Google's call for cloudier, taller disks is a tall order says analyst Storage analyst firm Trend Focus says Google's call for the storage industry to make more cloud-friendly disks is a tall order. Google recently threw down the gauntlet to disk-makers, asking for taller, denser and smarter disks that would help it to lower costs and scale its storage rigs to even greater heights. Trend Focus' …
Apple only changed connectors and electronics (to give direct access to temperature sensor bypassing SMART). That is trivial.
Google is asking for a different casing design to accommodate extra platters. That in turn means different mechanics, possibly different motors and in a worst case scenario re-computing the aerodynamic models of how everything moves inside the casing. That is a tall order.
> Google is asking for a different casing design to accommodate extra platters. That in turn means different mechanics, possibly different motors and in a worst case scenario re-computing the aerodynamic models of how everything moves inside the casing. That is a tall order.
I'm sure the National Museum of Computing can lend them a Western Digital 5.25" 5MB drive that they can copy.
I cannot see why the disk makers will avoid this. Google has the resources to do this themselves if they wanted, and adding this design to the Open Compute initiative will give Google more than enough scale. In the reverse, the first of the disk vendors to do what Google suggest and then incorporates this design into an Open Compute design wins. Big. Soon you say there will be only a few Cloud Service Providers, so it makes sense to get in bed with them now, before you get left out. You're going to loose the disk sales in any event. Google does as it pleases and a boycott wont stop them from achieving their goals.
This isn't a boycott, it is a business decision that the R&D expense to create these new tall disks isn't going to be worth the risk. They'll be fine with it if Google buys an old hard drive production facility and tries to roll their own - let them spend the money to trying to figure out how to make it work, and endure the loss if it turns out to not be cost effective.
The drive makers are struggling for survival in a shrinking market with the high end high margin products gone and only commodity products remaining. Google probably buys under a million drives a year, and won't be wanting to pay premium prices.
I take on board what you say but I can guarantee you this: there will not be a market for platters outside a data-center in less than 10 years - all 'personal' storage will be SSD or similar and the rest of your digital treasure in a data-center, on spinning rust. If you can fabricate a more efficient drive and you have a market that includes Google, Facebook, Rackspace, etc. and that's the only place you can sell a platter too, you should do it. And hope that you have SSDs as an SKU.
"If you can fabricate a more efficient drive and you have a market that includes Google, Facebook, Rackspace, etc. and that's the only place you can sell a platter too"
Then you won't have a large enough market to stay in business and/or the drives will sell for substantially more than they do now.
Either way, SSDs will out-compete spinning rust.(*)
(*) Generic. NAND or crossbar or any other tech, the future belongs to non-moving parts.
I think you're putting the cart before the horse. Google is a mighty behemoth on the Internet, no doubt there, but it is not a hard disk maker - yet. It therefor relies on the disk makers to get its storage products.
The disk makers have the imperative to be profitable in order to survive, and that profitability is getting thinner and thinner every year. If the disk makers decide that they cannot risk it, they won't and Google will just have to go buy existing disks.
Or it plops money on the table and creates the Google disk. That would be Google doing "as it pleases".
In any case, the number of Cloud Service Providers is irrelevant. They all buy disks, they cannot not buy disks and they're not going to change how disks are made simply by pouting.
Just how a purely cosmetic change destined to enhance consumer lock-in provide in any way a counter-point to said arguments ?
Were the disks of larger capacity ? Faster ? More reliable ?
Don't think so. The point is irrelevant from where I sit.
"Or perhaps he's right and Apple's famous commitment to integration of hardware and software meant it was willing to tolerate extra costs that Google clearly finds unpalatable."
Seems more plausible. Apple has the profit margins on its iThings to pay just a little bit more for something they want without getting hurt.
On a related note: with all the snooping and data slurping going on already and all the proposed snooping and data slurping that is to come you'd think disk makers would be working overtime...
One of the physical reasons that support the high bit densities of today's drives is the extremely short axle on which the platters (typically two, IIRC) ride. Today's drives are approaching or exceeding 1TB per square inch, which means one bit occupies a space approximately one millionth of an inch on a side. To accurately locate that bit, the head has to be accurate to about 1/2 that, or 5 ten-millionths of an inch.
Now you have an axle spinning little disks at 7200 (or 10,000, or 15,000) RPM. The longer that axle the more distortion is going to occur depending on even an extremely small amount that the mass of the system may off center. The axle will bend slightly, and there is a small amount of 'play' inevitably in the bearings that the axle rotates in. At these scales, there is little room for fudging.
This may all be fine - it may be possible to do. But it will also require substantial research and testing to develop what is effectively an entire new disk mechanical technology. For perspective, the two largest disk drive makers, Seagate and Western Digital, have been de-emphasizing (or outright elminating) there advanced research efforts. These companies don't see any future in disk drives, as today's SSDs have equivalent or higher capacity, equivalent or better reliability and lifespan, and no mechanical constraints. SSDs with capacities of 13+TB are being publicly mentioned, and higher capacities are reported to be in the labs now. For enterprise-level SSDs the write count limitations now exceed the projected lifespan of hard drives. As a clincher, I was told recently that at least one drive maker executive has said they see the end of hard drives within five years.
It appears that the last remaining hurdle for SSDs to completely replace hard drives in all applications is price, and that is following Moore's law pretty well. So all in all, I can not imagine any hard drive maker taking up this challenge - they would do better to convince Google to just move entirely to SSDs, and cut a deal with Google as a guaranteed buyer to supply the volume required, justifying a new, high efficiency plant that can bring production costs down to below the hard drive price.
They say "The current 3.5” HDD geometry was adopted for historic reasons – its size inherited from the PC floppy disk." The 3.5" drive form factor allows 8 of them to fit inside an 8" floppy drive bay. I have no idea why the 8" floppy drive size was chosen but I suspect it could be something as odd as the dimensions of rack mount gear is related to bee keeping frames.
"Bring back the Quantum Bigfoot!" ...
To add my part... The track-zero aka circumference of platter should be as large as possible inline to bigfoot concept, the spindle portion i.e. inner most track should also as large as possible. may the spindle axle remain same in diameter but the writable data be denied to several inner most tracks. .. say last/inner most track circumference should NOT be lesser than 30% of trackzero circumference. this will avoid the crawl speed reading of data from inner most tracks where spindle is rotating still no matter at 15k. Inner most 30% area of platter (think platter circumference) should be plastic not even metal to keep those shut asking about its wastage. or there should be a firmware unlock feature only those stupid will enable for themselves.
Also there should be a firmware based defragger. the world has suffered enough damage at hands of OS based defraggers and their fancies. the firmware based defragger should intelligently keep moving the hottest data towards trackzero aka outermost track/ring.
Well, if it hadn't been for the success of SSDs, Google's interest in producing a different kind of HDD might have some merit. HDD capacity has gone from 1TB to 10TB in nine years. The price for a 1TB HDD has fallen from $0.32/GB to about $0.05/GB. SDD capacity has gone from 1TB to 15TB in three years. The price for a 1TB SSD has fallen from $0.60/GB to $0.30/GB. HDDs still have a price per GB advantage, but SSDs has won when it comes to capacity. As production of HDDs declines, the price per GB will not fall much lower. As production of SSDs increases, the price per GB will continue to decline. Seems very likely that by 2025 HDD production will end. SSDs or their successor will have taken over in both capacity and price per GB. HDDs will fight on with SMR for specialized storage and HAMR, if it ever proves commercially workable, but it is a losing battle for HDDs. Not bad though when you consider that the first rotating magnetic disk storage device was commercially sold by IBM in 1956. Sixty years was a good run.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
