Intel releases eight-headed* beast Intel offered only a few new details about its latest processor offerings at the International Solid-State Circuit Conference (ISSCC) today, even though it essentially had the stage to itself. AMD, IBM, and Sun chose to sit out the Microprocessor Technologies session, forcing the hundreds of engineers hungry for CPU info to …
"Intel releases eight-headed* beast"
Cores and heads aren't the same.
This single core laptop has three heads ... the built-in laptop screen, a largish external monitor, and an IBM type 3151 terminal + IBM Model M keyboard.
Please learn the difference if you wish to be taken seriously as a tech writer. Ta.
hmm... Rock is first several years late and then it's killed ?
was it Nehalem or Tukwila that killed the Rock... or maybe it was IBM POWER !!!
Where is Sun technology heading now ? Fujitsu ?
here you can see a roadmap from year 2004, where Rock is planned 2H2006 : http://www.amdboard.com/sun_roadmap_0505.jpg
And then a roadmap from year 2008, where is no more commitments shown : http://regmedia.co.uk/2008/07/14/sun_chip_roadmap.png
and now the latest roadmap where the Rock has disappeared totally ?
And it's not looking any better for Itanic ( although Tukwila finally comes this year, being only 2 years late ) - Itanium roadmap from year 2004 : http://www.ferra.ru/images/171/171346.jpg
Interesting stuff.. Nehalem seems to be a killer product !
So it's Intel Xeon vs. IBM POWER from now on...
"ROCK is gone....I guess it did not make the January RIF" .... By Anonymous Coward Posted Tuesday 10th February 2009 05:06 GMT
AC,
It has gone Renegade Native/into Deep Cover Underground for Stealthy Uninhibited Progress. I noticed that Victoria Falls is also a "black" and unmentioned in the picture .... http://sunroadmap.weebly.com/uploads/1/8/1/4/1814344/5639922.jpg
Which is a good/bad sign just depending on who you would be Rooting and Routing for.:-)
Ooops. I should also have preceded "Don't count your chickens ........" with ...... "Managing SMT extracts a performance penalty, but it should be less than 10 per cent, according to Intel's Rajesh Kumar."
There is obviously a coding/architectural flaw, which will always punish Intel, whenever "simultaneous multithreading (SMT) technology to support two simultaneous threads per core." extracts a significant performance penalty, rather than delivering an additional performance boost.
Which is probably exactly what the competition has perfected.
that would be "metaphor" and possibly "humour" in this venue
on topic .. AMD releases 45nm with no significant advantage .. yawn ..
Intel ahead of schedule with good "product health" on 32nm process .. meaning they are getting good yields .. big investment in quality US plants .. all good ..
The new process means less power drawn, less heat generated, higher clock speed and lower manufacturing costs
some people live by "cheap is cheap", so I quess AMD will survive a while yet .. z z z z z z z z ...
I don't know where AC got his bogus roadmap from, but he should ask a Sun contact to show him the latest public document! The SPARC 64 VII line includes Jupiter+ and Jupiter++, as you'd expect, and the CMT line includes Rock and two 40 nm processors code-named RF (Rainbow Falls - formerly known as KT) and YF (Yosemite Falls).
Wow, the Sun roadmaps are so obviously faked. Notice the different fonts?
That begs the question. Who would be so scared of ROCK that they would go through the effort to spread such obvious FUD?
IBM? With their gigantic layoffs and forcing employees to move to other countries if they want to keep their jobs? Perhaps the dwindling Unix sales are killing them?
How about HP? They seem awful excited about the release of Tukwila... Perhaps the never ending feature fade and continual delay is getting to them? It's hard to believe that their customers would want to actually do the required forklift upgrade with times the way they are now.
Or is it just a HP or IBM Fanboy? Matt? Nah, Matt isn't too into facts, but would he out and out lie? It's probably just some other Fanboy with way too much time on his/her hands.
This isn't supposed to be an academic journal, you know!
Perhaps in your view the title should have been something as exciting as "Intel releases eight-core cpu". Well, it's accurate, but some folks also like a little humour in their lives, and I would be fairly confident of guessing that most people payed very little attention to this gross misuse of this word 'head'.
Perhaps you'll understand if I put it this way - (I presume you spend a lot of time coding!):
When compiling a sentence, if either the HUMOUR or JOURNALISTIC_LICENSE symbols are defined, then many more words are compiled with overloaded constructors, taking a variety of parameters to denote a variety of meanings!!
Sounds like a killer chip for the x64 world!! If it comes out ready to scale to eight sockets, then that's 64 cores in probably about a 10U server size rack server, or 256 cores in a 2m rack. You could probably squash eight sockets into a smaller chassis but with that many cores you're going to want to have plenty of space for memory and IO slots. Now that would make for a monster ESX system or very tasty MS SQL server!
And I can't wait to see how it gets slipped into blades, if QPI can be matched to a new c-class blade chassis backplane that acts like the cross-bar switch backplane in a Superdome then hp will be able to make an ES7000-like x64 "mainframe" in a 10U blade chassis. That would knock a few points of the IBM mainframe revenue figures.
"This isn't supposed to be an academic journal, you know!"
Well, duh. But there is a time and a place for everything.
"Perhaps in your view the title should have been something as exciting as "Intel releases eight-core cpu"."
In THIS particular case, it would make more sense, yes.
"Well, it's accurate, but some folks also like a little humour in their lives,"
I agree. That's what Odds & Sods and Public Sector (and to a lessor degree, Security) are for.
"and I would be fairly confident of guessing that most people payed very little attention to this gross misuse of this word 'head'."
Most people aren't qualified to run computers, either. I suspect that the article in question would have received a lot more page hits from professionals with a non-frivolous subject line. Speaking only for myself, I nearly missed it ... and it's a subject I'm interested in.
"Perhaps you'll understand if I put it this way - (I presume you spend a lot of time coding!):"
Actually, at this time of year I spend a lot of time getting ready for mares to foal and/or doing all the little groundwork things that help prepare the babies for when it comes time for more formal training. In general, I don't code anymore. I hire someone to do it for me. My time is more useful elsewhere.
"When compiling a sentence, if either the HUMOUR or JOURNALISTIC_LICENSE symbols are defined, then many more words are compiled with overloaded constructors, taking a variety of parameters to denote a variety of meanings!!"
Whatever. As I said, English is a precise language when used properly.
A scared competitor putting out an obviously faked roadmap? I still can't tell which vendor, and probably not really a vendor but an overzealous employee, is spreading this around... It doesn't even look real. Notice the wrong fonts?
My guess would be IBM as POWER seems to have lost a bit of the excitement that it carried in the past. Of course it could be HP as they are waiting on a woefully disappointing Itanic update. But Itanic is getting lot's of press right now and the HP faithful wouldn't feel the need to put out obviously false data. That is, except for Matt B, as he always finds the need to put out obviously false data.
Fake Roadmaps, incorrect proclamations, bad assumptions and a complete lack of fact or vendor information.
Sounds like someone is trying to hide some of their own problems.
If Rock is dead, nobody has told Jonathan, John F, or anyone else that matters at Sun... (See recent articles already cited)
Go on, El Reg - Post an article pointing to the real SPARC roadmap. I dare you.
And - see if you can find out when Intel actually plans to release Nehalem in anything other than 2 socket variants. I read it's not till 2010, and if past experiences with large, scalable NUMA systems is anything to go by, don't be surprised if it's even later than that. Think back to the last time they tried...
post it here...give a link....I would hope Sun has the latest roadmap on their website...I hope it has dates though....last one I saw had the dates deleted.....then Fujitsu's does not even have rock on it
http://www.fujitsu.com/img/SPARCE/technology/protection/roadmap01.gif
All we know is ponytail "thinks it's later this year"
you can easily find intel roadmaps
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
