TR Forums

No reviews anywhere as far as I could find on Google. It's available here. What are the chances that Scott's gonna have this thing shipped to him ASAP and write the world's first helium filled HDD review?

I'd love to know myself... or the 5TB that's inside the LaCie drives? (which I think is a Toshiba?)

http://www.storagereview.com/hgst_ultra ... ive_review

Rejoice!

Whoa... I didn't realize storagereview.com had started posting new reviews again. Looks like they've been active again since sometime in 2010?!??

Yeah they have been active again for a while. that was one of the first forums I signed up for there...and it ended up leading me to TR

Yep, they've got one of the only reviews on it that I've seen. Sadly the review types they do (they seem to have a few different benchmark setups) are a little confusing at times as to why they choose which benchmarks.

IIRC Tweaktown has the other review I've seen, don't know if anyone else has reviewed it...
http://www.tweaktown.com/reviews/6211/h ... index.html

So no surprises on the performance front. As expected, the price premium means it is a niche product aimed at use cases where storage density and/or capacity per watt matter above all else.

The other questions and implications raised by the tech are interesting though:

- I wonder if sealing the drive improves reliability? Although traditional designs do have a filter on the breather hole, gasses (e.g. water vapor) can still get through the filter, and this design eliminates that potential route for the entry of contaminants.

- I wonder if relying on helium reduces reliability? It is very good at leaking out of things due to its very small molecular size; can they really keep it contained for 5 years or more?

- From the pics in the Tweaktown review it looks like the lid is welded on (to keep the helium in, no doubt); if this is really true I imagine this has some back-end implications. If any internal components are bad the drive is probably not repairable even by HGST, and data may not be recoverable by any means even if the platters are intact.

- 7 platters in standard 3.5" form factor is insane! But it seems like IBM/HGST has always been willing to push platter count to the limit, using extra platters to get capacity while their competitors pushed the limits of bit density. They should be able to easily make an 8TB drive with this tech by increasing the bit density of the existing 7 platter design.

- For me, the most interesting thing will probably be the price drops on existing 4TB drives, now that a 6TB drive is becoming available in volume!

just brew it! wrote:

- For me, the most interesting thing will probably be the price drops on existing 4TB drives, now that a 6TB drive is becoming available in volume!

Some good points you bring up in your post! I don't see it affecting consumer 4TB prices much because as ya said the price is so high regular consumers don't have the option of buying these... but I do see it as the first stepping stone. Can't wait for Seagate's announced 6TB drive to come out, was expected to arrive this quarter. And to think, Seagate expects to have 20TB desktop drives shipping by 2020 based off the same HAMR tech...

Well, Seagate's 6TB is here....

http://www.seagate.com/internal-hard-dr ... ty-3-5-hdd

Figures that I'd post about it and a day later it shows up! It uses HAMR tech to achieve the 1Tb per inch areal density and they expect it to scale up to 20TB drives at minimum. Have to say I'm surprised it's only 10 cents per GB, kinda cheap for an "enterprise" class drive, but well within consumer's reach.

Here's an idea:

Torture the Seagate and Hitachi 6TB drives 24/7 and see how long they last. Should give everyone an idea whether helium sealed drives are more reliable or not. Of course, considering that these are enterprise drives, it might be a long time before we see any of them dying. But then again, maybe Geoff can devise some brutal workload that will make short work of them within a year.

The enterprise drive should be able to withstand continuous pounding for years. Furthermore, one of the concerns is potential helium leakage, which should be completely independent of workload.

Hmm... if you do lose the helium, does the drive die? Or does performance and power usage simply degrade?

just brew it! wrote:

Hmm... if you do lose the helium, does the drive die? Or does performance and power usage simply degrade?

Given the helium was there to stabilize the bit writes as well as the thermals, I'm guessing it would start throwing lots of read & write errors once the helium was gone. If the thermal issue turns out to be the case then there might be a new use for that freezer trick, chill the drive long enough to pull data off it.

I don't see how helium would boost the performance, especially as the Seagate drive outperforms it by a fair amount. As for power, if the drive could function without helium then why go through all the trouble and expense to put and keep helium in there for a couple watts at most in savings.

Kougar wrote:

As for power, if the drive could function without helium then why go through all the trouble and expense to put and keep helium in there for a couple watts at most in savings.

2W * the thousands of enterprise drives in a data center and we're talking real money both in electricity and reduced cooling costs.

Captain Ned wrote:

Kougar wrote:

As for power, if the drive could function without helium then why go through all the trouble and expense to put and keep helium in there for a couple watts at most in savings.

2W * the thousands of enterprise drives in a data center and we're talking real money both in electricity and reduced cooling costs.

On the face of it, sure. But DC's don't mass-upgrade drives all at once, nor do all the types of servers need 6TB drives in the first place, and lastly many don't even fit 3.5" models anyway. So I don't agree with that argument. The choice to go with helium would have to contribute far more in terms of reliability to the point that it's critical for the drive to function normally.