Posts by Jesper Frimann

Oracle tag teams Solaris and Linux

Well

ripratm.

The problem is that the main install base for Solaris is in the same space as the 3U dell Hardware space. Lowend Unix.

The last marked numbers I have seen from IDC that talks about the lowend/midrange/highend UNIX marked are from 2008

And here 35% of the SUN UNIX revenue came from the Lowend, where as it was 11% for HP and 5% from IBM.

In the high end it was 21% for SUN and 34% each for HP and IBM.

In number of shipped servers 84% of SUN servers shipped in the Low-end, where as it was 64% for HP and 43% for IBM.

And someone posted the numbers of Highend systems sold by IBM and Fujitsu+SUN:

http://forums.theregister.co.uk/forum/1/2010/01/21/sun_oracle_approval/

"IDC Server Tracker WW Q309 - M9000 (Fujitsu & Sun)

2008Q2 2008Q3 2008Q4 2009Q1 2009Q2 2009Q3

183 129 221 159 165 90

IDC Server Tracker WW Q309 - Power 595 (IBM):

2008Q2 2008Q3 2008Q4 2009Q1 2009Q2 2009Q3

149 211 545 246 380 328"

So Larry has a problem in the high end, and is vulnerable i the low end.

// Jesper

IBM's Power7 servers imminent

Just a moment.

Damn what a lot of posts.

David Halko.

I used to dream of a SPARC 3 WorkStation. I grew up on BSD 4.2/4.3 and later SunOS and HPUX. So I know the dreams. I just got converted from being a diehard SPARC/HP9000 man to like POWER servers in the early 90ties.

Kebabbrain:

I'll deal with you later. There are more important things in my world right now. His name is Peter he weights 3020 Gram and is 53 cm long and was born the nigt to Monday.

Matt Bryant.

Jup, no big fan of the Mainframe and that bunch, as a Hardcore Unix Fanatic. They have kind of always been the arrogant bigbrother, who got all the cool stuff, and all the funds. But I agree with you one have to respect the Mainframe. Specially if one works with AIX and POWER cause there is a lot of Mainframe goodies

But it is not always a good idea to go off old Mainframe or the like platforms. I have seen some hardcoded HP3000 to Unix projects that didn't really turn out the way they should.

// Jesper

Kebabbrain (2 of 2)

....

And why do you all of a sudden choose a x86 server. Why not use the T5440 ? Why cause a 256 Thread 32 Core 256GB T5440 does a medicore 4720 SAP users on the same benchmark.

SUN T5440 with 32 cores and 256GB RAM

4720 USERS/18 users per thread/148 users per core.

SUN X4640 with 48 cores and 256GB RAM

10000 USERS/208 users per thread/208 users per core.

Power 550 with 8 cores an 64 GB RAM

3752 USERS/232 users per thread/464 users per core.

"The POWER6 is inefficient, it NEEDS 5GHz to match one Intel/AMD CPU. Let us clock POWER6 down to 1.4GHz and see how good it is then."

It will most likely be very good, and very efficient. But how come that all the benchmarks that the T5440 does it has many many times the Memory of any other system that it is compared to ? It's a factor of 3 on the above SAP benchmark, a factor of 8 on the SPECjAppServer(R)2004 benchmark you mentioned...

And what do you think that the Oracle price is for the DB on the above SAP benchmark as it is a 2-tier benchmark ?

It's bloody 24 licenses versus 8 for the power 550.. and they do cost 40KUSD+ a piece...

"So how about the POWER7? Will it also suck? It has great specs, yes. Maybe it will even be faster than Nehalem on some things!

<Deleted strange stuff>"

See the cache calculations above. 'nuff said.

"Regarding the consolidation of 251 Dell Linux servers down to 24 SUN T5220, even better. The T5440 has four CPUs. The T5220 has only two Niagaras. Probably it would suffice with just 12 of the T5440. One POWER6 have trouble beating one Intel or AMD. I bet POWER6 could not consolidate all these dual cpu intel Linux servers. Because POWER6 is good at single threaded work, but not on this work."

You have no idea of what you are talking about. There is a think called overcommitment on the POWERVM hypervisor. On my little serverfarm we run with a overcommitment factor of 3. Hence we use a 16 way POWER 570 as it was a 48 way server. On our internal machines, we use a factor of 5, hence a 16 way machine becomes a 80 way. machine. I don't expect you to understand, but that is how it is.

"Bull shit. We both know that Oracle will decide. Oracle has not presented any decisions yet. You are just rampling, here. Maybe there will be a Venus box."

Well lets see then. But they had better do a better job than they did going from 2->4 cores. On SPEC OMPL2001, the scaling was an amazing 18% for double the number of cores. And an amazing 13% if you correct for an increase in GHz.

http://www.spec.org/omp/results/res2007q2/omp2001-20070417-00252.html

and

http://www.spec.org/omp/results/res2008q3/omp2001-20080714-00308.html

"You need a radical design to circumvent this problem, which Niagara T2 does. It does not need large cache to be several times faster than POWER6. Nor high bandwidth."

BLEH it does it by magic powder right. The T2 design is good, but not ground breaking. And it still has less cache, less memory bandwidth per 'click' than POWER6.

....

"It reminds me of when I showed links that Niagara is 13 times faster than a IBM CELL on large workloads (string pattern matching). Then you chimed in and said something like "that is not relevant, the CELL is 70% faster on small workloads!". Jesus. Who is interested in small workloads where everything fits into the cache? Real life workloads are always big and dont fit into the cache. Hence you get cache misses, and it start to thrash. Which is a BAD thing. Which is evidenced that CELL looses performance really fast, down to less than 5% of it's former peak, when it's cache starts to thrash. Maybe POWER6 has equal behavior."

Heh, what does cell has to do with POWER6, they are two radical different designs. And what I said was that you cherry picked the results that fit you. And threw the ones that you didn't away. You didn't bother to paint the whole picture. You only mentioned the stuff that fit your purpose. Or rather that is giving you to much credit, cause you got it from some marketing material where you get all your claims from. Again as you do now. You paint everything black and white. Things are grey in grey, in the real world. And I have maintained the whole time that the T5XXX is an excelent box if you use it for what it is designed for. And I still mean that. You have to do research yourself. Stop echoing Snoracle marketing material.

"When you talk about Mainframes, yes the old customers keep bying - because of vendor lockin. But we talk about CPU performance, they suck big time. If you could use a Nehalem to run native Mainframe code, you would need 14 Nehalem to match 64 of the "uuuuuberfaaast" Mainframe CPUs. So new customers are not likely to buy them. Just look at the wikipedia article on "turboherkules"

You don't get it, there is a big difference when having a z10 running zOS with DB2 in Parallel Sysplex, with CICS and exits written in PL/1 or COBOL perhaps with embedded assembler, it all running on a pair of 32 core server with 95% average utilization.

It is almost impossible for any Power/Sparc/Itanium/x86 box to compete with this. Both with regards to throughput and availability. Try to think about why.

" This is not true. Niagara is growing much faster than SPARC64, something like 30-40% each year. So this is a pure lie. Again."

Yeah, and it's always nice to make predictions based upon a graph that consists of 2 points. Yeaaaahhh... go go.

"Yes, of course you have some T2 servers and some P570 servers. And of course the P570 servers are much faster. Yes. And on MY job, which is in a large investment bank/stock exchange/nuclear plant/pick your fancy, we see that one T2 machine is several time

s faster than P570 on multithreaded workloads. What I am trying to say, I dont believe you. I bet you are the person always writing that he works on a large investment bank/exchange/whatever and they are now migrating from his beloved SPARC to POWER. Just lies."

Oh, so you actually have a power 570 box at work ? What is the serial number then ? What is the output from a 'lsdev -Cc processor', not that we don't believe you...

"I have actually read here in the comments (now, this is actually true that I have read this) that there are 4 solaris machines shipped for every AIX machine. If this is true, then no way people are migrating from POWER to SPARC so much as falsely claimed here."

I think it was this that you refered to was someone that posted this:

IDC Server Tracker WW Q309 - M9000 (Fujitsu & Sun)

2008Q2 2008Q3 2008Q4 2009Q1 2009Q2 2009Q3

183 129 221 159 165 90

IDC Server Tracker WW Q309 - Power 595 (IBM):

2008Q2 2008Q3 2008Q4 2009Q1 2009Q2 2009Q3

149 211 545 246 380 328

But as to number of servers, then yes SUN is shipping more UNIX servers than anybody else. But most of those are entry level servers, and SUN is in free fall with regards to revenue. In Q3 2009 SUN's server revenue from servers running Solaris fell an amazing 39%, according to IDC, in the quater before Q2 that it was 40% and in Q1 it was 27%. You have HP and IBM slaughtering SUN in the markedplace, and it's going to continue. Why ? Cause SUN has cut so deep in R&D that it will take a long time for them to recover. And it's not like Fujitsu is stepping up.

http://regmedia.co.uk/2009/09/11/sun_sparc_roadmap.jpg

Do you really think that a Chip with just more threads now at 128 on a chip, at the same GHz, produced by a company that has never before produced a serious server chip (TSMC) will be enough to stop POWER7 and an enhanced Nehalem?

"I dont see the point in lying? I rely on official verifiable benches, you IBMers FUD and lie and twist everything. We discuss which CPU is fastest, you try to move focus away, to core pricing. Jesus. You always write the SAME thing: "work at a large bank/telecom company/etc which are loving SPARC and now migrating to POWER". No phantasy at all. The same lie all the time."

Ehh.. You don't rely on benchmarks. You haven't quoted one single official benchmark, you have quoted marketing sites, forums where people positive to your case have put their interpertations on the benchmarks. That is not the same thing. Again this might work in your little local backyard newsites or Boards, but not here.

"Why would I be embarrased because Oracle has the TPC-C record right now (via SUN gear)?"

Jup, and badly scaling cluster where they have had to only lease you the software to keep prices down, and only give you websupport. Again if you don't examine what you get when you put down the big bucks, you get cheated.

But I forget you have trouble reading TPC-C disclosure repports. How long do you think it will take before IBM puts out a benchmark using DB2 purescale ? It will blow the Snoracle benchmark out of the sky.

You, really should stop trying to play with the big boys, when you get easily bruised. And damn it takes 2 posts to adres all your SUN marketing FUD.

// Jesper

Kebabbrain (1 of 2)

Kebbabbbbbbrain this is getting tiresome.

"For instance, on SPECjAppServer(R)2004 benches we see that two Niagara machines with one 1.4GHz CPU each, gets 70% better performance than two IBM POWER servers. Just the cost of the licensed cores in the P570 used as the J2EE server is 1.6 times the list price of the two SUN servers, delivered, set-up and fully warranted for three years!"

http://blogs.zdnet.com/Murphy/?p=1537"

Again you don't even understand the benchmark. You wouldn't even be dangerous if you were clever. Again no clue about what you are talking about. Quoting murphy is like quoting BMSeer, that is where he gets his ammo.

Why don't you try to do some basic research about the benchmarks you are quoting rather than just quoting others. People who again get their information from BMSeer.

First this is the Benchmark where a T5XXX box has to shine, this is what it is designed for. So you cherry pick a benchmark Fair enough.

1) Java version the T5220 on the benchmark runs Java version 1.6.0_03 the POWER 570 runs 1.5.0. If you have a look at the JBB2005 benchmark you will see that for POWER going from 1.5->1.6 will give you a boost of aprox 25%. Perhaps even more.

2) You keep mentioning two power 570'es. Actually it is one POWER 570 and one p570. The p570 is an oldish POWER5+ machine.

3) Both POWER servers are of the 570 series. But none of them are more than a 1 CEC version. It's not a fully configured power 570/p570. It could just as well have been a POWER 520, the smallest non POWER blade server.

4) The T5220 machines uses 10Gbit network interfaces, the POWER solution uses 1Gbit hardware.

5) The T5220 database machine uses a the POWER database server actually uses 2Gbit connections.

6) The T5220 Database server uses 64GB of RAM, the POWER p570 used 8GB.

7) The T5220 APP server uses 32GB of RAM the POWER 570 uses 16GB of RAM.

8) The response times on the two systems is very different, on the POWER setup the MAX request time is only 43% of what it is on the T5220 setup. For example the Maximum response time for Browse is 22+ seconds on the T5220, which is 5 times worse than the POWER setup.

9) If you knew just a little about UNIX you would quickly see that the people who setup the T5220's have done a much better job than the ones setting up the POWER servers. The most basic network and AIX java tuning hasn't been done. It's right in the manual:

(http://publib.boulder.ibm.com/infocenter/aix/v6r1/index.jsp?topic=/com.ibm.aix.prftungd/doc/prftungd/java_tuning_aix.htm)

So ... yes... sure you can compare but please stick to the facts.

"If you are discussing licensing prices, then your claim is relevant. But my post is about the performance of the POWER6 and POWER7 vs Niagara."

Eh.. I try to educate you, it is hard, cause you constantly resist knowledge. But HW cost is only a small portion of the cost of a system. Let me cut it out in cardboard for you. Perhaps that will help you.

It's like buying a car sold to you by SUN that cost 1000$ but will only run 10 mile to the gallon, compared to buying a 10.000$ car from HP/IBM that will run 50 miles to the gallon. Which is kind of a bitch if you have to drive 40.000 miles a year and the gas is 8$ a gallon. Because after driving 14000 miles the cheaper SUN car starts to cost you money.

Yes there are people that does this kind of thing all the time, they also have massive credit card misuse and a lot of unpaid bills.

"Regarding if POWER7 is a legacy designed POWER6 derivative or a POWER5 derivative, you posted a link. I have skimmed that link, but I would be glad if you pointed out on which page it says the POWER7 is a POWER5 derivative."

I wrote it in my post. It's a question about filosofy. POWER6 is a speed devil processor, much like POWER4. POWER7 is a braniac, just like POWER5. POWER6 is an mostly inorder processor, POWER7 is a OO processor.

Regarding the $17 million cost of the IBM P595 system that earlier had the TPC-C record. I see that all clients + client's software adds upp to 0.5 million. I think it is safe to say that the rest of the $17 million sum, is only for the P595 server then. I mean, that is quite a wicked pile of money, dont you think? $16.5 million!! For one sucky POWER P595 server!!! Who spoke about

SPARC having large margins? *chuckles* Actually, without the discounts, ONLY the P595 server including it's sofware costs $35 million!!!!!! "SPARC large margins" - eh? :o) :o) $35 million! Jesus! (Thanks for the link :o)

Begezus, you can't even read. Man and me who is trying to convince the wife to buy a house in sweden. Hmm.. Nahh.. Swedish education system is good I know it.. I choose to belive in my swedish roots. You just turned to the last page and didn't understand what it said.

Now if you want expensive then try to look at your precious SUN TPC-C submission:

http://www.tpc.org/results/individual_results/Sun/Sun_T5440_TPC-C_Cluster_ES_122309_v2.pdf

On page 2, they haven't even let you buy the software.. you only lease it.. and the support is Metalink only, no upgrade protection, no nothing. Just lease of the software. And for that you pay 2.6MUSD each year. It's the 10 miles a gallon car all over again. Buy some terrible entry servers and put expensive software on them.. yeah.. it's money out of the window. And you don't get

it. The Oracle software costs just as much to lease for a year, as DB2 costs in five years if you buy the software, and that is with full support.

Now here is the real reason why Oracle bought SUN.

"You just dont know that a high Hz does severely punish cache misses?

http://forum.beyond3d.com/showthread.php?t=54889"

Ever heard about SMT ?

And if you think that 2 threads running at 5GHz is bad for the caches ?

How do you think that 8 threads running at 1.6GHz with even less cache ?

Let os again do the math.

One POWER core has 128KB L1 cache 4096 KB L2 cache and 16384 KB L3 cache. This means that it has:

12.8KB per thread/GHz L1 cache

410KB per thread/GHz L2 cache

1638KB per thread/GHz L3 cache.

One Niagara core has 24KB L1 cache 512 KB L2 cache and no L3 cache. This means that it has

1.9KB per thread/GHz L1 cache.

40 KB per thread/GHz L2 cache

0 per thread/GHz L3 cache.

Ever heard the expression that you should throw stones when you live in a house made of glass ?

"<Cut out some strange ramplings that didn't make sense>

Also, on SAP we see that AMD Opterons is faster than 5GHz POWER6

http://www.hpcwire.com/offthewire/Sun-Microsystems-Sets-Records-with-New-x64-Server-78779357.html

Eight AMD opterons is 2.7x faster than four POWER6. A very expensive POWER6 can not justify it's much higher price against one AMD opteron."

Again you can only figure out how to read marketing announcements from SUN. Damn.. I have to ask.. I have a large steel construction in Paris that I want to sell.. it's kind of a tower, it is very nice, you want to buy ?

It's 48 cores and the only really suitable benchmark to compare it to is the power 550 which does around 40% of the 10.000 users with a 1/6 of the cores. So it's still 2.5 times faster per core. But ofcause it starts to fall behind now as it only has 2 cores per chip. Nothing strange about that.

And again nice cherry pick, take the newest just announced x86 AMD server and here in 2010 compare it to a machine from 2007, with it's replacement being announced right around the corner. A POWER7 box with half the sockets will outperform the X4640 by a factor of 2. Again cheap marketing shot, without any real thought behind it.

You are stating the obvious, that given the time with twice the number of sockets and 6 times the cores a SUN x86 system can outperform a POWER system.

...

What the...

"On SIEBEL v8, three IBM Power P570 servers with 12 of the 4.7GHz POWER6 CPUs is half as fast, as one SUN T5440 with 4 of the 1.4GHz Niagara T2. So assume you double the amount of POWER6, then you will get 24 of them. And then finally, those six P570 will match one SUN T5440. One P570 costs $413.000 and one T5440 costs $76.000."

Yes, we do think, yes many of us do have university degrees in Math, Computerscience, physics whatever. And many of us do have many years of experience in the business, opposed to you. Now I hav already on this forum, made a walkthrough of the Siebel bench

mark you are mentioning and showed you that yes, the T5440 is a very good one. BUT no where in any way just nearly as good as you keep trying to make it. You are maing a fool of yourself. Sorry.

"...So basically Niagara has a slim lead on POWER6 only when comparing chips to chips..." Are you joking? Who does NOT compare a CPU vs another CPU? Who compares ALU vs ALU? No one. If you try to establish which CPU is fastest, isnt it appropriate to compa

re... CPU to CPU? Or should we instead compare ALU to ALU, and then infer which CPU is fastest? Are you joking? It is very relevant. Why ?

CAUSE YOU USE CORES WHEN YOU DO CALCULATE SOFTWARE LICENSES!!! HOW MANY TIMES DO WE HAVE TO TELL YOU.

And thread level performance is also important, cause single threaded throughput is still a very important factor. One factor that all to many CPU's aren't really doing enough to adress IMHO. Damn.

Try doing some math here:

http://www.oracle.com/corporate/contracts/library/processor-core-factor-table.pdf

or

http://www-01.ibm.com/software/lotus/passportadvantage/pvu_licensing_for_customers.html

"Regarding the POWER7, I am asking if the POWER7 will also achieve only one third of it's theoretical performance just as POWER6 does. Is that not a fair question?"

*SIGH* Sure it is, but stop mixing the benchmarks up and using GFLOPS when talking about specfp_2006rate. It makes no sense. And you still don't understand virtualization. Which is perhaps the most important thing forcing up utilization and thus making sure that a system is well utilized.

"Also, I have heard that POWER7 is basically a couple of stripped down POWER6. But you claim it is more similar to a POWER5? Do you have any links so I can read more on this?"

Yes, if you read this presentation:

http://www.ibm.com/developerworks/wikis/download/attachments/104533501/POWER7%2B-%2BThe%2BBeat%2BGoes%2BOn.pdf

Note that POWER7 is a OO core, with no less that 12 execution units.. 12.. it's a fat core. It also has 4 way SMT, not coarse grained multithreading like Niagara.

"Also, earlier I read that the IBM P595 which had the earlier TPC-C record costed $17 million. But you claimed that cost was not

only the machine, but everything (clients, etc). Do you have any links on that, too?"

Yes, it's here:

http://www.tpc.org/results/individual_results/IBM/IBM_595_20080610_ES.pdf

Mattie Pattie Laddie,

Could you try to explain to me, what is the point of scalability when the CPU sucks? Why all the fuss about scalability? I mean,

I prefer a few CPUs that are several times faster than lots and lots of slower CPUs, that still can not catch up! Who does not

agree? If you can use 4 CPUs to match lots of other CPUs, which solution do you choose? Obviously the slower solution with many times more CPUs?? But hey, maybe you do. If you can claim something weird like "in my opinion that 2m guy is SHORTER than the ot

her 1.5m guy", then you surely can choose a solution which uses way more CPUs, and costs 10 times more! Strange logic.

The whole point is that it doesn't suck, power6 is damn fast. The project I am responsible for has no less that 16 power 570'es and a few p595's. And damn they are fast, and actually quite forgiving due to their high Ghz. Again if there is a CPU core that sucks then it is the T2. Now both Itanium and POWER has managed to keep One of the reasons why Itanium still sells fairly well is that it's single threaded performance actually is pretty good. And it will, IMHO, get better with Tukwila. And you still can't get it into your thick head that the one machine with the myriad of CPU cores is the T5440. It has bloody 32 cores with no less than 256 threads. Man that is half the threads of a maxed out M9000.

.

"You know, SUN migrated 251 Linux DELL dual cpu servers, than run 700 instances of MySQL down to 24 SUN T5440 machines. I bet you can consolidate way more than 240 old IBM servers on a modern SUN machine, maybe M9000. Dont you think?"

yeah Yeah, you keep quoting that story everywhere U go. First it is something you picked up from BMSEER here:

http://blogs.sun.com/bmseer/entry/mysql_consolidation_on_sun_sparc

And you couldn't even get the machine right, it wasn't a T5440, it was a T5220. Do you bloody research before wasting our time.And just quoting BMSEER isn't really particular serious, ever wondered why he/she isn't blogging anymore ?

He/she got fired, cause Oracle took over the business.Now you don't see me quoting, without using my head, some of the IBM/HP/DELL marketing slides. And the M9000 has even worse per

core performance with the quad core chips, compared to the T5400 when using large Core count. Damn..

Anyway, I am just trying to say that I would like to see benches and real numbers on the POWER7. Just like all you do. And see if it achieves as high as one third of it's theoretical numbers, just like the POWER6 does. Then I would like to pit it against the UltraSPARC 8-core Venus and see how much faster the Venus is. Let us continue when we see first benches on the POWER7. Let us see who is right. (Maybe IBM will refuse to publish benches, just like IBM refuses to publish benches on their dog slow Mainframe)

Again you are just rampling. There will be no SUN Venus box. Read their roadmaps. And yeah lets see the benchmarks and lets see a 128 Node Power7 box benchmark with DB2 Purescale.. now that would be fun. And be sure that POWER7 will blow anything that SUN can come up with out of the water and still have plenty of firepower left. Just wait and see.

"The point is, a traditional legacy 2.5GHz CPU idles 50% under _full load_ because of cache misses. A much higher clocked CPU, will be punished much more severely by cache misses. It will idle even more than 50% under maximum load. That is the reason a 5GH

z POWER6 sucks, because 3 to 4 of the GHz vanish, they are spent on waiting on data - even under max load. Sure, the theoretical numbers may look great, but hey, on a high clocked CPU you will NEVER reach them. As evidenced by POWER6. One third is just really really bad. POWER7 is also legacy construct. On the other hand, the Niagara T2 only idles 5-10% under full load, all its clock cycles are doing useful work. Why do you the much slower clocked Niagara T2 can outclass a much higher clocked CPU, on certain work loads? Because of efficiency. That is why. T2 is much more efficient than the old legacy designed CPUs. No use of big cache, nor complex prefetch logic. T2 is a small lean fast CPU. Which is evidenced by benches."

Again you have no idea what you are talking about. Have you ever made a solution or even been logged into one of the boxes you are talking about ? or are you just been overdosing on BMSEER. No wonder why the rest of the posters on this page, simply ignore you.

And yes the T2 is efficient.. it is .. it is a GREAT CPU for running many light independent threads which share the same code, so you can benefit from shared libraries.. but.. well... Xeons are just better at that. Sorry. And much chearper.. and has more RAS features and and and..

// Jesper don't wanna waste more time.

Sigh...

Kebabbert. SPECfp_rate2006 is has nothing to do with GFLOPS. As the benchmark does not measure FLoatingpoint Operations Per Second, hence FLOPS.

And POWER7 is perhaps more like POWER5 than it is to POWER6 in design strategy.

And the predictions on how much faster POWER7 is compared to POWER6, then the factor that has been thrown around are somewhere between 5-5,5 times per chip i 'chip performance'

1 Thread:

SPECfp_2006:

POWER6 24,9

T2+ n/a

SPECint_2006:

POWER6 21,6

T2+ n/a

32 Threads

SPECfp_2006rate

POWER6 544,0

T2+ 68,5

POWER6 is 7,9 times faster

SPECint_2006rate:

POWER6 542

T2+ 97

POWER6 is 5,6 times faster

64 Threads

SPECfp_2006rate

POWER6 602,0

T2+ 133

POWER6 is 4,5 times faster

SPECint_2006rate:

POWER6 832

T2+ 183

POWER6 is 4,6 times faster

128 Threads

SPECfp_2006rate:

POWER6 2184

T2+ 270

POWER6 is 8.1 times faster

SPECint_2006rate:

POWER6 2155

T2+ 360

POWER6 is 6 times faster.

1 Chip

SPECfp_2006rate

POWER6 58,0

T2+ 68,5

Niagara is 1.2 times faster.

SPECint_2006rate:

POWER6 60,9

T2+ 97

Niagara is 1.6 times faster.

2 Chip

SPECfp_2006rate

POWER6 116

T2+ 133

Niagara is 1.1 times faster

SPECint_2006rate:

POWER6 122

T2+ 183

Niagara is 1.5 times faster

4 Chips

SPECfp_2006rate:

POWER6 222

T2+ 270

Niagara is 1.2 times faster

SPECint_2006rate:

POWER6 263

T2+ 360

Niagara is 1.4 times faster

8 cores

SPECfp_2006rate

POWER6 222

T2+ 68,5

POWER6 is 3.2 times faster.

SPECint_2006rate:

POWER6 263

T2+ 97

POWER6 is 2.7 times faster.

16 cores

SPECfp_2006rate

POWER6 544

T2+ 133

POWER6 is 4.1 times faster

SPECint_2006rate:

POWER6 542

T2+ 183

POWER6 is 3 times faster

32 cores

SPECfp_2006rate:

POWER6 602

T2+ 270

Power6 is 2.2 times faster

SPECint_2006rate:

POWER6 832

T2+ 360

Power6 is 2.3 times faster

So basically Niagara has a slim lead on POWER6 only when comparing chips to chips, and at ithe best result is a 60% lead and that is only against lower clocked POWER6 on SPECint_2006rate. On specfp_2006rate (16 way) each 8 core, 32 threaded T2+ chip is only 20% faster than 2 core 4 threaded POWER6+ chip.

Now you think that a POWER7 chip with four times the cores 8 times the threads, and fatter cores are only going to match a T2+ chip ?

Sorry but you are not being serious.

// Jesper

Oracle to detail delayed Sun roadmap

RE:Ivan Hallworth

1) Well FJ have made better servers than SUN, for quite some years IMHO. But they are still very small in the Server Marked compared to HP and IBM. And with SUN loosing SPARC revenue in the range of 25-40% in each of the last 3 quaters according to IDC, th

en it's not like things are looking bright. But there is another area where FJ is suffering, and that is in the Virtualization area. They are heavily relying on SUN here, who has been cutting R&D like hell.

And it is here it starts to get nasty. Cause if you are able to overprovission a POWER7 power 595 server with a factor of 3, a server which is already 4 times faster per core in the POWER6 version (For example on SAP R/3 SD 2-Tier M9000 versus POWER 595 http://www.sap.com/solutions/benchmark/sd2tier.epx), then you are really up the sh*t creek. Sure you can to a certain extend use Solaris container, but they do not provide the same isolation as a seperate virtual machine.

2) You cannot have both cheap server, high performance and high RAS, good profits at the same time, as a Server vendor. And right now the front of the pack with regards to general business servers, are powered by Nehalem and POWER6. With Nehalem having the

best per socket throughput (as it is a quad core) and POWER6 the best per core throughput. THis will most likely swing to POWER7's advantage when it becomes available. With 8 core chip version of Nehalem close behind.

I must admit I fail to see SPARC Venus being a game changer. POWER6 is already a factor of 4 faster per core/factor of 2 faster per chip. And FJ servers haven't got the best reputation for scaling on benchmarks with lots of cross board communication.

For example have alook at the

M9000 128 Threads 128 Core 64 socket@2.4GHz result: 1230446

http://www.spec.org/omp/results/res2007q2/omp2001-20070417-00252.html

That is:

4005 per thread per GHz of the CPU

4005 per core per GHz of the CPU

8011 per socket per GHz of the CPU

M9000 512 Threads 256 Core 64 socket@2.52GHz result: 1456653

http://www.spec.org/omp/results/res2008q3/omp2001-20080714-00308.html

1129 per thread per GHz of the CPU

2258 per core per GHz of the CPU

9032 per socket per GHz of the CPU

So doubling the number of cores and the number of threads only gave a medicore 13% more on a Board Machine Destructive workload.

I know this is a Technical benchmark, but it is one of the better to illustrate what happens when you run some very board unfriendly heavy memory access workload on a FJ MX000.

6. Nahh.. Not Mainframe. I don't agree on that.

// Jesper

Well

1. Yes, that is perhaps the most important thing talking for SUN. A large install base, and a pretty fanatic one also. As you can see on this board, they are perhaps the most fanatic Unix crowd, and can sometimes be beyond normal reasoning. Have a few college that are like that.

2) Well what options did Fujitsu and SUN have ? They could have their chips made by Intel or IBM. What has to be seen is can TSMC manufacture a chip that is equal to Tukwila, Nehalem and POWER7 in Raw compute power, reliability and operating temperature ? There is a big difference in making Grahpic Chips for NVidia to go in Gamer PC cards, and to making highend 'must not fail' server chips... a big difference.

Yes many are saying that, we just need fujitsu's venus 8 core chip, then we are equal with the rest But try to have a look at this picture of a watercooled Venus board with 4 chips and 32 cores.

http://regmedia.co.uk/2009/11/25/fujitsu_venus_board.jpg

and compare it to a 4 chip and 32 core POWER7 MCM chip.

http://regmedia.co.uk/2009/11/26/power7_switch_packages.jpg

and a POWER7 power 575 node with 32 chips and 256 cores. (also water cooled)

http://regmedia.co.uk/2009/11/26/power7_ih_node.jpg

It's a whole different league.

3) Both HP and IBM from what I understand.

4) No matter what anybody say. YOU do NOT gain any Business Value or higher ROI as a customer by having to pay for 24 licenses rather than 8. (Niagara T5440 versus Power 550). You simply don't. What you do gain Business value on is using a Enterprise version of Oracle rather than a standard edition. But giving your money to the software supplier and buying inferior hardware is just bad for business. Sorry.

5) Sun makes great x86 gear, or so I am told :)= I am a Unix guy. But their volume is tiny compared to the other big players.

6) Again, it has been to easy for them, kind of reminds me of the Mainframe sales people of the 80ies. But times have changed, the bean counters are reaching higher up in the trees to pick fruits. And software licenses is one of the ones where you really really can save money.

7) I think that a marriage of AIX and Solaris would have resulted in them both getting married to the penguin in the end. Now that would have been sweeeet. A linux with all the AIX and Solaris gutties. Yum Yum. And man that would have been bad for Microsoft Windows.

8) The most important thing here is perhaps that we will have 3 not 2 big 'we can deliver it all' vendors. And competition is good, real good. So no arguing there. I am just not sure that it was such a good deal for SUN to be swallowed by Oracle. Fujitsu or IBM would have been better, for the customers but perhaps more expensive due to less competition.

// Jesper

RE:Ivan Hallworth

I am sorry I don't see what I associate with SUN, their SPARC servers running Solaris, making a comeback any time soon, if ever.

SUN has cut so deep in R&D that development of SPARC processors is more or less reduced to the Niagara line which is a niche product.

Rainbow Falls is going to be manufactured by TSMC. A company that has never made a serious server chip before.

And it is basically a 16 core chip each with 8 threads, that all have to be active at the same time to get max throughput. This means that a 4 socket box have to utilize 512 threads. It simply does not make sense. And it will be no match for Intel/AMD x86 products and IBM POWER servers.

So I predict that in 3 years Snoracle hardware will be almost pure x86.

And Larry might think that it's a good idea to give away Coolthread servers for free if they use Oracle.. Sure.. it's like giving away Ford F-350 pickup trucks when you are selling gasoline to the same people at 100$ a gallon. Ok, oracle licenses are way way more expensive.

Customers aren't that stupid, here in Denmark Computerworld is currently writing articles daily on customers that are having Oracle sales people come to their door trying to squeeze more license money out of them.

One place where I used to work, got a bill of 4MUSD, cause they merged with another company, I know the systems, I helped design the solution. I know the Oracle license rules so there is nothing to come for. And they will be moving to DB2 on SAP, I talked with some of my old colleges. Simply cause they have had enough of Oracle's arrogance.

Then we just have to hope that IBM doesn't pull the same stunt on them, but I don't think so as they will be buying everything from SAP.

And it is with sad eyes that I see Oracle destroying SUN, if the SUN IBM deal had gone though Solaris would have lived on for quite some time before merging with AIX to become THE UNIX OS, that would have been grand. I grew up on HPUX and Solaris at the university.

I also have friends who work at SUN, and they are scared. And it is sad, it is.

// jesper

Sorry

In a time where the once great company SUN have been mishandled and tens of thousands of SUN employees, are loosing their jobs.

I must admit I find it offensive and bad taste that you promote an Oracle sponsored sport teams, like Americas Cup that is estimated to cost as much as 60 million Dollars to participate in.

If Oracle had chosen to give up that sponsorship, hundreds of SUN employees could have kept their jobs.

I know that it's lonely at the top, and hence you sometimes loose a bit of perspective, and I know that wearing a tie might limit the blood circulation to the upper most part of the body. But your attempt to take focus away from the terrible layoffs by promoting a upper class sailing event is IMHO distasteful.

I can only hope for our fellow IT people that there are other companies like HP, Fujitsu and IBM that need your excellent skills.

// Jesper

Guinness to hit three quid a pint

Brew it yourselves

This can only be seen as an excellent opportunity to take up brewing beers by yourself.

Tastes a hell of a lot better and a lot cheaper also.

// Jesper da brewer.

Neon sues IBM over 'anticompetitive' mainframe tactics

Well it's not like it's any different than...

lets say you bought a SUN server with a hefty graphics card. You have to do serious graphics presentation of data that is located in a Oracle database, on the same machine.

And lets say that a smart company then let you offload certain parts of the Oracle code to run in the graphics card, which then means that you only have to buy half the Oracle licenses.

1) Would you expect to get support when the graphics card corrupted your data ?

2) Would you expect Oracle to get on your butt and try to charge you extra licenses as you are basically circumventing their license rules ?

3) Would you expect Oracle to void any support you would be able to get ?

and and and ...

That is what the problem is all about.

// Jesper

IBM: Power7 to rollout throughout 2010

RE: David Halko (I give up.)

If you can't see it, you can't see it :)=

// Jesper

RE David Halko on Ellison whips out his Sparc TPC-C test

Damn, thread "Ellison whips out his Sparc TPC-C test" closed and I just used time to write up an anwer to David... well he's on here so here it comes out of context:

David Halko wrote:

"Good applications, however, are not always available for the architecture that may suit it best. The market takes care of the latter situation, over time."

Look, SUN tried to jump ahead of the other server vendors by using a lot of threads, and they will continue down that road. Victoria falls has what 128 threads per chip, that's just as many as many as a M8000. So to get the maximum throughput out of a single chip you will have the same thread complexity, with locking, memory use, monitoring and amdahl law issues, as you have on a complex Big Tin server, which is many times faster.

"I covered the main drawback above in a previous quote. I can add additional arguments (production line manufacturing of a single-chip solution is normally less expensive than an MCM - the consumer market size drives the most profitable implementation.) The AMD-Intel battle for quad-core illustrated it best, in the most recent near-term. If I need me to explain this, I can. I am really not trying to pick an argument - technology is what it is and I appreciate it as it is."

No your original comment was "Some may argue that the MCM's were innovative engineering, disagree and suggest MCM's are a pragmatic business-man's short term solution to a technical problem." That is not "technology is what it is and I appreciate it as it is.".

And I'll just repeat myself. It is not either CMT or MCM, you can use both. Sure it is faster, cheaper and better to put it all on one chip, but when you've filled the chip up, then an MCM's is next best thing, rather than having to use more sockets. From what I've read here at 'The register' then the next POWER7 will just as POWER4 use MCM's to make high density servers with 2 socket servers that houses 2 MCM each with 4 chip that each houses 8 Cores. Now that will give you 64 cores in a 2U box. Not a general purpose server btw, but a HPC box. And witout MCM's I doubt that that box would exist.

If you want to say that something is business man like, then it is that IBM _didn't_ make an MCM version of the POWER6, it's not like there weren't enough memory bandwidth in the sockets.

"Sacrilege is not the issue, accuracy is. Also, Solaris and SPARC are historically based upon community efforts, as Open Communities, which have been guided by Sun, external companies, and external organizations - so inaccurate information used in a slanderous way offends many people who invested their university, research project, masters, phd, and/or life works into it. Offense is to be expected when inaccurate information is used to slander large groups of people."

Yes I know, at the university I attended, VAX and HP was for the students and SUN was for the researchers and PHD's. But it still doesn't change that it's bloody anoying when you get people in projects that base their decissions on feelings and not cold facts. This does not have to Solaris followers, but might as well be mainframers or Vaxers or windowers or Linuxers.

"Some would suggest that the behavior merely objectifies a group of supporting individuals and slanders their life work. There are lots of people who feel they can slander groups of people calling them "blood sucking..." (fill in the blank) - it is just the same behavior. It shuts down inter-group dialog instead fostering healthy inter-group competition and later teamwork."

Well you don't live in Denmark now do you. Here insulting/making fun of people and then yourself (or their work or..), is kind of of a accepted way of bringing people to the same level. Rather than praising others and then yourself almost as much, you 'make fun' of the other and then yourself. Like 'What an ugly dog you have, but it's not as ugly as mine.', could be a good ice breaker. Odd yes I know, but quite true.

"(i.e. OpenOffice, VirtualBox, NFS, Lustre, Xen, etc.) When you consolidate it all together, Sun contributes a lot."

Then don't call it linux, OpenOffice VirtualBox .. are all applications that run on pretty much every platform, or can be made to do it. Now if you had said that SUN offers up opensource great applications, I would just have agreed with you.

I use virtualbox myself and absolutely love it, for it's simplicity and usefullness. I run OpenOffice and I like it, I sometimes get a little fustrated cause it isn't as stable as it used to be. I feel very strongly for open document standards and is actively engaged in promoting these.

But these applications are not a part of the operating system called Linux. Period.

// Jesper

Again Bigbrother cause something is wrong in the state of Denmark.

RE:Kebabbert

I'll just repeat myself:

"So again this is a case of not telling the whole story, which is that For small dictonaries Cell is faster for large ones Niagara is faster. And that there is quite a good reason for this, the two processors are targeted at different markeds/workloads.

But cause you are picking a slam dunk x10 benchmark from a SUN FUD site then you get burned."

I don't dispute the SUN benchmark results. I just point to the fact that they aren't telling the whole story, but cherry picking where they are best.

"Sorry no points for quoting BMSEER, this is SUN FUD central, And you get no points for quoting HP the real story or 'IBM's pseudo business technology statements' either."

Jup, I have no problem quoting SUN/HP/DELL/IBM/Oracle technical sites, there are a lot of serious good people working at all these companies.

But BMSeer is a FUD site just like the others I mentioned. Sorry.

// jesper

Adam 61.

"Citing benchmarks is fine - but how realistic are they in the context of what people do on a day to day basis?"

Well you are absolutely right, and well.. the problem is that much sizing information is actually derived from benchmark results. If you for example take the SAP sizing data you can get from different hardware vendors then it is usually based upon SAP SD benchmarks. So in my day to day I indirectly use benchmark information, all the time.

And we don't have the time to actually go and try things out, which certainly would be the best, we do a qualified guess and then add a little more MEM and CPU and see how it goes, we can then always deduct a little afterward.

"However I also tire of his constant references to "Slowaris", "

Well that is a part of the game. I always use the therms SlOwARIS, HPsUX, AIntuniX and Toonix. And not to forget Wintendo. Then you are sure you get to insult everyone, which is a very danish way of doing things, you start by insulting people and then you insult yourself to put you on the same level as them. Strange yes yes.. I know.. :)=

And well a bit of teasing is ok, I mean it's not like people are seriously saying that Solaris is not a Unix, or not a fairly good one. Ever heard the amount of sh*t that AIX admins have to take about AIX, specially from Solaris admins ?

If Slowaris should be an insult then it should be to the SPARC platform.

// Jesper

RE:Kebabbert

"I forgot to post the links that show that 1.4GHz Niagara is more than 10 times faster than a 3.2GHz Cell. Here it is:

http://blogs.sun.com/bmseer/entry/extremely_fast_pattern_matching_on"

Sorry no points for quoting BMSEER, this is SUN FUD central, And you get no points for quoting HP the real story or 'IBM's pseudo business technology statements' either.

You write that it might not be right to compare Cell to Niagara. And i agree the first is a HPC/gameconsole processor, the later is a Webserver/light application server product.

Again this is a cherry picked benchmark.

What our good friend BMSEER has done is to point to the place where Cell doesn't perform well on pattern matching. Actually for small dictonaries a single cell, using it's SPE's will do around 41 Gbit, which is 1.7 times what the Niagara processor will do.

But this isn't really mentioned. Here is a presentation connected to the paper that BMSEER referes to:

http://sti.cc.gatech.edu/Slides/Petrini-070619.pdf

So again this is a case of not telling the whole story, which is that For small dictonaries Cell is faster for large ones Niagara is faster. And that there is quite a good reason for this, the two processors are targeted at different markeds/workloads.

But cause you are picking a slam dunk x10 benchmark from a SUN FUD site then you get burned.

// Jesper

RE:Hopefully By Kebabbert

Kebabbert.

You should check the facts,

1) The two benchmarked solutions actually run somewhat different mix of transactions, the p570 runs more Partner Relationship Management transactions which, judging by the response time are the heaviest.

2) You have to adjust for utilization of the machines as stated. The power 570'es run at low utilization on DB and Web, where as the T5440 runs at 80+ utilization.

3) The T5440 consumes quite a lot of memory 94 versus 46 GB

4) Opposite the recent TPC-C benchmark, then the response time for the p570 are much faster than the T5440

5) The T5440 runs more or less in a 2-tier mode, where as the p570 uses physical network.

if you try to adjust a little due to 1) then I get to the fact that the T5440 is aprox 2.4 times faster per chip on this particular benchmark. And this is what the T5440 does best, webserving, light application serving and a small database that can run on a few cores in this case 3 cores/12 threads.

// Jesper

Ellison whips out his Sparc TPC-C test

RE:Afara, Multi-Chip Modules, Linux

"Jesper posts, "I am tired of having fanatics trying to get me to use these servers in my designs where it isn't appropriate."

Join the club. I feel the same way about various other servers."

My whole point was that I don't think that Niagara based servers are general purpose servers.

So you are saying that you don't feel that one of more of Itanium, POWER, Xeon or AMD based servers are not general purpose servers ?

"Doubtless, there are many benefits to multi-chip modules, as well as drawbacks"

And the drawbacks are ?

And originally you said:

"Some may argue that the MCM's were innovative engineering, disagree and suggest MCM's are a pragmatic business-man's short term solution to a technical problem."

Short term solution ?

MCM modules have been used in mainframes since the 70ties.

Intel used MCM's on the PentiumPro, and later on Presler,Dempsey,Clovertown, Kentsfield and Yorkfield. And who knows they might do it on Itanium also.

IBM used POWER2,POWER4[+],POWER5 [+],POWER6[+] and POWER7.

And from what have been written here on the register then we are gonna see a 4 chip MCM power7 module. Now that is going to be a 32way MCM module which you can have in your hand

like this: http://upload.wikimedia.org/wikipedia/commons/1/1f/Power5.jpg

And that MCM module will most likely have the several times the Umpf, than a current M8000 does. Hence such a module will have all the busses, all the infrastructure and part of the backplane that goes into the CPU part of seven or eight CMU boards of a M9000.

Oh no that is not innovative. I suspect that the only reason you say it's not innovative is cause SUN doesn't do MCM's.

"Anyone can have a good idea, but when there are technical hurdles, it takes innovation to overcome them, to bring an idea to fruition. It also takes an innovative spirit in order to select a good idea to bring it to fruition."

IMHO there is nothing wrong in that, large corporations like HP, IBM and SUN do it all the time.

But it still doesn't change that the original innovation was done by another company.

My problem is that when people say things like this, it treated as a sacrilegious act, by the followers of the SUN. Sometimes I feel like Solaris and SPARC are religous icons, when speaking to the followers of the SUN.

And I can say SlOwLARIS just as well as I can say HPsUX and AIntuniX, TooNIX which is what the OS'es are known at where I work. It kind of softens things up between the different fractions.

"Sun is in the top 30 corporate contributors, according to the August 2009 update from the Linux Kernel Development"

Ok, found it, so this year they made it into the top 30. Wheeee!

Hardly impressive when compared to their main competitors INTEL/HP and IBM.

But still it doesn't change the fact that the number of people at SUN that are maintaining a part of the Linux kernel is one. So it's not like SUN has invested a great deal of effort in Linux.

So your original statement:

"The Linux community looking at Sun in a negative light for "buying the rights" to open-source Solaris seems rather myopic. Sun has always been a significant Linux contributor, contributed A LOT of their own code & engineers into Linux, and ships a significant number of Linux platforms."

Still doesn't stand. It is simply not true, period.

// Jesper

PS: And I think it's very sad that SUN has to let more people go.

PPS: Bigbrother cause something is rotten in the state of Denmark.

RE:David Halko

"Afara did not have a piece of SPARC silicon when Sun invested in them. Sun brought SPARC CoolThreads into fruition. Almost a half-decade later and Sun is still the most significant octal-core CPU vendor on the market." (octo-core) I guess SPARC still operates in binary :)

We were talking about inovation. So you would agree with me that the actual inovation behind the TX000 and came from Afara. SUN did what you should do as a big company that buys a small inovative company, they turned the inovation into a product. And as I have said again and again, there is nothing wrong with the Niagara based servers if you use them for what they were intended for. But they were not designed for running a moderate number of heavy independent threads that do a lot of cross communication between the threads.

So, the reason why I am p*ssed, is that I am tired of having fanatics trying to get me to use these servers in my designs where it isn't appropriate. And the worst part is that I've lost some battles, and most of those solutions now have to be redone.

"While Sun was doing the heavy-lifting by engineering heavily multi-core and multi-threaded processors into a single piece of silicon, others were cobbling together multi-core using multi-chip modules. Some may argue that the MCM's were innovative engineering, disagree and suggest MCM's are a pragmatic business-man's short term solution to a technical problem."

I think you have missed the point all together. It's not multicore | MCM. It is multicore & MCM, if you have read some papers on MCM technology, you would see that there are quite a few benefits.

The MCM module that have been in the press the most, besides those from the mainframes, are the POWER4 and POWER5, which basically allowed you to plock 8 cores into a socket. Now nobody has claimed that those were in a single socket, or that IBM somehow beat SUN to the octo core by using these MCM modules. On the other hand the Quad core modules found in the power5+ based p5X0Q servers, did plug into a standard pSeries socket. And you might call that for a business man solution, but damn it worked, both Oracle and IBM counted it as being one socket. So a box like the p560 could run Oracle standard edition, just as the T5440 can today. And you might make fun of MCM's but for example the 505Q servers were just as fast as the T2000.

You make the creation of the Niagara chip sound like it a revolution. Hey, the original T1 was 8 SPARC II cores, that each rans 4 threads in a statically scheudled round robin multithreading mode. Clever and a good piece of ingeneering. But not a revolution.

Now as for the whole Linux/SUN/SCO/IBM deal...

"Had Sun been trying to assault Linux, code contributions to Linux from Sun would have ceased contributions (instead of continuing to increase Linux contributions), and Sun would not have paid to protect their Linux customers."

Ehh what code contributions to Linux ? I think you are kind of in the dark here. Try to find SUN's contributions to Linux 2.6 kernel here:

http://www.linuxfoundation.org/publications/linuxkerneldevelopment.php

They are not even in the top 31. Lets see how many sun developers that are active in maintaining parts of the Linux kernel...

[frimann@roedmette ~/linux-2.6.31.4]$ grep -i sun.com MAINTAINERS | sort -u | wc -l

1

One guy...

Or lets try to scan the whole source code for Sun contributions...

-> 11 hits on 6 different persons.

So lets just put the SUN was/is a major contributor to Linux in the grave.

And for fun I did the same with IBM, and there are thousands of contributions by hundreds of people.

And you state:

"The Linux community looking at Sun in a negative light for "buying the rights" to open-source Solaris seems rather myopic"

In law inforcement you normally operate with a concept called 'Cui bono', 'Cui bono' means who benefits in latin. Lets look at the timeline.

January 2003 IBM Software Boss starts talking about merging Linux and AIX.

In early 2003 SUN buys the Unix licens from SCO that gives SCO the financials to sue IBM.

In march 2003 SCO sues IBM and generally threatens everyone involved with Linux.

June 2003 the SUN marketing machine starts to crunch

http://www.sun.com/datacenter/migration/aix/AIX_WSJ.pdf

January 2005 SUN launches Opensolaris

Now what have SUN benefitted from this whole deal ?

// Jesper

RE:David Halko

"That's OK - Sun pioneered in heavily multi-core'ed and muti-threaded CPU architecture... and the rest of the market continues to emulate with Intel being the fastest to catch up."

Well it wasn't really SUN, SUN bought the company that made the TX000 servers, it was AFAIR one of the 'original' SUN guys who made that company. Just like the highend servers which they got when they bought Cray's Unix business back in .. 1996 I think it was, wasn't their own creation. And yes the TX00 were innovative and the Niagara family is *GREAT* for what it was designed for. But there isn't not a general purpose box.

As for pioneered, multicores and multi threaded cores, Weeeel.. then the Power RS64 IV had up to 4 threads back in 2000, and you forget that the cell microprocessor came out at the same time as Niagara. And power4 was the first CMT processor back in 2001, where as the UltraSPARC IV (Jaguar) is from 2004.

"That's OK - Sun pioneered in heavily multi-core'ed and muti-threaded CPU architecture... and the rest of the market continues to emulate with Intel being the fastest to catch up."

I don't think that either HP nor IBM could have convinced their customers to give up their single threaded performance. And as for catch up, U forget that SUN is being absolutely trashed in the UNIX markedplace, sad but true. So there are no business reasons to emulate SUN's CMT. If IBM were really worried about the T5XXX line they could have put out a 'Q' version of their POWER6 servers as they did with POWER5+. Actually the 1.8GHz p560 has a higher socket throughput with 0,56 rperf than the power6 power 570. And such an MCM based Q server would have rocked, in benchmarks. But in the real world the server wouldn't have had enough memory to be useful, as when you run in a virtualized environment you use a hell of a lot of memory.

And early here next year we will have POWER7. 8 cores 32 threads on one chip, but still with the same or better single threaded throughput as POWER6.

So right now a T5XX0 might be 1.5 times faster per socket than a power 5x0 box. But with four times the cores and a bit more umph on each core, then it will be something like a factor of 2-3 the other way around. Sure rainbow falls with 16 cores will make a difference but IMHO it's just more of the same 128 statically scheduled threads versus 32 SMT threads where one thread actually can use the whole processor.

And as for the whole storage setup on the T5440 RAC benchmark, I tip my hat at that one.

But for the fact that you can quickly burn your way through a flash card if you write alot. Now I know that some SSD vendors simply just reserve half of the drive so as to extend the lifespan of the drive. And that would work for me.

Now the whole 'lots of disks on the benchmark' is a hack, In my world you have Big Tin Disk boxes from Hitachi,HP,STOR^H^H^HUN,EMC,IBM and Netapp that can do a lot of IO/sek by using cache. But as the TPCC benchmark is also about price, they are to expensive to use on such a benchmark. Hence you use a lot of disks. I find the disk price of the benchmark irrelevant, as in my world the disk systems are a constant, no mather what server or vendor we use.

"Sun's innovation in benefited and will benefit the entire global computing community for years to come while IBM's innovation benefited only themselves and a few customers for a short period of time.

There is good reason to argue that this behavior is a primary reason why Sun was not consistently profitable and IBM typically is.

Sun = Innovation ; IBM = Business"

I think you are very wrong. I grew up on BSD and Solaris Unix wise. I learned like a good UNIX boy at university to hate IBM, cause they were the big bad mainframe monopoly. And they were in the late 80'ies. But in the early 90'ies in my first IT job I had to use AIX and power, I lobbied for a HP or SUN box, but actually I got to like AIX with it's logical volume manager, SMIT menu interface, and easy to use cli where everything was like chxx mkxx lsxx rmxx. And you were the underdog as a AIX admin, that had to listen to how good Solaris, HPUX and Whatever name digitals Unix used at the moment. But us in the AIX admin department could "outadmin" the other admins, cause while they were editing files, and sending signals to deamons we just did a chxx and then a refresh -s subsystem. Or when they were formatting drives, and tying in sectors and sh*t, you simply did an extendvg and then that was done.

This is where my preference for AIX on power was founded.

But you seem to forget that the whole basis of SUN+Oracle's business depends on IBM inventions. IBM and others (not SUN) invented the RISC processor. IBM invented the relational database and SQL. And if they were such a monopoly, they why are SUN and Oracle in business at all ? Not to mention EMC, cause IBM invented the hard drive.

And if it hadn't been for the IBM PC, then neither Microsoft nor INTEL would be the giants they are today. What ground breaking technology have SUN invented besides Java ?

If you search for "SUN inventions" on google, the whole first page is about a novel.

How many nobel price winners have SUN fostered ? IBM is 5 or something, and not in economics, but in physics with discoveries like the tunneling microscope, and high-temperature superconductivity.

And you say SUN is innovating for the global computing community. Ok, that is why SUN was the very first to finance SCO's attack on Linux. Simply great, buy the rights to make Solaris code public, while trying to put Linux in the grave. No it wasn't Microsoft who were the first to go after linux using SCO as a proxy. It was SUN.

http://news.cnet.com/2100-1016_3-1024633.html

And we are many pasionate linux users who hasn't forgotten that move.

So yes IBM is all about business, but damn so is SUN. And when it comes to innovation, SUN is, or rather was, no match for IBM.

// Jesper

@Kebabbert

Dear kebabbert

Listen I stopped flaming people on the internet 15 years ago cause life is way to short to use it on bringing unhappiness to people. And I'm not going to start now. Although I must admit you last post here did tempt me a tiny bit :)=

Making fun of my name is cute btw.

I would never spread FUD about you, I simply said that you have a nasty habit of calling people FUDders and liars,when they don't agree with you. Now I have never called you a liar, I have simply stated that you were wrong. There is a difference, lying is stating something that you know is wrong, and I don't know if you are lying.

You seem to ask the "If POWER6 uses 500-600 watt ?" question again and again.

You seem to have a habit of asking that same question anywhere you can sneak it in. Here are 3 links beside this debate:

http://www.theregister.co.uk/2009/07/21/ibm_power7_details/comments/#c_543822

http://www.theregister.co.uk/2009/07/22/sun_sparc_t_crank/comments/#c_543769

http://techworld.idg.se/2.2524/1.242411/osaker-framtid-for-suns-16-karniga-processor?articleRenderMode=listpostings

And such questions might get you cheers, at a SUN xmas lunch, but not here.

But at least now I have directed you to where you can find an answer to your question so now you don't have to keep asking it. Right ?

And for your whole little 17 Million for a machine, let me correct you yet again, it is the price for the whole power 595 benchm arked configuration. With software, disks, servers and all. So it's not the power 595 server. So now you know that.

And I cannot understand your claim that nobody can afford these machines. Where I work we have a lot of M9000,M8000,Superdomes,p595 and power 595. Well it's not ours all of them, many of them are owned by our customers.

And I still don't think that the T5440 is so stunning. 7717511 tpcc for 12 machines gives 643126 tpmcc per machine, now the old 4.2 GHz version of the power 550 does 629159 tpcc. So I gues it's a pretty good bet that the 5.0 Ghz version would do 700000+ tpcc.

// Jesper Fnesper Lesper Desper Klesper whatever.

RE:Kebabbert

*SIGH*

You accuse people who don't agree with you of being liars and spreading FUD.

But remember when you point at someone there are 3 fingers pointing back at yourself.

Lets take one of your claims:

POWER6 uses 500-600 watt per CPU, and with CPU I presume you mean CHIP, hence a

Now doing the math this would mean that just the CPU power usage would make a power 550

use 2400 watt. So why don't we check out your 'facts' with one of the tools that are

actually used where I work to calculate power usage.

http://www-912.ibm.com/see/EnergyEstimator

A fully loaded power 550 with

8x 5GHz cores

256 GB of RAM,

DVD-RAM,

Tape,

Dual 10Gbit,

6x15K SAS disks

5 PCI Cards

uses around 1429 Watts.

And if we strip away all the RAM and other stuff then it uses 878 Watt.

If we use the 3.5GHz Power6 then we hit 738 watt, hence going from 3.5 GHz

to 5GHz uses 35 Watt extra per Chip, as the machine has 4 chips.

So can you answer me why the p550 isn't using 3000-3500 watt as it should,

if your numbers were right ?

And another one.

"And for you FUDers who say that Niagara doesnt scale beyond 4 sockets and therefore produces inferior solutions: well you are WRONG as evidenced by this article."

You don't get it, do you ?

This benchmark is done using Oracle RAC, it is a cluster. You could have done the same with a processor that didn't support SMP. Hence your argument

that cause you cluster a bunch of 4 socket T5440's using RAC then it proves that the T2+ processor scales beyond 4 sockets is.. well.. not valid,

I would even go so far as to call it ignorant.

And last "Who can afford $17 million for the P595 machine?"

You didn't bother clicking on the links I posted for you. This benchmark is all about selling software licenses.

Server price:

Now the price of the power 595 is 5.833.031 USD with 3 years of support.

Now the price of the 12xT5440 are 2.113.991 USD with 3 years of support.

So the T5440 are a good deal cheaper as expected. But we need software to run on the benchmark.

Now the price of the Ora software is 7.954.800 USD with 3 years of support

Now the price of the IBM software is 1.277.477 USD with 3 years of support.

So for the physical servers and the software the cost is.

Ora price Servers + Server Software is 10.068.791 USD

IBM price Server + Server Software is 7.110.508 USD

The reason why oracle's price end up being lower for the whole solution is that they use flash drives, something that wasn't really available when the power 595 benchmark was done.

What you can see is that the Oracle software costs more than the Power 595 + IBM software. And this is the whole reason why Oracle

is pushing Niagara, based clusters. To sell more software.

And if you look closer at the Oracle prices then you will notice that you only get to lease the software for 3 years. After that you

have to buy yet another lease. And the software support is Metalink (web only).

If you were to buy the software the price would be around 20MUSD - Discounts.

Furthermore the IBM software stack also has 3 Tivoli magement products priced into the solution. ( Management, TADDM and accounting)

Again you can't back up your claims, this is actually getting a little boring.

// Jesper

RE:Kebabbert

You don't get it do you.

This system is actually quite expensive, when you realize what you are getting.

You pay 8 MUSD in software licenses to run a complex database on some entry level servers for 3 years. When those 3 years are gone you will pay another 8 MUSD, to run this system for another 3 years. Who is stupid enough to do that.

The real price of the software licenses if you were to actually buy them is almost 20MUSD in list price with 3 years of real support, (you only get Metalink access in the benchmark as your support)

So to answer your question "Who can afford $17 million for the P595 machine?"

One that is clever enough to buy a power 595 with a real software license and real support.

Again as it often is with you you forget to check your facts.

If you look at the reports for the the power 595 and the T5440 benchmark reports you will see that the price of the actual power 595 server + the server software is less than that of the 12xT5440+server software, even though you are buying time limited licenses on the T5440. If you bought normal licenses and software support the T5440 hardware + software licenses would be double that of a power 595 + software license.

So all this benchmark has proven is that it is in real life it would much more expensive to run your workload on RAC running on a bunch of T5440 compared to running it on a big tin SMP box.

Do the math

http://www.tpc.org/results/individual_results/Sun/Sun_T5440_TPC-C_Cluster_ES_101109.pdf

and here

http://www.tpc.org/results/individual_results/IBM/IBM_595_20080610_ES.pdf

// Jesper

Just an expensive lease.

As the First AC poster pointed out the licenses are 3 year limited licenses. So you don't buy the software you license it, for 3 years. If you were to buy the software it would cost you 192 licenses times 82.000 USD=15.744.000 USD+10391040 USD=26.135.040 USD List price. - Discount.

And notice there is not any normal Oracle Support offering on this

http://www.oracle.com/corporate/pricing/technology-price-list.pdf

On page 8 it clearly states that on 'term licenses' you still pay 22% for your normal software license update and support.

Oracle Incident Server Support Package is basically metalink access.

http://www.oracle.com/support/collateral/oracle-technical-support-policies.pdf

So basically on the software side on this you get to lease the software for 3 years with Metalink as your support.

I mean that is just laughable.

And putting up 384 cores and only getting 7.7M tpmc is not impressive.

And the per thread throughput is simply awful, I mean 384 cores x 8 =3072 threads, that gives you a staggering 2512 tpmc per thread. That less throughput per thread than results using 200Mhz PentiumPro's from 10+ years ago.

I am not impressed.

// Jesper

Sun tunes its VirtualBox

VB

Well I run VB on my laptop, my host OS is a RHEL and I have a XP running inside VB. And it just works like a charm. Most of the apps that I run inside the XP are quite IO intensive with sync IO and they benefit from the fact that my Linux caches the VB disk file, hence they run faster than on the bare metal XP, that I have in another disk partition.

And as Antov Ivanov wrote:

" VirtualBox is an excellent Type 2 supervisor."

And I couldn't agree more, but SUN needs to start to make some money on this product.

// Jesper

Sun, Fujitsu crank Sparc64-VII clocks

RE:hmm....

It's very simple. Lack of product development, that actually get products out on the marked.

// Jesper

Oracle to 'out-Sun' Sun on hardware and software

Re:@Matt Bryant -- HP Ripe For The Picking!

Ehh... ?

So you honestly think that putting a 'We know best' filesystem between the database and the disk is going to get you a very good performance ?

I think you have been reading the wrong blogs at SUN, like:

http://blogs.sun.com/roch/entry/zfs_and_directio

Man the guy must have hit his head, or haven't worked much with databases. Luckily there are more sensible guys like this one:

http://blogs.sun.com/realneel/

Who actually know what he is talking about and makes sure that ZFS does not stand in the way of the database. Which is what you should do.

ZFS is way to hyped IMHO.

// Jesper

IBM wins benchmark crown with SSD power

RE:SegFault

Well I don't think you understood what this benchmark is all about, and that is not meant in a negative way :). I must admit I didn't get it first either, I was like "what the f***", but after having a look at the FDR I got it.

This is not a storage benchmark. Basically the SDD's are kind of irrelevant.

This is a benchmark that shows what the Virtual IO Servers, that you use to virtualize disks inside a power server, can pull through of IO's per sek.

So if you for example compare this benchmark to the IBM DS5020 benchmark (I use IBM<->IBM so as not to bring any platform religion into this), then the difference is that in this benchmark what corresponds to the DS5020 control unit, the whole cabling, switches and HBA's on the servers that run the workload, are all implemented in virtualization software inside the power 595.

And seen in that light it's a pretty good result, that server virtualized storage can beat dedicated hardware and do it with a 5 ms response time.

The actual stack is psysical SAS->Virtual server SCSI adapter ->Virtual client SCSI adapter.

It could most likely be done faster with 'virtual SAN' NPIV

I would very much have liked to see the CPU utilization and RAM usage of the VIO server under this benchmark run, as that is what is really interesting.

And as for the price.. well hey the IBM dorks priced the whole machine load generator and all into the price. So basically the price that is listed here is the storage + a highend server that would run your whole DB workload and all. Not just the storage part as it is in all the other submissions.

// Jesper

TPC slaps Oracle on benchmark claims

RE:Kebabbert

I have never said that Nobody would buy SPARC because of too high a price, sure customers will, buy SUN servers.

But things aren't really looking that good.

http://www.sun.com/aboutsun/investor/earnings_releases/Q409_SLD.pdf

SUN sold 6530 MUSD worth of servers in 2008

SUN sold 4657 MUSD worth of servers in 2009

that is a drop of 29%

And before you start to dissect the numbers and say "Yes but CMT is doing well", notice the other system product category :)= Moving revenue around in your financial statements is nothing new. And from what we hear it isn't like things are getting bright for SUN (no pun intended).

Do I think that current SPARC Servers are to expensive ?

Yes, in the M series you simply get less bang for the buck compared to for example power.

The CMT series are not well suited for many workloads IMHO, and I think they are over hyped.

Sure the band new T5440 with 32 T2+ Cores will outperform a Power 570 with 16 cores, on for example specJBB2005, nothing much to say there there are benchmarks to show it.

http://www.sun.com/servers/coolthreads/overview/performance.jsp

But there are some things you should keep in mind here.

1) The Power 570 that SUN are comparing with is more than 2 years old, note that there was a benchmark out on the 5.0GHz version and the 32 core version of the power 570 which both beat the T5440 score. A little cherry picking is always nice

2) This is on a workload that is very well suited for the T5440, lots of threads and embarrassingly parallel. The T5440 uses 32 JVM instances versus 8 for the power 570. The T5xxx'es are good for benchmarks, but in the real world where single threaded performance is a key factor, things do look different.

3) I don't really know if it is fair to compare a high RAS system like the power 570 with an entry level box like the T5440. I mean you can't even hotswap PCI slots on a T5440. It is more suited to compare against the DL585 which is also mentioned on the page.

4) And funny enough the machine that is in the same league, with regards to RAS, as the T5440, the DL585 they don't really mention. But Ok it's also only 6 times cheaper, than the T5440, and faster.

It's not that I think the fact that SPARC is doing badly is a good thing, we need competition. But Snoracle needs shut up with their marketing bull, and start to ship some better products.

I mean, in Q1 2010 POWER boxes will start to ship with POWER7. That means that the CMT advantage with 8 cores per chip is gone. And from what I understand POWER7 will be packing a heavy punch, with 8 cores each with better per core throughput than power6.

This basically means that a 4 socket T5440 will be competing against a 2 Socket POWER7 box, and on some benchmark, like SAP, get the living crap beat out of it by a 2 socket power7 box.

And how do you think a M9000, that today barely can beat a 64 core power 595, will fare against a a 256 Core power 595 ?

Specially seen in the light that the next 3 years the only upgrades to the M9000 is minor speed bumps.

Lets hope Snoracle gets their act together, but I seriously doubt that they will.

// Jesper

RE:Kebabbert

As this was a real story, the machine in question was actually a p690.

"Oracle would be dumb to price the SPARCs to high."

Yeah right. T2+, is 0,5 license per core, but T2 is still 0,75. So sure keep milking the install base.

And the oracle license count for a T5220 is still 6 licenses, which with sw maintenance on Oracle means that the bloody DB is almost ten times as expensive as the box.

And for at T5440 it'll still be 16 licenses, for a box where you need to utilize 256 threads to get the maximum performance out of. Not that is not a straight forward job.

// Jesper

RE:All your CPUs belongs to Oracle + SUN

"All your CPUs belongs to Oracle + SUN"

<true story>

Yes, heard that one before. Lets see.. you have 16 Oracle partitions running on a physical machine with 32 cores... He He he.. that is lets see..

0,75 licenses per core times 32 cores times 16 partitions, now that is 384 Licenses costing 47KUSD. And you have only paid for 32 cores.. ohh.. ohh.. Ohh.. yes that means you have to go and buy another 360 licenses, and You have been running this for 3 years.. that is 22% a year also.. so basically *tap* *tap* *tap*, and with your great discount. 16.852.320 $

Oracles own license rules have never stopped the Oracle sales machine. It took intervention from the Server vendor and the customer top management and an escalation to oracle on a higher level than the country level. To fix this.

</true story>

So face it. Oracle bought the SUN hardware as a license cash cow, and the Java part cause it was better that they owned it than SAP or IBM get their hands on it.

Not that Oracle products aren't great, but the prices..

// Jesper

Again Big brother cause something is rotten in the state of Denmark

re:Matt Bryant

Well I am not 50, and not really an old geezer yet. But I must say that there is a whole of a lot difference from my current 4GB 2GHz dual core Laptop, to the 2.5Mhz z80 based Piccolo with 32KB ROM and 32K RAM, from 'Regnecentralen' that I learned to program on, at the university. And you had to bring a bag cause the bloody 8 inch floppies couldn't fit anywhere.

// Jesper

Re:perf

"I personally don't believe that the DB2 on the i platform is that much worse than the DB2 for Unix boxes."

Well i is not about performance, it has never been. And the DB2 of i is also different than the 'distributed platform' version. Furthermore there is an extra sw layer(s) between the hardware and iOS, and that doesn't speed things up. One thing I've learned is that you must never insult a sysadmins, iSeries box. You can scoff at his wintel servers, say that his Unix servers aren't really Unix servers and say that his mainframe is to expensive.. but insult his iSeries and he'll smack U up.

And he DB2 of yesterday, performed like sh*t on AIX, but with version 8 and newer versions things changed, and now Oracle/Informix/Sybase weren't the only choices for AIX in my book.

// Jesper

RE:Oracle gets free publicity

Well, to the untrained eye. But if you read this article the number of machines needed to get double figures in the TPC-C benchmark, have more or less doubled. Now it's 16 (2 Racks) from One rack in the original Oracle commercial. So yea Oracle might have slashed the license multiplier, but it is still hugely expensive.

16 machines x 4 sockets x 8 cores x 0,5 license/core=256 licenses.

at what.. 82KUSD a piece ? that is 21 MUSD just in licenses..... which gives you 0 more value.

// Jesper

IBM readies Exadata killer

Re:Victor 2

You wrote:

"The single traffic cop is one massive point of failure. On Oracle's Exadata, any component on any node, or even a whole node (be it an Exadata or a Database server) can fail and the system will continue running as if nothing happened"

You described RAC as being able to continue runnning witout any disruption with a node failure. That is simply not true. And neither would any DB2 on POWER product. Hence my comment.

"Big deal, so has Sun... It doesn't mean THIS particular case of creating a DB cluster had to be done using Infiniband... why not 10Gb?... Let's see what IBM does to address the flash accelerator part, but why use Infiniband if you are going to use standard and SLOW hard drives?"

I didn't say that SUN hadn't used Infiniband for years, only that your comment about IBM copying SUN by using infiniband is a not true.

And you don't get it. The external IO drawer is connected to the server via Infiniband. And one main reason to use Infiniband on POWER is that you have the option of plugging it into the GX bus, which means lower latency and higher throughput.

"Oh, oh... so, you say using flash is flawed, yet IBM will come up with something flashy too?.. contradiction alert..."

Now, seriously, who told you you had to replace flash memory every month?... remember also, that Sun/Oracle are NOT using SSD drives, they are arrays of memory modules. Ellison emphatically stated so.

If you do ALOT of write IO to the same block, you will very quickly burn through a flash memory chip. This is the same for all.

"I'll grant you that, yet, how does it compare to Exadata v2 grand total of 15TB of data when compressed?"

Better AFAIK, compressions isn't really something I normally use.

"SPARC roadmap has hardly changed over time, ROCK was set to be a parallel architecture to that of CMT and Fujitsu's SPARC.. new scheduled projects can get canceled, that's why they are called "projects"."

Call it what you want."

I call it p***** on it's customers. If you hype a product that much, yes it was hyped way way to much, and then just cancel it.. you must expect that customers (like where I work) gets p*ssed.

Just look at these few links:

http://news.cnet.com/Sun-burnishes-next-gen-Sparc-chips/2100-1006_3-5561693.html?

http://blogs.sun.com/jonathan/entry/rock_arrived

http://www.theregister.co.uk/2009/01/28/rock_sparcs_on_track/

http://www.physorg.com/news97329357.html

http://sun.systemnews.com/articles/107/3/news/17539

http://www.infoworld.com/t/hardware/suns-rock-rolls-further-along-898

http://sun.systemnews.com/articles/111/1/news/18046

http://sun.systemnews.com/articles/114/4/news/18516

http://research.sun.com/spotlight/2007/2007-08-13_transactional_memory.html

....

It's one of the most hyped products in recent years, and it turned out to be just hot air.

"SPARC is alive and well, thanks for asking."

Yes it is alive but barely. Don't get me wrong I WANT SPARC to be there, we need competition and good SPARC products. But on the APL line nothing is happening. And to be honest fujitsu isn't really moving toward being more of a server vendor, than they are today. They would rather buy the PC part of the joint Fujitsu/Siemens vendor, than buy SUN. That basically leaves us with the TX which is terrible for workloads that cannot be parralized, in the extreme.

One of my wife's friends who is a very competent Oracle DBA, she quit her job after they went from APL to T5XXX'es where she worked. Simply cause the DBA's had to take the blame for the bad performance, massive increases in install times.

And that is what Larry don't understand when he praises Niagara, he is going to p*** off much of his loyal base which is the Oracle DBA community, when they get forced to rethink much of what they today consider best practice.

"Then, don't use for other things than what it was intended too... You can't buy a toaster and expect it to serve you coffee.

Oh, and Power isn't all that cheap either."

Jup, and I am not using it for anything that requires just decent single threaded throughput.

Power boxes are expensive, and you should always batter your IBM sales person into giving you some decent discounts. But the savings in software licenses, and the pure utilization that you can get these boxes to run with. Means that it is actually a good deal IMHO.

"No, look closely... why is that "virtualization" so similar to solaris 10 "virtualization"?"

Ehh... Solaris 10 virtualization is containers, and yes AIX 6.1 had WLM extended a little bit to make Workload Partitions, that had more functionality than containers, but I would hardly call that virtualization, rather workload management and user and process isolation.

Now when an AIX admin talks about virtualization (s)he talks about what the power hypervisor can do for you. And here SUN is no match.

"So, you want to use Oracle and not pay for it?... execution compatibility is not al there is to Oracle products."

No, but I don't want to pay more than I have to, I don't want to pay for 24 licenses if I can design a solution that can run on 8 licenses, with the same throughput.

It gives me no value NO VALUE what so ever, to pay for more licenses. And if it comes down to the buy cheap hardware with expensive software, I'd take the better hardwarewhich saves on the license cost any day. Good hardware gives you value.

Note that this also goes for any software, not just oracle.

"Leave that part to the people that knows."

Yes, I am. Myself and I'm good at it.

// Jesper

Re:Victor 2

"The single traffic cop is one massive point of failure. On Oracle's Exadata, any component on any node, or even a whole node (be it an Exadata or a Database server) can fail and the system will continue running as if nothing happened."

Yeah right.

"IBM's cop fails and your system is dead. Guess IBM rushed too much to get this on the market to gain attention, didn't they?"

You don't know much about RAC do you?

"They copied up to using HPC's prefered Infiniband for networking... "

Ehh.. POWER servers have been using Infiniband for IO connectivity for years.

"What will IBM offer to compete against Exadata's flash memory?"

Perhaps something that works. Let me turn the question around, what will Oracle do so that customers don't have to replace their flash memory drives every month ?

"Or Oracle DB's data compression?..."

DB2 have been doing data compression for years.

"Power's processor have since a long time missed their release schedule, "

They might but six month, 1 year is something completely different than the SUN policy of just hyping up a processor and then... cancel it. BIG difference.

"and only recently did they announce the capacity to have more than two cores with POWER7... which even with DARPA's funding, took all these years and still isn't out in the market... A feature SPARC processors have had for quite some time now..."

Again, it is most likely going to be a product that is going to ship, where as Rock is... dead.

And honestly Niagara is terrible if you don't use it for what it was intended for. And expensive also.

"AIX latest releases and features are all about copying Solaris 10 features."

Not really it is all about virtualization.

"DB2 is all about it's new "Oracle DB Compatibility"."

Yes, perhaps cause many customers are tired of Oracle's pricing policies, not their products.

"So, TPM, tell me... who should we follow, Oracle/Sun or IBM who in turn is just trying to catch up with Sun/Oracle?"

The one with the best and cheapest products, and a product that actually have a future. *cough*

Oracle on some points yes. Sun not likely.

// Jesper

Re:Matt Bryant

"Personally, I think IBM should have hidden the DB2 bits under the hood and just given it a brandname like "Xadataplus", just to wind Larry up.

Heh, yes that would be fun.

// Jesper

Sun's Sparc server roadmap revealed

RE:@WinHatter: wait for it

"Does that mean you believe Sun is planning on investing in the manufacturing processes?"

SUN isn't actually investing in the manufacturing processes, they don't manufacture processors.

T2 line is manufactured by TI and the SPARC64 is made by Fujitsu and Manufactured by TSMC.

Furthermore Rainbow falls and onwards is also going to be manufactured by TSMC, as TI is exiting that marked.

So basically Snoracle is relying on TSMC, who are going to make their first server processors, with these new SUN and Fujitsu deals.

So I wouldn't be surprised if some delays crept in, with SUN going from their long time partner to a new one.

// Jesper

RE AC on RE:LOU

AC wrote:

"Wow! Who's been drinking the IBM Koolaid? That's the most ignorant piece of drivel I've ever heard - OK, that's exaggerating, but it's close! IBM went from Power6 4.2GHz to Power6 5GHz (20% increase in GHz), but when looking at IBM provided rPerf numbers this increase only yields 11-15% increase in Performance."

Well let me guess, math isn't your strong side. Or at least you don't really bother checking out the facts before you make an unfounded statement.

rPerf: for the Power 550

2x4.2Ghz does 18,36 rPerf equals 2,19 rPerf/Ghz/Core

2x5.0Gh does 21.18 rPerf equals 2,12 rPerf/Ghz/Core

That is 97% scaling.

4x4.2Ghz does 36,28 rPerf equals 2,16 rPerf/Ghz/Core

4x5.0Gh does 41,81 rPerf equals 2,09 rPerf/Ghz/Core

That is 97% scaling.

6x4.2Ghz does 52,24 rPerf equals 2,07 rPerf/Ghz/Core

6x5.0Gh does 60,20 rPerf equals 2,01 rPerf/Ghz/Core

That is 97% scaling.

8x4.2Ghz does 68,2 rPerf equals 2,03 rPerf/Ghz/Core

8x5.0Gh does 78,6 rPerf equals 1,97 rPerf/Ghz/Core

That is 97% scaling.

That means that the scaling in performance is 97% of the scaling in frequency. That's better than putting more cores on the same die, in scalability.

Now if you look at specint_rate numbers:

16 way 4.2GHz power 550 does 212 in specintrate which means 3,15 specintrate/Ghz/Core

16 way 5.0 GHz power 550 does 263 in specintrate which means 3,29 specintrate/Ghz/Core.

Again my statement stands.

AC wrote:

"The numbers start looking even worse when you compare the doubling of GHz between Power5+ and Power6 where IBM could only get a 35% performance increase..."

And I also specific wrote that you could not use my statement to compare between different versions of the Processor.

AC wrote:

"Don't believe me? Check out IBM's performance numbers:"

I did and you were wrong.

// Jesper

Big brother cause something is wrong in the state of Denmark.

Re:Lou 2

Well if comparing between different types of the same processor, for example from a 3.5 GHz POWER6 to a 5 GHz POWER6 processor, for example on a power 550, then it's actually just as fast or rather a bit faster than the increase in GHz suggest.

Why ? Cause the interconnects and the busses scale with the CPU frequency.

// Jesper

Ellison: Sunacle is an IBM killer

RE:A very clear statement...

The problems for Larry are many. Not least that it is a huge difference from being a Software company and being a HP/IBM.

But the biggest problem IMHO is that there are quite a few Oracle customers, including where I work who are pretty pissed by they way oracle have been jacking up the prices. these last years.

So basically if you think that the something like MYSql will have a chance in this licens factory, you might want to reconsider.

// Jesper

Bigbrother cause free speech is a thing of the past in Denmark

Oracle, Sun set to light FlashFire

RE:RE:Moving up the product announcement 1 month

"Enough information has been published about the October benchmark publication and SPARC road map sine the acquisition announcement - this indicates that the SPARC line is secure for the future."

Eh ?

Did we see the same roadmap ?

In the next 3 years the APL line will ...

1) Be on the same processor manufacturing technology.

2) Have the same number of cores per chip.

3) Have the same number of threads per core

4) Increase a staggering 500 Mhz in speed.

5) Be manufactured by TSMC, which haven't really manufactured Highend Server Processors before. No Intel Atom, and Graphics chips from nVidia don't count.

As I said before the IBM and HP sales people who sell highend systems will have a field day.

// Jesper

Big Brother cause something is rotten in the state of Denmark.

RE:Shaun.

Working for a company that is both a big Oracle and SUN customer, I must say I don't share your excitement, at all.

Well the problem here is that, if you look at the roadmap that was 'leaked' here on theregister then it's really a disaster in the non-niagara space.

The APL line is simply not going to evolve in the next 3 years. 500Mhz is really not that impressive.

IMHO, this is a clear signal from Snoracle, push the platform that will enable you to sell a lot of software licenses, and put the consolidated large servers, which are expensive to develop, out to slowly dwindle away.

You can be certain that HP and IBM will be throwing good offers at all the SUN costumers that have 'largish' M[589]000 series or equivalent older machines.

A shame IMHO.

// Jesper

Big Brother cause something is rotten in the state of Denmark

Re:David Halko

Yes the big looser here, with regards to the 'Database Machine', seem to be HP.

// Jesper

Bigbrother cause something is rotten in the state of Denmark.

Oracle and Sun taunt IBM with Sparcs

RE:Flash in the pan?

Matt is right.

Any vendor can make a clustered benchmark and get a 'big' result. The reason why the Vendors stopped doing these mega clustered benchmarks is that, we the customers, could see it for the bull it was.

And if Soracle opens this 'battle front' again they will just get their butts kicked.

I mean for example a distributed DB2 database is something that has been around for a long time, it is even a shared nothing database so it will scale really really well.

So for example lets say that IBM did the following:

1 x 42 inch rack.

4 x 9U Bladecenter each holding

14 x JS23 Blades each holding

4x 4.2GHz Power6+ cores.

Each blade will do about 350-400K TPMC, based upon it's rPerf rating. And lets just be seriously pesimistic and say that the scaling factor i 95% on a shared nothing database.

That would put one rack in the range of 14-19 Million TPMC.

Or Imagine a rack full of HP half hight blades with Xeons with for example DB2 ?

This is a can of wuuuppa** that Soracle shouldn't open.

// Jesper

IDC outs the worst quarter in server history

RE:Itanic@AC

Eh. Your numbers are wrong.

SUN had a 40% drop in Unix sales.

HP had a 35% drop in Unix sales

IBM had a 16% drop in Unix sales.

// jesper

IBM touts Power Systems prowess on SAP tests

Re:Thread vs Process

Fenton, I think you might have mixed up the threads and processes of the Operating system with the concept of processor core threads.

On for example a niagara based processor each 'processor core thread' is actually capable of running a whole separate OS image, which then again can run many processes and threads.

And I have seen the same problems when migrating to Niagara based servers as you describe.

And don't get me wrong I think Niagara is great, for what it was designed for, workloads that will perform well with many light independent threads. But there is a long way from that to a general purpose Business platform.

// Jesper

Big Brother cause:

1984 was meant as a warning, not a manual.

Kebabbert

It is very seldom I say this about someone posting here.

But you have absolutely no clue about the things you are talking about.

And your troll techniques are kind of infantile.

// Jesper

RE Kebabbert

>Here we see some old benches where T2 outperformed Power6, on spec_int, ORACLE, SAP, Lotus Notes, etc.

>

>http://johnjmclaughlin.blogspot.com/2007/10/utrasparc-t2-server-benchmark-results.html

Are you serious ? These are BMSeer quotes, do you seen any other people participating here using links HP or IBM marketing fud ?

These are chip versus chip comparisons. Yes they are perhaps interessting to you. But if you are talking Licinsing, setting

up real world applications or even comparing servers versus servers they are utterly meaningless.

Have you ever tried to setup a 4 threaded system versus a 64 threaded system to do the same workload ? It's a completely different

ballgame.

You just put yourself in the same category of people that you are pointing fingers at by using SUN's own FUD central.

>SUN has publicly stated that TPC-C is meaningless. It is favours pathological servers that no one use in real life. A while ago, IBM held the record.

>The server used many many Intel CPUs (not Power cpus) and 2TB RAM! And it used many short stroked hard drives! Who uses short stroked hard drives?

>How much does 2TB RAM cost? This config is really silly. And this IBM machine costed like 120 million USD, built for one purpose: TPC-C but nothing else.

>People shun away from TPC-C, they say "if you are good at TPC-C, then it proves that you are good at TPC-C and nothing else.

>Literally nothing else, certainly not average DB usage".

Who cares what SUN says. TPM-C have done more for Database performance, cause the hardware vendors actually learn something from this benchmarks, that goes

into their products, than SUN has done since they choose to stop making those benchmarks.

> If you really think that a average DB admin has access to this machine, then you are wrong. TPC-C is really artificial.

Sure TPC-C is artificial, it's a benchmark. But jup I have access to several, You name it POWER5 p595's with 2TB of RAM, SUPERDOMEs with 64xPA8900

or 128 way x I2 Montecitos, power 570 with 16 4.7GHz cores. And yes they are getting used some with quite large partitions. 56+ cores.

> SUN could also build a server with many intel CPUs and compete in this artificial test, but why would they? It says nothing about real life applications.

So why don't they ?

> Your point does not prove that Power6 is better than Niagara, which is not.

> This reminds of Top500 supercomputers, which are basically a bunch of nodes in a large network.

>On rank nr 5 we find the IBM Blue Gene super computer. It has.... dual core 700MHz PowerPC CPUs! Does a high ranking at Top500 mean that

> IBM makes better CPUs than SUN? Not quite.

You don't really get it do you ? You hail systems with lots of slow cores and lots of threads and then you scoff at Blue Gene, which kind of like uses

the same principle. I mean... you make less and less sence.

>"Even Oracle confirms that, since the processor factor licensing table shows Niagara is the slowest, then comes Intel and IBM are the fastest."

> Come again? Do you mean that Niagara is slower than IBM?

Sure it is. Don't confuse speed with throughput. A niagara Chip will deliver more throughput than a POWER6 chip. But on a core to core basis there

really isn't a match. On a throughput basis it's different there the 8 core 64 threaded niagara chip will outperform a POWER6 chip, if...

1) Your workload can be threaded to enough threads to exploid the 64 threads of the niagara. Not much good if you only have 4 threads.

2) All your bloody threads doesn't block eachother.

3) Your job isn't depending on serialization like the job I'm doing tuning on at the moment. 10 min with 1 thread, then 20 min with Many threads,

then 10 min with one thread and result. The job is heavily CPU bound.

Now who would run this job fastest ? This currently runs on a 3 Core POWER6 4.7Ghz Virtual machine.

Now on a niagara I would most likely get 80 min with 1 thread 14 min with many threads 80 min with 1 thread, oh yes sure I could start up a whole

development project to paralize the code, and pay more for my Oracle License. Welcome to the REAL world.

>It depends on the work load. If I have to transport 1000 persons from one place to another, then what would YOU choose?

>Niagara is like a huge slow bus, whereas legacy constructed CPU as Intel/Power6 are like a porsche.

>Imagine you have to transport 1000 persons. Which would be first to finish? Niagara or Power6?

Wrong again, Niagara is the 64 small slow minibusses, where you need 64 drivers, power6 are 4 fast Large busses using only 4 drivers. So sure

>Niagara shines on multi threaded work under huge loads. It never chokes, which Power6 does.

What are you talking about ? Try slamming your 64 minibusses and 4 Large busses onto a public road, and see who messes up. The sheer

complexity of having 64 threads versus 4 is exactly why the niagara chokes.

>"IBM has 100% right to count cores instead of sockets"

>

>Yes of course. But if you talk about the highest performing CPU, the FASTEST CPU, the you should compare cpu vs cpu, not compare core vs core. Right?

You seem to forget that before Intel and SUN started calling multi core, chips for CPU's, a CPU was equal to what we today call a core for.. well 60 years

or so. And it worked it confused people like you. So why don't we start to call a pig for chicken. It will do wonders for the PIG meat sales.

> Maybe all SUN lovers should write that to? Post as "Anonymous Coward" and write "We love IBM and Power6, but when we tried the new Niagara it smoked!

>And it is one fifth of the price. So we are now swapping our Power6 to Niagara. Sadly.

>I really liked Power6, but clearly Niagara is better and way cheaper. We just can not afford 413.000USD P570 vs a 76.000USD Niagara.

>And the Niagara smokes P570 too".

Well I've seen people throw away their new Niagara boxes, and replace them with partitions on old POWER5 partitions on p595's. Why ? It made sence from

a licing perspective, and running RAC on the Niagara boxes performed so badly that they actually tried to move it to the older v880 test machines to get

some throughput.

And if people are stupid enough to only focus on the cost of the hw box, then let them I don't care. When they mess up their IT infrastructure, then come

to us and we'll throw the lot onto a large POWER server and we'll give you more performance, better uptime at a lower cost.

// Jesper