Slashdot Mirror
Are Consumer Hard Drives Headed Into History?
Lucas123 writes "With NAND flash fabricators ramping up production, per GB prices of solid state drives are expected to drop by more than half by this time next year to about 50 cents. Even so, consumers still look at three things when purchasing a computer: CPU power, memory size, and drive capacity, giving spinning disk the edge. SSD manufacturers like Samsung and SanDisk have tried but failed to change consumer attitudes toward choosing SSDs for their performance, durability and lower power use. But, with the release of the new MacBook Air (sans hard disk drive), Steve Jobs has joined the marketing push and may have the clout to shift the market away from hard drives, even if they're still an order of magnitude cheaper."
Even if the Per GB price dropped by 80 or 90% SSD's would still be more expensive and have a lot shorter life expectancy than current HDD's, we are many many years before the possibility of SSD's fully replacing HDD's becomes even conceivable
$0.50 per GB is still about five times the cost of a magnetic drive. Put another way, each user has the choice between paying $50 and $250 for the same amount of storage. Does anyone think there is a real competition here?
And of course, that's by next year. How much denser/cheaper will magnetic drives be by then? Please stop with these "year of the flash drive" posts.
Why would I switch to SSD? I've had 1 drive go bad in my lifetime. They've lasted in some cases 20+ years. Plus they are cheaper. Why would I bother buying SSD's when they have a known failure point at after given number of writes?
This is very much like the blue-ray issue. It's not surprising folks aren't interested in jumping on board because, frankly, there is no real reason to run out and BUY it.
CD's and DVD's had huge adoption because you saw a large improvement on your existing hardware. Bluerays required a new TV to see that improvement - and it was a very expensive TV at the time.
Once people have purchased new TV's (it will probably take another 5-10 years for the older TV's to all fail so that the mom and pops of the world HAVE to go buy a new one) blue-rays will have come way down in price and they'll finally replace the DVD.
Likewise the SSD. I'm sure many other folks are as tired as I am regarding these silly... strike that... STUPID press releases trying to push their sale.
They will be bought when there is a need. There is none at this point, except in very specific applications, like the high-vibration atmosphere at manufacturing plants.
Shame on Slashdot's editors for continuing to run this hokey marketing BS, and shame on the people who continue to send articles like this. It's quite silly, frankly.
In the MacBook Air, the SSD chips are soldered to the logic board. It is not like there is a choice on what kind of drive can be installed. When 64GB isn't enough, there is no way to upgrade. When the SSD gets a fault, there is no drive to swap out - it would be time for a new logic board. With NAND Flash having a finite lifetime, soldering the SSDs to the logic board is a prime example of planned obsolescence. When the SSD dies (when, not if), there is only Apple to turn to, so Apple effectively has vendor lock-in as well, but we have come to expect that from Apple.
No, the SSD's are on a removable board. See http://www.ifixit.com/Teardown/MacBook-Air-11-Inch-Model-A1370-Teardown/3745/1 (It's the thing that comes off from above the RAM)
Back when CD-Rs were new, we were hearing how they'd last for well over 50 years. Now we're finding that CD-Rs last only 3 to 5 years, and that's when they're stored in conditions that are near-perfect.
It's pointless to take media lifespans measured in decades as anything other than marketing bullshit, especially given that the computer industry itself has only been around for about 65 years.
It's not only the bigger size and lower price that makes HDDs attractive. The worst case random write time, for example, is generally far worse for SSDs, and if you absolutely need to commit within a guaranteed time frame, SSDs might not be an option even if they're much faster on average, and orders of magnitudes faster for random reads.
Don't underestimate the power of a rack of short-stroked 15k rpm drives.
feeling like a 14 year old girl is not a good thing at all.
Not once I'm through with them.
My experience differs. My employer does on-campus interviews at around 50 schools nationwide, and over 70% of the potential recruits were equipped with macs.... according to the web logs from a site they were required to visit individually.... Our public web site has seen a marked increase in 'mac' traffic as well.... averaging around 10%, up from 3% a couple of years ago.
We're not hating. We're just tired of getting a story every time Jobs farts.
-- I have a private email server in my basement.
Um... Small random reads are the primary pattern in desktop usage. Are you a complete idiot? That's the SUBJECT under discussion, not dumb shit like sequential transfer speed. That's only important for marketing people who like big numbers with MB on the end.
No, small random reads are NOT the primary pattern in desktop usage. Almost NO file on your file system is under 4k in size, which is the "chunk" size for most 8mb to 64mb hd caches.
Even DOS didn't have average file sizes that small. And many of today's hard drives also have implemented the elevator algorithm in hardware, so head seek times, especially for small random files, are much less of an issue than they once were.
4 drives with 32mb hardware caches will outperform your sdd in every scenario, including small random writes - especially since, for the same capacity, they can be grossly under-stroked - limited to the outermost few tracks. Understroke a 1TB drive to 32 gigs and its' seek times drop to almost zero. Throw 4 of them into a 4-drive setup as /, /home, /var, and /srv, and you'll beat the 128-gig SDD in small file r/w, and massively beat it in large file r/w.
NPD is a reputable source, but I'd like to see the actual study myself, rather than hearsay.
Note that it's 20% of the "consumer retail" market share. That is to say, individual people buying boxed computers from a shop. A rather large percentage of computers on the internet are purchased by companies and either (1) installed at their offices and then used by employees for personal purposes during their breaks or (2) loaned to employees to take home. This happens very rarely with Macs; in my 15 years as an IT consultant I've worked with precisely one company that installed Macs. A smaller, but still nontrivial, percentage of the sales of computers is in markets that aren't typically classified as retail: all those people who either buy components and self build, or buy from local "we sell to trade only, honest" shops, or from computer fairs are probably not included in these stats.
Also, market share != installed base. Note that if the average Mac user changes their machine every 3 years, while the average PC user only bothers upgrading every 6, that will double the market share of Macs relative to their installed base.