Seagate-Steve trash for Apple-Steve flash pash The MacBook Air is not a pointer to a diskless notebook future, according to Seagate chairman and CEO Stephen Luczo. Luczo made this pronouncement while discussing Seagate's poor results with financial analysts. The headline first quarter results for Seagate's financial 2011 set the scene: revenue rose $34m to $2.7bn compared …
In case you hadn't noticed, El Reg has been thoroughly indoctrinated into Jobsian cost-related doublethink. The fact that apple hardware is stupidly expensive is always glossed-over by reg hacks, who evidently have deeper pockets (or less sense) than their plebeian readership.
"If one is to rule, and to continue ruling, one must be able to dislocate the sense of reality."
...we don't talk about the cost of Mac toys because we know better than to buy them? Something about getting better value for money.
--Sent from my netbook using my HTC Desire as a perfectly legitimate and carrier-allowed MiFi point. (The pair of which still cost me less than an Air.)
I sold my six-year-old 1.67 GHz Powerbook a few months ago for £500, which wasn't a bad return on my original investment. I then went out and bought an ex-demonstrator 2.8 GHz Macbook Pro and got it for 800 of your English pounds. It's not out of warranty until December and it's a thing of beauty.
I still run a Umax Powerlook scanner on an 11-year-old Apple G3. It has a G4 processor inside it, mind you, but it works like a charm after all these years. I dropped it accidentally a few weeks back when I was moving it across the office, and when I plugged it in to see if I had killed it, it fired up and is still in perfect working order.
When I buy a computer I treat it as an investment that will not only perform well and reliably for years to come, but will have a high resale value when the time comes for me to get something newer. This makes Macs an absolute bargain, even if one does have to pay a little more at first to move away from the herd's standard fare of disposable Airfix PCs.
I wish you well with your netbook and hope it makes the full two years before it gives up the ghost and dies. That's not inevitable, of course, but it is likely. I expect my Macbook Pro to perform superbly well for the next four or five years or more, when I can hope to sell it for around £500 - perhaps a little more. So you paid a couple of hundred quid for your netbook and I paid £800 for a not-quite-brand-spankers Macbook Pro. I know which I'd rather have.
Treat your toys well and they will treat you well in turn.
You see, the issue here is one of perception. I blame the low value/low quality vendors such as Acer. If you lop the bottom 20% of the PC market off, then everything you are saying is crap. (BTW, the nVIdia eutectic bump bit with the 8000 series? Affected the Macs too…)
I have a Dell XPS Gen 2 – one of the first shipped – that has survived more crap than I would throw at a toughbook. Fallen off of moving cars, dropped down concrete stairs, many coffees spilled. It has never taken much TLC to bring it back after a major incident. Shock and gasp though…being a PC, parts were available! I have a Toshiba 386 Laptop that is still around and doing it’s job. I have a dozen PCs of various flavours all older than 3 years old. I have a FLEET of 15 Pentium III PCs from 1999 that are still in operation!
As to netbooks, well so far I have seen zero degradation on any of the ones I have worked on…except Acer. The Asus one, the HP ones…they’ll be around for a /very/ long time. See, the trick is to take care of them. Periodically remove the cover from the bottom of the device and blow the dust bunnies out. Don’t force the power connector: treat it properly and realise it’s a weak point. IF you spill coffee on the thing, then turn it off, pull the battery, remove the mainboard and soak it in distilled water. (Let dry for three days before reassembling.)
My newest box is an Alienware M17. It’s about two years old. This one is a completely Novel device in the PC world because it has air filters on the fan intakes. (Shock!) This means that in all the time I have owned it I have never found a dust bunny inside. Oddly enough, I don’t actually pay all that much for these things. Even the Alienware was 25% less than the highest end Mac, whilst still being double the specs.
Macs have literally nothing to offer over PCs except the Magsafe connector. Frankly, I buy XPS and Alienware laptops for my primary because of the enormously beefy power connectors they use; they are powerful enough to take any banging on the power connector.
So to my mind, your entire argument would rest on “Buy a Mac because it has a Magsafe power connector!” I don’t think so. I’ll buy TWO PCs for the same price (of equal spec) and just throw one away if/when the power connector gets killed.
As always, the above logic disintegrates if you pay any attention to Acer/Gateway/E-Machines/low-end-Dells/etc. Use Asus/HP/Toshiba/Fujitsu/Lenovo/High-end-Dells and suddenly you have reasonably reliable gear.
I should also point out that we have to maintain a Macbook at my work so that we can run our Java-based order entry software on it (HAH!) in the environment our Mac customers have so as to troubleshoot their issues. I have had to rebuild the thing three times in it’s four-year lifespan because it’s a poorly put together piece of crap. It doesn’t go anywhere. It just sits on a desk in a fairly dust-free office and gets turned on once ever two weeks to troubleshoot some odd thing or another. I really wish it would die once and for all so I can stop supporting it. Thanks to Jobs’ Java faffery, it won’t be getting replaced!
So yeah, I’ll enjoy my netbook. For the next six or seven years.
Given the lowering life of hard disks as the capacities rise I am almost glad that this downturn is happening. It will hopefully make companies like Seagate turn their attention to driving up the capacities of solid state storage (which will make them viable alternatives to hard disks thus reducing their price as they become more competitive) as well as finding ways around some of the limitations of solid state (such as the limited writes of Flash memory).
Also, solid state will hopefully make computers quieter!
SSDs are simply too pricey yet vs the performance benefits. Flash is approaching HDD speeds however and is a good option. However, mass storage isn;t going to compete with disk for a long time. On a Tablet or ultra portable, 64-128GB is sufficient. on a main machine, hell no it is not!
That said, hybrid IS the future. Segate is dead wrong, as is the writer of this article. By Hybrid, what I'm really talking about it tiered storage. The EKY to tiered storage is not the hardware price, it;s the OS and the user experience.
If what i suspect is the case with OS X 10.7, one of the fundamental features we have not heard about is that the OS "anonymizes" the storage it finds. Put a Flash pack on the motherboard, and a storage drive on SATA, and the OS will present a single disk to the user. The OS will manage what parts of the dataset (at the block level) are more frequently accessed, and put them in Flash. This is a SIGNIFICANT performance improvement if it is done intelligently, and comes at only a small cost increase vs a single traditional drive. In fact, having mroe than 1 spinning disk provides similar benefits, as the OS can alternate blocks between the drives, then sort them out later when idle.
We've been using 2, 3, and 4 tier storgae models like this in the SAN world for a decade or more. The technology can finally tricle into client-side OS. If the OS handles it, a small SSD (or flash) can handle the OS and some core apps, and any file writes a user is active in, and the OS migrates unused blocks over to HDD as a background process when the drive buffers are low.
Seagate is right, spinning disks won't go away, not for anyone's core machine. For mass media storage its the best option. in fact, it;s going to be replacing tape in the enterprise here soon too. Disk sales should increase dramatically in the channel, however, in point systems, tablets, business machnies, and places where mass storage is not required, flash boot provides significant advantage at a small cost.
As for media consumption, I think we're going to see the cloud start meeting performance needs in a couple years, 2014 at the latest, where people will simply stop buying large scale disk at home entirely. The holdup is not the technology, it's the bandwidth, and the pricing models for content (and protability and sharing of the content) keeping it from wide adoption.
Seagate is deluded if they think we'll keep buying HDDs heavily as consumers past 2014. By then simply flash (let alone SSD) will exceed HDD performance, and possibly some of it may exceed HDD capacity, at comparable prices. Seagate better get on board shifting their business plan, or they'll be in the same shitstorm record companies are.
Your initial comment that 'flash is approaching HDD speeds' is strange, the current generation of SSDs operate at speeds which simply annihilate the fastest mechanical drives - double or triple the sustained read/write rates, almost-zero latency and a ridiculous improvement in I/O operations per second.
You seem to be drawing a line between flash and SSD which in fact does not really exist; both are controller devices slapped onto 1 (in a flash pen) or more (SSD) NAND chips. An SSD is in essence a flash pen with a more sophisticated controller device, a SATA interface and multiple NAND chips operating in parallel.
I mostly agree with the rest of your essay, although I think you are overly optimistic to think that consumers will not be buying HDDs in 2014 - unless there is a MAJOR leap in the production of NAND chips in the very near future it will take considerably longer than that for SSDs with 1-2 TB capacities to be affordable alternatives to HDDs.
As for hybridisation, this is already taking place in the 'custom or self-built PC' market, where it increasingly common practice to use an SSD as the operating system drive and then larger mechanical drive for everything else.
Sorry Seagate, and sorry Michael C. Yes, spinning rust is dying. It won't happen overnight, but it is highly likely that within 5 years there will be no new 15K or 10rpm drives shipping. We'll have some around for spares and replacements for another 5 years after that, but that's about it.
I'm guessing Michael has hung onto his Adaptec SCSI cards for a while too. Sorry, SCSI as a physical transport is dead (but lives in FCP and SAS), just like spinning rust will be.
Yes, at the 3.5" form factor, and at the 7200 and 5400 rpm speeds, we'll continue to see large capacity drives for at least 10 years, maybe even 15 or more.
But, for smaller capacities (read under 1 TB) and higher performance the future is all flash. Not just the current crop of NAND flash, phase change, and others are coming too. Silicon improvements continue at around 2X every 18 months, spinning media can't get much faster, and the paramagnetic effects are starting to be problematic.
So Apple got this one right. If you need less than 200 GB, and want high performance, and low power, flash is the right answer... today. In the future, that will be even more of the case.
The concept that ZFS uses is truly what Seagate is trying to accomplish. A large array (or just a platter) for actual data storage with a high-grade cache storage in front of that (SSD array, or a small flash cache in Seagate's case). The cache doesn't actually permanantly store data, it just caches the data from it's permanant home on the platters. DIYers have attempted a close approximation of this by having an SSD boot drive, but the problem is this still makes Crysis crap on load times, or perhaps something in your Adobe suite slow, simply because it all got lumped on that "mass storage" drive. What's the solution? How about a proper 32+GB "cache" for the hard drive and a "right-click -> Add folder to flash cache" OS option? The more user-picked cache data, the less space for the "hot data" algorithm, but who would know best that you wanted to load ALL your Crysis map data into the cache? Permanantly. (well until you delete the data or "right-click -> Stop caching". A wimpy sudo-algorithm with a miniscule 4GB of cache isn't going to help much. I'd by a 64GB flash cache + 2TB platter space drive in a heartbeat over a 64GB "boot drive" and a 2TB spindle drive. Especially if it had a "right-click -> Add to cache" option.
/jogs down to the patent office
SSD won't replace spinning disks ... the limited number of writes is a no-no, especially after one of my pendrives failed after 2 years because of this limitation. However, it does make sense on laptops or netbooks that are used in bumpy environments ... you can shake, bump and hit an SSD drive without crashing the disk. They're also impervious to magnetic fields as well, so these environments are ideal for SSD.
SSD won't replace spinning disk simply because of cost and capacity. There will always be demand for larger devices. There will demand for larger devices than are available in SSD and demand for lower cost solutions. That demand might be mainly driven by corporations but it will remain.
iPods got up to about 120GB in capacity and then Apple replaced them all with flash based players of much smaller capacities. Sales didn't fall and they still sell well.
Power users are always going to want the biggest disk, most RAM and so on. Not everyone wants to store loads of video or audio on their laptop.
The only reason Seagate would be concerned is that in recent years their products have become crap. I have had all my Seagate drives swapped out for WD. I used to swear by Seagate until they started failing sooner and on a grander scale and saw it increasingly noted in the trades that Seagate build quality dropping more and more in recent years.
Bye Bye, Seagate.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
