Many factors in play here....

1. PC purchases are way down, folks keeping their machines longer due to a) lack of need to upgrade, b) small incremental performance increase generation to generation and c) cost of GFX cards

2. SSDs are sexy. Yes SSDs give great benchmarks but what do you call the IT Manager at the Fortiune 500 company who submits a proposal to switch all the comapny'ds PCs to SSD storage ? Easy answer ... unemployed. While extremely beneficial in rare niche usages, SSDs have no impact on personal productivity. No legal secretary every typed an extra legal brief. No CAD operator ever laid out an extra room in a building and no gamer ever reached a further save point because they had an SSD. It only can be financially justifies on rendering boxes, video editing and other small segment markets.

3. It really doesn't matter than I can backup 2 TB of files much faster from SSD to SSD than from SSHD to SSHD ... that happens at 3 am while Im sleeping. It really doesn't matter whether a MS Word benchmarkwith 1,134 actions can run much faster on an SSD cause a use has to type at least 1,134 keystrokes in between each one.

4. If you can switch the boot device in a box between SSD, SSHD and HD over 6 weeks and no one notices, you aren't benefitting from having one. In years gone by, we saw enthusiasts using RAID to get their most out of there machines but from a productivity or real life performance standpoint, neither has had an effect. Oh we could prove how much faster it was using benchmarks ... problem was , day to day, we never did any tasks representative of those benchmarks outside very small niche markets.

5. Over the last 7 years, we have had 3 storage device (SSDs + SSHDs) failures, all SSDs. Over the last 5 years we have had no failures. The baseline set up has (1) SSD (Samsung Pro 250 GB) and (1) SSHD (2 TB or more).

6. Reliability of SSDs is well past the shaky initial years, but costs for multi-TB systems is just not competitive yet. But regardless of what you are doing on the PC, again outside those niche markets, productivity is not being impacted by the storage subsystem. I expect that costs will be not worth thinking about within 5 years but I don't see the HD going the way of the floppy disk and zip drive before 2025.

John Naylorsnip

I guess you've never worked for a large company, because at mine every new PC that is bought comes with an SSD. The cost of a 128GB SSD is comparable to a 500GB HDD when you're buying from an OEM like Dell, and for our IT guys the M.2 form factor is a massive improvement over SATA when it comes to maintenance (no cables, only 1 screw). Plus for people in the standard office-type roles, the smaller capacity of the SDD discourages them from installing crap, so it's a win-win.

As for "no impact on personal productivity"... you forget that giving people nice things makes them appreciate working for you. And yes, there is actually a productivity gain, because the old excuse of "oh I'm making coffee while waiting for my computer to boot" is no longer valid.

SSDs are also pretty much invaluable for tasks that touch a lot of files, such as software development. I have an SSD and HDD in my work PC - the HDD is used for backup purposes, but sometimes I run builds off there, and the difference to the SSD is night and day - it's almost impossible to imagine we used to spend our lives waiting for HDDs.

Xx Tek Tip xXIdiots don't care about hdds then I'll take my money to seagate and Toshiba it's their loss

What do you mean “idiots don’t care about hdd’s“? You make it sound so personal and juvenile they did this.

It’s a BUSINESS decision based on not selling quite enough to warrant 3 factories. They still have 2 more; HDD’s are still a big part of WD’s business and will be for years to come. As long as there is a demand they will build them.

rtwjunkieWhat do you mean “idiots don’t care about hdd’s“? You make it sound so personal and juvenile they did this.

It’s a BUSINESS decision based on not selling quite enough to warrant 3 factories. They still have 2 more; HDD’s are still a big part of WD’s business and will be for years to come. As long as there is a demand they will build them.

They are definitely doing the wrong thing - as everyone needs higher capacity drives the business will move there - i don't see no 3-10tb ssds and when I do they cost more than a kidney and don't sell.

Xx Tek Tip xXThey are definitely doing the wrong thing - as everyone needs higher capacity drives the business will move there - i don't see no 3-10tb ssds and when I do they cost more than a kidney and don't sell.

This should tell you everything you need to know:


You can see the clear wipeout of supply from the floods in Q4 2011, and then the sustained drop (to the tune of 30%) from 2014 to 2016. The market has spoken: only datahoarders, enterprise and cloud will accept HDD-based storage, and only because it's cheap. The moment NAND (of any sort, really.. QLC is fine...) reaches 1.5-2x the per bit price of HDDs (currently it's around 4x), expect the HDD industry to completely crash as demand evaporates basically overnight.

Of the 3 remaining, WD and Toshiba are the ones most likely to actually survive, both of em being fab owners. Meanwhile, Seagate doesn't even have an in-house NAND controller (they bought LSI.. then sold them off....), let alone a NAND fab, and I really doubt that Micron is interested in being acquired by Seagate, let alone acquiring Seagate.

It's kind of a shame that Optane (or simply caching) didn't pick up much even for business. It's invaluable to me. I run a small SSD for the main drive, and a 10TB Seagate with caching. A lot of games get put on the Seagate and while it may not be an SSD, it's a helluva lot better than regular HDD speeds. You get the best of both worlds here. And save some cash. No way in hell am I going to fork out cash for huge SSDs (not that they make them at 10 TB anyways.. or the ones that are 10 TB are rare and for servers or something).

StrayKATIt's kind of a shame that Optane (or simply caching) didn't pick up much even for business. It's invaluable to me. I run a small SSD for the main drive, and a 10TB Seagate with caching. A lot of games get put on the Seagate and while it may not be an SSD, it's a helluva lot better than regular HDD speeds. You get the best of both worlds here. And save some cash. No way in hell am I going to fork out cash for huge SSDs (not that they make them at 10 TB anyways.. or the ones that are 10 TB are rare and for servers or something).

Give it 1-2 years and it'll be much more buyable

Optane is great, but the prices are crazy callous for the moment....

ZeDestructorThis should tell you everything you need to know:


You can see the clear wipeout of supply from the floods in Q4 2011, and then the sustained drop (to the tune of 30%) from 2014 to 2016. The market has spoken: only datahoarders, enterprise and cloud will accept HDD-based storage, and only because it's cheap. The moment NAND (of any sort, really.. QLC is fine...) reaches 1.5-2x the per bit price of HDDs (currently it's around 4x), expect the HDD industry to completely crash as demand evaporates basically overnight.

Of the 3 remaining, WD and Toshiba are the ones most likely to actually survive, both of em being fab owners. Meanwhile, Seagate doesn't even have an in-house NAND controller (they bought LSI.. then sold them off....), let alone a NAND fab, and I really doubt that Micron is interested in being acquired by Seagate, let alone acquiring Seagate.

We'll see how it unfolds - hopefully those hdd factories can meet the supply and demand requirement or competitors will smash through their units shipped and dominate the hdd market.

StrayKATIt's kind of a shame that Optane (or simply caching) didn't pick up much even for business. It's invaluable to me. I run a small SSD for the main drive, and a 10TB Seagate with caching. A lot of games get put on the Seagate and while it may not be an SSD, it's a helluva lot better than regular HDD speeds. You get the best of both worlds here. And save some cash. No way in hell am I going to fork out cash for huge SSDs (not that they make them at 10 TB anyways.. or the ones that are 10 TB are rare and for servers or something).

Caching will catch on a bit thanks to Ryzen and StoreMI. It's a really good start, but they've got to increase the dedicated DRAM amount further in the next motherboard generation and one's after it. It's far too low and DRAM is abundant and even now it's cheap relative to the performance it provides. What I said about the deidcated DRAM amount that can be allocated is even more true with Threadripper/Epyc as well because of the memory channel bandwidth being 2 to 4 times higher thus way more bandwidth advantage to capitalize upon. It's the same scenario with APU's one based on Epyc platform's memory channel bandwidth would really kick the pants off a AM4 one.

InVasManiCaching will catch on a bit thanks to Ryzen and StoreMI. It's a really good start, but they've got to increase the dedicated DRAM amount further in the next motherboard generation and one's after it. It's far too low and DRAM is abundant and even now it's cheap relative to the performance it provides. What I said about the deidcated DRAM amount that can be allocated is even more true with Threadripper/Epyc as well because of the memory channel bandwidth being 2 to 4 times higher thus way more bandwidth advantage to capitalize upon. It's the same scenario with APU's one based on Epyc platform's memory channel bandwidth would really kick the pants off a AM4 one.

I didn't even think about RAM, but you're right.

And if I had the cash, even a RAM drive seems realistic to me now. My board can hold 128 gigs.. but I'd never use that all of that as just RAM.