WD Wants to Save Rotating Magnetic Storage

[nord_musician]

oh really? thats what you really want? Lower the prices then!

 

[somebodyspecial]

Having worked for a computer repair company, I can say that hard drive failure rates are way too high. Spinning drives need to be ditched completely by 2020.

Agree on failures. I was dealing with 50 dead HD's a year on only supporting 700 PC's/laptops. And those people aren't too happy when you get to their desk either...ROFL. SSD's give you a far better way to avoid the issue with warnings long before they die so you can replace before then thus avoiding the pissed off user. SMART just doesn't seem to detect much before the problem usually for HD's. Most enterprise people could do fine with 120GB SSD's. I see these taking over enterprise soon for a large portion of the market. They are just far more reliable and easy to deal with in massive quantity.

 

[yankeeDDL]

Personally I am a big fan of HDD. Today, of all times, large storage is absolutely inevitable on a decent home PC.
With Cameras spitting out 16 (if not 32)GB of photos every holidays, 1080p videos, video games almost always above 10GB each, for me, having 6~8TB in RAID in a PC would seem a sensitive option. For that, today, there's still not an easy solution: desktop HD are around 2TB with some 3TB models, so 6TB in RAID means 4 disks and a decent controller.
So I would love to see single-drives with capacities up to 10TB costing below 150usd. Get a pair of them, put them next to a speedy 512GB SSD, and you've got yourself a reasonably future-proof system.

We're not there yet, and I don't believe that we'll see reasonably priced SSD with 2~4TB any time soon: $600/TB is amazing, by "yesterday's" standard, but waaay too expensive for most non-power users.

 

[danwat1234]

Prices aren't much better than just before pre-flooding prices, around $70 for a 2TB hard drive, something like $120 for 3TB.
Please make HAMR or bit patterning or shingled magnetic recording a reality sooner rather than later. Platter densities haven't increased for over 2 years now!

Yeah, SSDs certainly are closing the gap. Approaching $0.50/GB, an easy choice for a system drive on a laptop or desktop.

 

[danwat1234]

it is really simple, just make 10K rpm HDD mainstream. close the speed gap between HDD vs SSD when the adaption of SSD is slowed down, so are the price drop slowing down, then the snowball effect.

But then areal densities go down because (I think) head technology can't keep up with the faster spinning platters and the high areal densities that 7200RPM and 5400RPM drives have. So prices per GB go way up. They are still way slower than an SSD. Bleh. Good effort though.

 

[InvalidError]

But then areal densities go down because (I think) head technology can't keep up with the faster spinning platters and the high areal densities that 7200RPM and 5400RPM drives have. So prices per GB go way up. They are still way slower than an SSD. Bleh. Good effort though.

Electronics can keep up with faster-spinning platters without problems.

The problem is head positioning (mechanical) precision cannot. So you need lower densities to afford the extra slack in case the heads are slightly off-target while writing.

Once SSDs become cheap enough per GB to replace 7200RPM drives for OS+apps+games drive (say 500GB/$100), we will likely see HDD manufacturers focus more on higher-capacity lower-power 4500-5400RPM drives to keep their $/GB advantage for people and companies who archive tons of data to near-line storage.

 

[mynith]

SSD's are the fastest way to store and retrieve data in computers today, that much is true. However, much of that data isn't needed immediately, so prefetching can make a lot of those advantages irrelevant. Of course, the reduced latency means applications get loaded immediately, and a video starts playing immediately, but a hard drive is fast enough to keep it playing. SSD's are a brute force approach to make a computer faster, not the cheapest. They consume less power, but not that much less. Certainly not if you only need one HDD to have the same storage capacity of about four SSD's. I think it's important to not see flash as a replacement for spinning media, but rather as an add-on. Tiered storage is already being implemented in hardware, like Seagate's SSHD's and Intel's Smart Response Technology. There are software implementations too, such as Windows ReadyBoost, and on UNIX-based operating systems, the ZFS filesystem has it, and BTRFS has it planned in the not too distant future. Sure, there will be some cache misses, but does it really matter that much? I don't think not having that small hiccup once in a while is worth paying extra for a big-ass SSD. That money is better spent elsewhere. So WD is right, the fat lady has not sung.

 

[GreaseMonkey_62]

The future will probably be close to Seagate's hybrid drives with large flash caches and dense mechanical platters. SSD's and mechanical drives have their pros and cons.

 

[stevenrix]

Just before the flood, i was able to buy a few hard-drives of 2 TB for a merely $60 per drive, and 2 TB was almost the top of the line, supplanted by the 3TB hard-drives. After flood, these prices are still very high: 4 TB HDDs with 64 megs cost almost $400, and around $300 for a 3 TB. Comparatively, i would say that the price is inflated by at least 3 times since the pre-flood prices. So yes there are still a big margin to go down on these prices. If you get lucky you can find sometimes cheap hard-drives on promotion, i was able to get a 4TB last time at Fry's for around $127 (2 weeks later they sell it for $237).Since the flood, there hasn't been any innovation at all on the hard-drive market, we were supposed to see 6TB drives by last year, and it never happened.
Do what i've done, i never bought a hard-drive after the flood, and will only do it if prices become reasonable. Until then I will choose other alternatives such as the SSD, since I need lots of access time to run my virtual machines.

 

[actionjksn]

If they want to remain competitive in the future they will have to stay cheap, improve their dependability and get their speed up at least close to that of one of the slower current model SSD's. Because SSD's are gradually getting cheaper and bigger. I don't know if it's even possible to get much more read and write speed out of a disk. But if they don't figure it out then they will we in trouble.

When the day comes when we can buy something like a 320+ gig SSD for a $100.00 or less, that's when it will be over for the hard drive makers. As it is I had to change out my 320 gig drive in my laptop for a 750 WD Black because I filled up the 320G. So I would say 320 is probably the bare minimum for most people as far as needed storage space. Right now SSD cost and capacity is the only things keeping hard drives afloat. The hard drive makers really need a technological breakthrough within the next two or three years.

 

[actionjksn]

Instead of pooling their resources for their PR campaign, this group of hard drive makers need to pool their resources to figure out how to make their disks have a much faster read and write speeds. Everybody already knows a hard drive holds a lot of data and is cheap. We want a more speed campaign not a spin campaign.

 

[InvalidError]

Instead of pooling their resources for their PR campaign, this group of hard drive makers need to pool their resources to figure out how to make their disks have a much faster read and write speeds.

Why fight a battle they cannot win? Mechanical devices cannot match semiconductors on speeds and access times so pursuing "more speed" is a total waste of time - anyone who wants speed above anything else will end up using SSDs as they get larger, cheaper and faster.

Making HDDs faster requires faster rotation, higher-power actuators, much more precise head positioning, lighter parts, much tighter manufacturing tolerances, much stronger materials to handle increased mechanical stresses, a lot more power, etc. so mechanical storage has extremely limited potential for cost-effective increased transfer speeds. Access speeds on modern HDDs are only 30-50% faster than HDDs from 12 years ago while transfer speeds only have increased by about 10X despite going from 5400 to 7200RPM and a ~100X increase in aerial density.

Speed on mechanical storage simply does not scale. The performance of cost-effective modern HDDs is pretty much as good as it will ever get.

HDDs have a much better long-term outlook as slower, higher capacity, cheaper low-power storage media. I expect 4500RPM HDDs to come back in fashion 2-3 years from now with higher capacities per platter, lower power and hopefully improved reliability from reduced stresses thanks to affordable 300-500GB SSDs kicking HDDs out of the boot/application drive market.