SSDs Have Bleak Future, Says Researchers

[Energy96]

I would never, ever, go back to HDD. It's like driving Yugo after having a Bugatti Veyron.

 

[maestintaolius]

Extrapolating trends generally a dangerous thing to do and I'm guessing even more so when done in areas where breakthroughs can totally change the dynamics.

 

[slabbo]

PRICE, PRICE, PRICE, PRICE. This is all going to be about price. Even if bigger SSD drives tend to be slower, they can offset this disadvantage by it's price point. Look at SSD's vs HDD's. HDD's are still thriving because of the price, not the technology.

 

[nonoitall]

I wouldn't say SSDs have a bleak future; I'd say NAND has a bleak future.

 

[danwat1234]

[citation][nom]Tomfreak[/nom]Sometimes I wonder, desktop HDD are not limited to power consumption, but why arent the HDD manufacturer push 10K RPM HDD down to mainstream to "slow the SSD" adoption . Surely the latency gap would have improved significantly, 8.9ms vs 4.0 seek time. It may not reach the SSD ones, but it is still better to delay the SSD getting into mainstream. While many of us like the capacity of HDD, but we I would prefer having slighly smaller 500GB with boost speed over the 2TB ones. The size of the mainstream SSD are still less than 64GB, win7 64bit takes up almost half of it, so the leave not much room for games/other thing.[/citation]
I agree completely. Western Digital could make a 1.2TB Velociraptor soon, the platter density is there now.

Also in the mobile, sector, 10K hard drives would be great in gaming and professional laptops. I don't care if my 2.5" hard drive takes 4 watts at idle, 6 watts at full random I/O.

 

[danwat1234]

[citation][nom]dspider[/nom]By 2024 flash storage will be analogous to floppy disk magnetic storage.[/citation]

No, by 2024, the hard drive will be analogous to floppy disk magnetic storage. I think a 12TB SSD in my computer will be plenty

 

[Shin-san]

Even if SSDs end up hitting a barrier, I'll still get them for my OS drive because it made a large difference, unless something faster comes out. 256 gigs is quite a bit for my copy of Windows, and I install a lot of heavy software.

 

[osmosium]

St

 

[superjesus]

So Tom's is now offering the insightful 2024 technology predictions of a grad student as news? Must be a pretty slow news day.

 

[Supertrek32]

Computers have been "mainstream" for what, about 20 years? Yes, they were around earlier than that, but not mainstream.

And we're trying to predict the technological progress of the field 12 years in the future?

Anyone else see an issue here?

 

[marraco]

At 2024 and 6 nm, every component will be at his limit. Even transistors, GPU, APU and CPU.

 

[tomfreak]

[citation][nom]caedenv[/nom]@TomFREAKThe reason 10K drives do not do well in the consumer market is not due to the performance or price increase, it is due to the noise. In a world where your 7200RPM drive is the loudest part of most computers, a 10K drive just screams. If they managed to make a quiet 10K drive people would be all over them.[/citation]If u ask me there are still a number of things that are louder than HDD in term of noise, things like CPU fan/GPU fan these are easily more noisy. currently the price of Velociraptors are just ridiculous, They should be offering twice the capacity than what there now to be consider worth buying.

[citation][nom]Wamphryi[/nom]I have two Velicoraptors in RAID 0 and they cant compete with my Crucial M4 in terms of performance. The Raptors are very useful for video editing scratch drives and indeed there is more capacity on them but that is really the only advantage they have over the Crucial M4. I put a Crucial M4 in one laptop and a Vertex 2 in another laptop and the difference was night and day. A 128 GB is more than enough for applications and games on a well managed system and 60 GB is enough for a well managed laptop imo. .[/citation]U may claim 128GB is enough for most application and games, but IT ISNT A MAINSTREAM SSD. Mainstream SSD is much smaller

 

[IQ11110002]

An upcoming scientist saying can't or saying there will be upcoming problems, I think it's time for a new career before you get off to a bad start!
In the words of Oddball(Donald Sutherland) in Kelly's Heroes, Always with the negative waves, Have a little faith baby have a little faith.
They will find a workaround or a completely new technology by 2024, These university studies really piss me off, I might as well say the sun is going to die out in roughly 5 billion years OHHH NO what are we going to do.
Point is enjoy the present,For who knows what tomorrow will bring!

 

[kinggremlin]

[citation][nom]Tomfreak[/nom]Sometimes I wonder, desktop HDD are not limited to power consumption, but why arent the HDD manufacturer push 10K RPM HDD down to mainstream to "slow the SSD" adoption . Surely the latency gap would have improved significantly, 8.9ms vs 4.0 seek time. [/citation]

For the most part, the additional cost of designing and manufacturing the drives does not provide enough of a performance benefit to justify its release. It's not a noise thing as someone below mentioned. The current Velociraptor is one of the quietest drives you can buy. The only performance metric that improves with higher rotational speed is average latency. That's a measure of how long the read/write head has to wait for the proper location on the platter to rotate around to the head. For a 7200RPM drive, average latency is 4.2ms. For a 10kRPM drive 3.0ms. That's an improvement of only 1.2ms, or about 29%. But that's for latency, which is only half the equation for average access time which also includes seek time. Seek time is how long it takes the read head to find the proper track on the platter. This is not in any way affected by rotational speed. So, if you have a 7200rpm drive with an average seek time of 5.0 ms, if you increase the platter speed to 10krpm, the average seek will still be 5.0ms. If you add average latency to average seek time, you get an average access time of 8ms for 10k rpm, and 9.2 for 7200rpm. This means that average access time is only improved 13% moving from 7200 to 10k rpms. That would be a hard to preceive difference. As RPM's increase to 15k and beyond, you rapidly reach a point of diminishing returns where the decrease in latency is simply not worth the increased complexity in design and manufacturering.

So why do 10k/15k RPM drives advertise average access times that are so much lower than 7200RPM drives when latency isn't really improved significantly? Because higher RPM drives, for a variety of reasons, have to use smaller diameter platters. By using smaller diameter platters, the heads reduce the maximum distance they have to travel across the platters which reduces seek time. So when you add the lower latency to the lower seek time, you see a pretty decent improvement in average access time.

Here's the trick though. You can achieve the same smaller diameter platter effect with any 7200RPM drive by simply short stroking it which means partitioning off part of the drive and not using it in order to keep the heads within a smaller portion of the platter. You've then eliminated the seek advantage the 10k rpm drive had and likely ended up with more capacity even though you're not using the whole capacity of your drive.

With almost any current SSD, the average access time is, near as makes no difference, instantaneous when compared to any mechanical drive so the seat of pants improvement is dramatic. With the prices of smaller SSD's dropping so low recently, there are very few circumstances where a 10kRPM drive would make sense for a home user.

 

[alidan]

so... what? no one takeing into account a 3d process? which will probably be more viable for ssds up front than a processor.

depending on how long the 4d process takes, it should at least double the size of the storage without doubleing the cost.

you also have to take into account that even at reduced speeds, ssds will still out preform hdds... however a hdd sould be capable of almost 2tb read speed if the reading ability scales with size of the disc... if you eliminate seek time, such as load the while os onto ram, and figure crap out on the fly, a hdd could potentially be faster than ssd, because whats the point of seak time if its nothing vs 16ms if you only do it once? but i doubt that it will ever come to that, we will figure things out to compensate.

 

[diellur]

Isn't this the reason for R&D? Existing technologies aren't going to work as well in the future, so we develop newer variants to do what we need them to do. 2024? That's 12 years...look at the difference in PC technology over the last 12 years to see how things can change; that's what hasn't convinced me in that article. Right now, SSDs are still finding their feet in the mainstream...with the level of benefit they give, I highly doubt that they'll be based on the same technology in 12 years.

 

[alyoshka]

Tech chaps are certainly very unsure of what they send out into the markets.... is this it?

 

[nikojagochi]

Not a too smart oppinion, we can see why he's just a graduate student...
As densities increase someone will find a technology to keep the drives larger with a high performance. This happened with everything else in IT history.
And, finding out new technologies is what gives competitive advantages to companies as Intel, Apple.
There is someone out there already having ideas how to solve this and sell it to some company...

@article
"This is just dumb." Please find some doctorate students that have a job for some time in the IT industry for oppinions.

 

[Guest]

By 2024 I'll be sh*tting self-sentient datacrystals and driving an antimatter-catalyst fusion spaceship!

 

[catatafish]

This article was about as profound as "Windows is going to change their logo in Windows 8".

 

[Burodsx]

If you look at the past we come up with a new technology to replace the old...

 

[Guest]

SSD's aren't doomed, NAND processing is doomed with this current design. What's important is that SSD's w/ NAND were the breakthrough step away from magnetic hard drives, but we will never take a step back to magnetic...they will find a way. Besides, too much money is invested at this point, there is no going back now.

 

[fuzzion]

This graduate is as dumb as a brick. By 2024, holographic memory arrays will be used in domestic households.

 

[Guest]

In 2024 we will have 1 PB (1000 TB) drives.

 

[caskachan]

BUT wait, you mean to tell me that my new 500/200mb/s read/write drive is slow and new bigger drives will be slower? 120 and 250gb versions of these disks exist, wuts the problem exactly , i only see ssd getting cheaper and faster did i miss anything?