• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Why no other hybrid drives?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
If SSDs become cheep enough soon enough, there will be no need for a HDD, except for specialty purposes- which, by the way will widen the price difference between the two drive types in the opposite direction.
Not likely. HDDs still pump out higher capacities, and we keep making up things to fill them with. They also keep improving overall performance. It will take decades for them to become products for small niches. HDDs aren't going to go away any time soon, they are just going to not look so good for a primary system drive, for more buyers, as prices fall.

What we might see are normal users getting a SSD inside the box, and then external HDDs. Most normal people already get an external drive, so as not to have to open up their machines, or pay someone to install a drive.
 
Not likely. HDDs still pump out higher capacities, and we keep making up things to fill them with. They also keep improving overall performance. It will take decades for them to become products for small niches

1. He said IF SSDs become cheap enough faster enough, if you read his posts you would realize that he doesn't think that will happen for years yet.

2. SSD already have higher density than HDD, a fully packed SSD can physically contain more info than the most dense HDD (with 5 platters @ 3.5 inch). The ONLY reason HDD have the "size advantage" right now is because of price, a 2TB SSD would just be ridiculously expensive. Here is a 1TB SSD http://www.amazon.com/OCZ-Technology.../dp/B003DYF6YA note the 3000$ price tag.
with current tech you could cram ~3TB in an SSD, more than any spindle drive in the market... but it would probably be 10,000+$.

3. HDD performance only increases due to density increases, in fact the trend in the last few years was to make LOWER performance HDDs (with worse performance than their predecessors) which are cheaper (to further accent the "cheap and large" comparison to an SSD). SSD performance is rapidly increasing while HDD performance is inching along. There is absolutely no comparison in terms of performance and there never again will be. current and future tech SSD are massively faster, period.

What does all of this mean? it means that for the foreseeable future we will have SSDs for faster system drives (fast and expensive, so most will get a small one), and spindle drives for bulk storage (cheap, people will be buying large capacity drives). As prices on SSD goes down, they will cannibalize more and more of the HDD market. The only thing stop them from completely overtaking the market right now is price.

PS. I have no idea what specialty purpose will require a spindle disk over an SSD... SSDs are more durable and robust in every way shape and form... can't think of a single case where a spindle drive is preferable.
 
Last edited:
While I like to tinker with my desktop computer to keep it running at top speed with some of the latest bells and whistles, my laptop need not be so snazzy. The things I do on the road are pretty mundane, basically e-mail, a little word processing, and download/watch movies/programs.

With that in mind, and the fact that I just bought a new laptop (Acer Travelmate 8172), I'm pondering whether to get a Momentus XT drive. Even here in Thailand, I can get a 320GB version for just under US$125.

My main concern probably shouldn't be a concern, but I just hate it when I buy something -- in this case a 320GB Momentus XT -- and the next week a newer, better, cheaper generation is released. I say it shouldn't be a concern, because objectively I need to decide now if the 320GB Momentus XT will be a wise purchase based on price and performance. What happens tomorrow, objectively, should not matter. But subjectively I would pout. 🙂

Based on the discussion above, however, it doesn't sound like any manufacturer is even paying attention to hybrid drives, much less getting ready to roll out new models.

That said, I'm going to (try to <g>) wait until 2011 to see what unfolds this fall after the introduction of the new generation of Intel SSDs.
 
You guys are talking like like average Joe PCs are the majority of the hard drive market.

Google farms, cloud computing, and office PCs are the market.
These are either extremely price sensitive markets or extremely capacity sensitive markets.

Platters aren't going away in 3-4 years. Maybe for enthusiast PCs, but until SSD is cheaper than platters at any one capacity point, platters will be predominant.

Expect the big platter companies to fight tooth and nail. They will design for specific low cost drives at popular capacity points to keep the advantage they have. Joe IT director makes decisions on cost, not performance.

Hybrid isn't popular because there's no way to market it. Benchmark a hybrid drive against a platter drive, and every synthetic benchmark barely has an improvement. Actually use one on a regular basis, and it's a noticable improvement. How do you market that to Joe IT director, even if he is performance conscious?
 
You guys are talking like like average Joe PCs are the majority of the hard drive market.

Google farms, cloud computing, and office PCs are the market.
These are either extremely price sensitive markets or extremely capacity sensitive markets.

Platters aren't going away in 3-4 years. Maybe for enthusiast PCs, but until SSD is cheaper than platters at any one capacity point, platters will be predominant.

Expect the big platter companies to fight tooth and nail. They will design for specific low cost drives at popular capacity points to keep the advantage they have. Joe IT director makes decisions on cost, not performance.

Hybrid isn't popular because there's no way to market it. Benchmark a hybrid drive against a platter drive, and every synthetic benchmark barely has an improvement. Actually use one on a regular basis, and it's a noticable improvement. How do you market that to Joe IT director, even if he is performance conscious?
I agree with your general sentiment, but addressing your middle paragraph: I think SSDs in business laptops is exactly where this market is going to explode sometime in the next 2 years.

The other place things have a potential to explode is PCIe SSDs in servers/SANs.
 
Seagate Branded SSDs

In August 2007 - Seagate announced it would launch a genuine solid state disk product sometime in 2008. (...Later:- That didn't happen, and the company merely repeated itself later - shifting the date to 2009.)

There has been much speculation about whether this will be done by rebadging, in-house design or acquiring an SSD oem.

Whichever route is taken - I predict that Seagate will almost certainly fail to achieve significant long term market share in the SSD market. The main reasons for Seagate's SSD failure will be:
insufficient enthusiasm to develop the new SSD business at a time when the hard disk market is still growing - also dampened by disillusion with hybrid products by that time.

Why should Seagate invest manufacturing resources in an industry where in February 2008 iSuppli said flash prices were already dropping below costs and the overcrowded market for 2.5" flash SSDs points towards an upcoming costly shake-out?


unwillingness to kill off its own profitable hard disk products by introducing new SSD products aggressively (a hesitation which its competitors will not have)


lack of corporate management skills - which are different when you're competing as a newcomer in a multi-vendor fast changing market like SSDs - compared to the sluggish manufacturing led pace in the decades old hard disk market - when you're the main encumbant.


http://www.storagesearch.com/seagate-fail-ssd-art.html
 
PS. I have no idea what specialty purpose will require a spindle disk over an SSD... SSDs are more durable and robust in every way shape and form... can't think of a single case where a spindle drive is preferable.
You could have just done that. The only diverging part of your reply was that (I'm not aure what #2 was about: I clearly typed out, "capacity").

However, we already have a situation where a HDD is preferable: my funds are not unlimited. As data sizes that need to be stored grow, that will not change. SSDs don't have a long way to go to become common main drives (I wholly agree that business laptops will go almost all SSD as soon as it is affordable to do so), but they do have a long way to go to replace cheap storage with spindles. My whole computer has only just barely cost me as much as it would to store my data on SSDs. As long as HDDs keep boosting capacity, and data that people want to store grows in size, SSDs will face an uphill battle to compete with that (on the bright side of things, such competition doesn't really exist).
 
we already have a situation where a HDD is preferable: my funds are not unlimited. As data sizes that need to be stored grow, that will not change.

Ah, but there is the rub. The thought is that SSDs will overtake disks in $/Gb. At that point, it will be both cheaper and faster to store all your data on SSD. We are not too far, relatively, from that juncture. I would guess, within a decade.

Incidentally, I don't know what a nitch market needs disks either. But there seems to be a nitch for most everything, so I wouldn't count it out.
 
Last edited:
we already have a situation where a HDD is preferable: my funds are not unlimited.

By definition, a "specialty purpose" is more expensive. We were discussing a case where spindle drives are effectively gone from the market, where SSDs are cheaper and plentiful.

There have been specialty SSDs in use for decades before they became consumer goods. If you need a computer on an airplane, a smart bomb, etc... you would pay ridiculous price for an SSD because it can withstand orders of magnitude more Gs.

SSD are:
1. More reliable, no moving parts to break.
2. Can operate under significant vibrations (again, no moving parts, no risk of head crashes)
3. Can sustain orders of magnitude (an order of magnitude is 10x) more acceleration (aka, "G forces")
4. Faster
5. More heat resistant.
6. More humidity resistant.
7. More cold resistant.
8. More resistant to changes in air pressure.
9. Can operate in a vacuum.
10. More resistant to solar radiation.
11. Produce practically no heat.
12. Consume a lot less electricity.
13. Completely silent.

In any odd military, manufacturing, aeronautics, etc situation that I can possibly imagine an SSD will be better... the ONLY reason to use a spindle drive is cost (SSDs are already more physically dense in size per unit of volume, only reason we are seeing smaller SSD sizes then spindle drives is extreme cost of SSDs)... if spindle drives will cost more I predict that they will NOT become a "specialty product" (very expensive, made for a few special purposes), but will instead become completely extinct.
 
Last edited:
if spindle drives will cost more I predict that they will NOT become a "specialty product" (very expensive, made for a few special purposes), but will instead become completely extinct.
Of course. The question is merely how soon.
 
The HDD will still be used where write-to-read ratio is high. Think logging (e.g. video logging).

logging is a very IO intensive task and quite a bottleneck, SSDs are actually taking over as logging devices. While a singular block on a HDD can be written to more often (in theory, I don't think it has been tested to the level that SSDs have), with wear leveling the total size of the disk is taken to account. So again it is merely cost. Pay more for an SSD log device to be larger, and it will handle more writes.
 
unwillingness to kill off its own profitable hard disk products by introducing new SSD products aggressively (a hesitation which its competitors will not have)
Remember that Seagate is a MANUFACTURING company building disk drives, not a marketing company and not a memory manufacturer. Seagate isn't going to be able to support its operations by assembling SSDs using memory made by somebody else. That can be done by a marketing operating (contracting out the assembly) or by a small manufacturing operation.

If Seagate can't find a way to sell spindle hard drives, they are going to have to develop other products that can use their manufacturing technology and assets. Or start over and become a memory chip maker. Or get rid of their plants and people and become a marketing operation.
 
Basically this is a big part of the reason Intel spent the last couple of years building new foundries around the world starting with the US. There has been a NAND shortage this last year thus the rise we've seen in memory for awhile.

Once Intel announces their new smaller, cheaper, NAND we'll see some prices drops in multiple industries. Intel is supposed to have doubled the size of their MLC SSDs while keeping the consumer prices the same if not cheaper. After all the chips are smaller thus made in great quantity.

If I can get an 80gig Intel SSD for $100 I'll be happy. If I can get the rumored 300gig for under $400 I will very happy. If they're SATA 6gbs I'll be estatic. Since Micron already has a SATA 6gbs drive out they should be since Intel & Micron are partners.

Edit: Damn no SATA 6gbs

http://www.anandtech.com/show/3965/intels-3rd-generation-x25m-ssd-specs-revealed
 
Last edited:
As noted, spindle disks will be around for a while, still. Over the next five years, the market for "small" disks will go away, overtaken by SSD. But "large disks" will still be sold in large volume. There's lots of places where the speed of SSD is not a big advantage and low prices ARE a big advantage.

Seagate could go into the SAN/NAS business, where spindle disks will continue to be used in large numbers. That'd use their disk-making expertise, as well as their controller expertise.
 
Last edited:
Still, at this point spindle drives only have a price per unit capacity advantage. If I had it my way I would choose to have no rotating media but obviously a 1.5TB drive, however possible it would be to stuff that many memory chips into a 3.5" package, would cost far more than the $100 or less you'd pay today for rotating storage.

Speed, reliability, noise, and power consumption all favour a (well-designed) SSD.

Hybrid drives are a (reasonably good) compromise for now since the mainstream is used to spending under $150 for storage upgrades, and a 500GB Momentus XT is well within that territory while the same users will scoff at getting 40-80GB of solid state storage for the same price.

That said, I have no regrets paying almost $400 for my first G1, and $200 for my second.
 
Oh, and I always forget to mention disks used for backup. In my small business and home installations, there are often as many disks dedicated to backup than to live usage. SSDs don't make good backup disks because they are being frequently totally re-written. Current SSDs would wear out far too quickly. A 500 GB SSD disk that's totally re-written every third night night isn't going to last long.
 
I don't run my backups the same way. Only changed data is re-written, and the rest is not changed. Since relatively little of my data actually changes, there aren&#8217;t all that many writes.


I don't know how long an SSD can sit on a shelf and maintain it's data compared to a HDD, but with periodic power an SSD would have an advantage here also. It could quickly check all it's cells, and refresh or move data that is going stale. Here again only cost/Gb is the disadvantage.

Like taltamir notes, when the price/space goes the other way, spinners will become extinct. If the HDD manufacturers wish to remain in the storage business, they must convert to flash. Developing hybrids now, while their disks still have value, would give them valuable experience with flash, while still using their current products, and for very little additional cost. I think they are foolish if they don't take this route.
 
SSDs don't make good backup disks because they are being frequently totally re-written.
Sounds like a rather strange backup strategy. I'd wager most businesses use something along the lines of incremental backups on site and monthly/weekly/whatever backups to other sides - both things where not that much data will be written to the same drive.
 
Also many businesses that do backup to disk go through the trouble of getting a solution that supports de-duplication, which really limits the writes to disk. The site I'm responsible for gets 1:38 storage:data results from dedup'ing our backups. It's an extreme example but it sure does make budgeting for more backup space easy.
 
Sounds like a rather strange backup strategy. I'd wager most businesses use something along the lines of incremental backups on site and monthly/weekly/whatever backups to other sides - both things where not that much data will be written to the same drive.

I use second copy... it cost 20$. it compares source and destination files and only copies / deletes changed files.
I don't know why anyone would erase everything and rewrite it.
 
Back
Top