When do you think we will see Hard drive manufacturers move to 5.25" form factor?

[cbn]

With NAND prices dropping and controllers getting better the price per GB gap between NAND and spinning platters (while still fairly large) is decreasing.

Also I have to wonder how much areal density will improve. If it doesn't improve quick enough maybe it is better to move to bigger platters rather than increasing platter count and using Helium to reduce friction.

Personally, I would prefer to a single platter 2TB drive in 5.25" form factor to one 3.5" with two 1TB platters if it meant reduced cost and power with improved performance.

P.S. One downside to 5.25" platters (vs. 3.5") might be decreased storage density (re: larger platter in a square box will probably have more empty space in the corners when compared to a greater number of smaller (3.5" or 2.5") platter drives of equivalent capacity.) With that mentioned, I'll bet this is fixable at the cost of some of other metric.

 

[dbcooper1]

30 - 35 years ago? Remember Quantum Bigfoot? 5.25" full and half height? Not sure that will return; there were manufacturing and economic reasons the trend went the other way.

 

[cbn]

Not sure that will return; there were manufacturing and economic reasons the trend went the other way.

It makes sense to me that back in day HDDs would downsize. Areal density was strongly improving and eventually capacity (even in smaller form factors) became enough for most people. And whatever competition these HDDs did face was obviously weaker.

But now, the progress in HDDs appears to be stalling. Consumer HDD pricing per GB hasn't been dropping very much as of late. And for the data center, it is taking high platter count and helium to hit 10TB (This due to the lack of progress in spinning platter area density).

P.S. Another thing I am wondering about is URE ( unrecoverable read error rate) and how much increasing areal density in the future poses challenge to keeping that in check. If so, maybe the larger 5.25" platters would be worth it (even if the areal density was a bit lower than what a smaller more expensive platter would have).

 

[ClockHound]

You guys are making me nostalgic for my old 50lb (that's what it felt like) CDC 5.25" Full Height 40MB beast.

When it spun up a small magnetic storm and large acoustic field surrounded the case. When it seeked, so did the desk. When it spun down, felt like something important had happened and the world could finally get some rest. If it had to taken out or system moved - had to engage the mechanical head lock or suffer the platter damage. After re-installing, disengaging the head lock drastically improved seek times.

Can get a refurb for only $314 (super cheap!!!). Won't mention the cost per MB back then. There was no cost per GB, because the Gigabyte hadn't been invented yet. :biggrin:

 

[Auric]

Putting aside the cost of new factories, the base cost regardless of capacity would presumably double due to increased robustness and materials. Larger also implies lower RPM and thus lower random access performance -yet increased power and cooling demands along with the expense of an incompatible form-factor. So it would be difficult to justify over simply adding more 3.5" HDD even if that were the only option. However, archival/cold/long-term storage is better served by ODD.

http://arstechnica.com/information-...lu-ray-discs-are-perfect-for-the-data-center/

https://en.wikipedia.org/wiki/Archival_Disc

 

[cbn]

Larger also implies lower RPM and thus lower random access performance

A larger diameter platter spinning at lower rpm still covers more area per revolution though.

Therefore if areal density was the same, random access time could decrease despite lower rpms.

-yet increased power and cooling demands along with the expense of an incompatible form-factor.

Power consumption would depend on how fast the platter spins and how many of them there were.

We definitely know 3.5" is becoming a problem because WD is having to resort to helium to keep friction and power consumption in check.

Also 5.25" is not an incompatible form factor. Many desktops can accommodate this (and there is always future external enclosures for Thunderbolt III to think about as well. These can be made for any size drive).

 

[Cerb]

I suspect that if we're going to see change, it's going to be 2.5" only, not the other direction.

MLC SSDs will have to come down by a factor of 5 or better to start competing w/ HDDs at all, 8-10 to wipe them out, and they will need to increase in capacity by similar factors at the same time.

We shall see, but by the time 2TB SSDs becomes affordable, we should be seeing 4TB+ 2.5" drives, and 10TB+ 3.5".

I'd love to see both SSDs killing off HDDs, and "raw" flash devices coming out (so instead of worrying about performance consistency of the controller, and it's likelihood of black box corruption, we could just have CoW/log filesystems for flash). But, I'm skeptical of the former, based on history.

Now, I do think that with 3D NAND, if the scaling promises hold, that we can and will see HDDs die off in terms of common OS drives for PCs, as time progresses.

 

[cbn]

I suspect that if we're going to see change, it's going to be 2.5" only, not the other direction.

For a data center, I have to wonder how much of a move to 2.5" (from 3.5") is due to power reduction or increased density at the cost of reduced performance and higher price per GB.

If those are the reasons, that needs to be addressed for the 5.25" form factor as well.

 

[Auric]

Also 5.25" is not an incompatible form factor. Many desktops can accommodate this (and there is always future external enclosures for Thunderbolt III to think about as well. These can be made for any size drive).

Was referring to data center racks and such as.

 

[IndyColtsFan]

I suspect that if we're going to see change, it's going to be 2.5" only, not the other direction.

MLC SSDs will have to come down by a factor of 5 or better to start competing w/ HDDs at all, 8-10 to wipe them out, and they will need to increase in capacity by similar factors at the same time.

We shall see, but by the time 2TB SSDs becomes affordable, we should be seeing 4TB+ 2.5" drives, and 10TB+ 3.5".

I'd love to see both SSDs killing off HDDs, and "raw" flash devices coming out (so instead of worrying about performance consistency of the controller, and it's likelihood of black box corruption, we could just have CoW/log filesystems for flash). But, I'm skeptical of the former, based on history.

Now, I do think that with 3D NAND, if the scaling promises hold, that we can and will see HDDs die off in terms of common OS drives for PCs, as time progresses.

Surprisingly, I haven't seen it posted here yet, but Samsung announced a 16 TB enterprise SSD in a 2.5" form factor last week:

http://petapixel.com/2015/08/15/samsung-16tb-ssd-is-the-worlds-largest-hard-drive/

http://www.pcworld.com/article/2971...argest-storage-drive-crown-with-16tb-ssd.html


Yeah, it will be insanely expensive and gears toward the enterprise, but it does give us a hint of the capacities that will be coming soon.

 

[Puffnstuff]

Seems to me that the trend is towards 2.5" form factor and it wouldn't surprise me to see desktops move to it as ssd's become the prevalent storage medium.

 

[PliotronX]

A self-contained dual-stack platter or 6xSSD unit in RAID would be interesting in a 5.25" bay.

 

[RecoveryForce]

I think, if anything, we will see traditional HDD manufacturers go more with SSHD drives at higher capacities in the 2.5" form factor. With enterprise drives, it might just make more sense to stick with the 3.5" sizes. Most enterprise drives already have smaller platters (a slight bit larger than the 2.5" platters) inside the casing. But, with the heat that those drives generate, I think it is more about heat distribution than anything.

 

[KingFatty]

We definitely know 3.5" is becoming a problem because WD is having to resort to helium to keep friction and power consumption in check.

Surely it can't be all that bad? Did they use helium because it provides better performance, but they could have used regular air and just had slightly lower performance? I dunno, it would be nice to see some support before I can agree with the conclusion that helium is a last-resort sort of tactic, instead of another incremental improvement.

I would guess that 5.25" was used in the past because tech was just too big to fit in smaller form factors, so there was no choice but to put up with its greater expense. With today's tech, there simply is no reason or justification to use 5.25" that outweighs the various negatives.

 

[rsutoratosu]

A self-contained dual-stack platter or 6xSSD unit in RAID would be interesting in a 5.25" bay.

well considering samsung has 16tb ssd on 3.5, why would you want 6 2tb max on a 5.25 ?


http://petapixel.com/2015/08/15/samsung-16tb-ssd-is-the-worlds-largest-hard-drive/

 

[dave_the_nerd]

The reason data centers moved from 3.5" to 2.5" HDDs is because they offer higher storage density in a smaller volume or smaller amount of rackspace, and use less watts per TB in the bargain.

To match a 2U enclosure of 24x 2TB 2.5" HDDs pulling 2-4w each, you'd have to have 6x 5.25" HDDs, which would have to have >4 times the capacity, <4x the power draw, <4x the per unit cost, and >4x the IOPS, to be in any way a convincing argument.

And a 400+ IOPS HDD is... probably impossible. 5.25" or otherwise.

Also, your RAID rebuilds now take 4x as long! Congratulations!

A 5.25" HDD makes more sense on a consumer desktop, maybe. But IMO, with more and more consumers relying on flash storage and cloud services, any progress in spinning rust tech is going to be watered down versions of whatever companies like Western Digital are willing to sell companies like Backblaze.

 

[fuzzymath10]

The flexibility trade-offs might be considerable. The casing might need to be thicker and more robust and cover more surface area which all add cost. The motor needs more power to drive heavier platters. On top of that, if there is binning, you now have standard capacity 5.25" drives which aside from cost don't have a lot of appeal.

SSDs don't have a lot of these weaknesses associated with fixed costs. I think that's partially why the floor on hard drive prices is so high relative to capacity.

 

[Cerb]

For a data center, I have to wonder how much of a move to 2.5" (from 3.5") is due to power reduction or increased density at the cost of reduced performance and higher price per GB.

It's due to reduced power consumption for a given performance level, and overall improved performance per U.

Two 2.5" give a bit more sequential performance than one 3.5" (30-50%, unless things have changed a lot recently), and around double the random performance. With a case that was made primarily for 2.5" drives, more than twice the number of 2.5" drives can fit as 3.5" could have.

Since even simple backup jobs can sometimes get stuck with random seeking that's not cached, 3.5" make less and less sense all the time, outside of uses like NAS. But, SSDs are still way too expensive for many, that still need their TBs of capacity. Meanwhile, power use will be the same or lower, depending on comparison drives (IE, a 5400 RPM NAS drive wouldn't be compared to 2x10k or 2x15k 2.5" SAS, but instead a 10k or 15k 3.5" SAS drive).

 

[hal2kilo]

You guys are making me nostalgic for my old 50lb (that's what it felt like) CDC 5.25" Full Height 40MB beast.

When it spun up a small magnetic storm and large acoustic field surrounded the case. When it seeked, so did the desk. When it spun down, felt like something important had happened and the world could finally get some rest. If it had to taken out or system moved - had to engage the mechanical head lock or suffer the platter damage. After re-installing, disengaging the head lock drastically improved seek times.

Can get a refurb for only $314 (super cheap!!!). Won't mention the cost per MB back then. There was no cost per GB, because the Gigabyte hadn't been invented yet. :biggrin:

Nice door stop.

 

[cbn]

The flexibility trade-offs might be considerable. The casing might need to be thicker and more robust and cover more surface area which all add cost. The motor needs more power to drive heavier platters.

Regarding the motor required, I am thinking of two things right now for 5.25" vs. 3.5":

1. Inertia

2. Air resistance.

For a given amount of TB, I bet inertia would be the same....but air resistance would be less for a fewer number of larger diameter platters vs. a greater number of smaller diameter platters.

 

[Red Squirrel]

It could be interesting to see just how big of a drive they could make with that. 10TB fairly easily I would think. Downside is it would not work in typical 3.5" bays and it would require a bigger server to get a decent amount of bays to do a decent raid set.

I actually see a move towards 2.5" more than anything as they can manage to cram more space into a smaller area, whether it's spindle or SSD tech.

 

[fuzzymath10]

Regarding the motor required, I am thinking of two things right now for 5.25" vs. 3.5":

1. Inertia

2. Air resistance.

For a given amount of TB, I bet inertia would be the same....but air resistance would be less for a fewer number of larger diameter platters vs. a greater number of smaller diameter platters.

I'm not an engineer, but I also drew high level conclusions from the fact that the current and voltage requirements are greater for a 3.5" drive than a 2.5" drive and so would follow for a 5.25" drive vs a 3.5" drive. Then there's vibration/noise.

I'd say the main attraction is more storage for the price rather than a lower price for the same amount of storage, but it's still an edge case. Most people probably have moderate storage needs and a preference for a compact/durable form factor at an affordable (!= lowest) price. Many systems today have only one 5.25" bay, if any.

Even 20 years ago, capacity was still limited even as density rapidly increased (CDs, high res images, digital video), yet we did not see a return to 5.25" drives.

The 3.5" and 2.5" form factors are now mature from a manufacturing standpoint and while you can easily run 2.5" drives in 3.5"/5.25" bays (especially with SATA vs PATA since the connectors are the same), you cannot force 5.25" drives into a smaller space.

 

[MongGrel]

When do you think we will see Hard drive manufacturers move to 5.25" form factor?

Never.

 

[cbn]

I'd say the main attraction is more storage for the price rather than a lower price for the same amount of storage

This is one reason I like 3.5" over 2.5" for desktop. (A 4 TB 3.5" HDD costs about the same as a 2TB 2.5" HDD).

With 5.25", I would hope the same trend continues. (An 8TB 5.25" HDD would cost the same as a 4TB 3.5" HDD or 2TB 2.5" HDD)

P.S. That calculation, of course, is based on scaling of 2:1 for platter area (5.25" vs 3.5") but it could be the storage increase is 2.5 to 1 or greater. (Area of 5.25" platter - inner part occupied by spindle / Area of 3.5" platter - inner part occupied by spindle)

 

[cbn]

I think, if anything, we will see traditional HDD manufacturers go more with SSHD drives at higher capacities in the 2.5" form factor.

Yes, I think the SSHD is something to look out for.

Then eventually SSD will take over the 2.5" SSHDs.

Then at some point before this happens hard drive manufacturers will sense the stress on their the 3.5" drive from a value standpoint.

But while this is happening the need for archival should be increasing.

So they will need to make decisions on how to approach this need.

3.5" HDD
3.5" SSHD

vs.

5.25" HDD
5.25" SSHD

^^^ These devices mixed with various other advanced techs.

P.S. Besides SSHD, I have to wonder how well an archival oriented HDD would work with a SSD cache for mixed duty (active use/archival) vs. other options. (unRAID implemented this with version 6.0)