This should solve our fast storage problems...

.... in a couple of years but it will likely be very expensive, at least at first....
https://www.dpreview.com/news/375252360 ... -ssd-drive

Wow!

E.J. Peiker wrote:.... in a couple of years but it will likely be very expensive, at least at first....
https://www.dpreview.com/news/375252360 ... -ssd-drive

Expensive indeed.  Can't wait to see this in the mainstream. 

When this thing comes down to $2K I'd likely get 3 of them and not have to worry about storage again for years. As it is now, every year or two I am buying the latest and greatest high density HD's - this would skip around 3 to 4 generations of HDs.

Toshiba announces a 100TB SSD:
http://www.theregister.co.uk/2016/08/10 ... tb_qlc_ssd

Considering that current top-of-line 2TB consumer SSD's (the Samsung 850 Pro) are still going for $1150 CDN, I'm not sure that I see a 60TB or 100TB SSD at a $2K price point for a long time. Not without some combination of revolutionary factors both on the production side and on the market side. I.e. a terminal upheaval of the spinning platter HD market.

Given that Seagate is behind the 60TB SSD, they'd have to be willing to utterly destroy their own HD business to make this happen. Most mature companies have shown, time and time again, a very strong reluctance to engage in that kind of self-predatory innovation. Partly because they become hostage to their own success in their well-established client base, and also because even the more innovative firms know that one false move in a major discontinuous innovation can spell the immediate end of a major line of business, possibly even the entire company.

This is why it normally takes somebody from outside the dominant market players to do it. In this case I don't know that I see the combination of technical and marketing forces required permitting a small, scrappy player to upset everyone else's apple carts. Samsung and Toshiba aren't as dominant in the HD market. They're both big enough to pull it off, and perhaps would have the willingness to sacrifice their HD businesses to do it. But it would be a major undertaking even for them.

By the time the price comes down enough to go mainstream, 60TB or even 100TB won't seem as large as it does now. I'm already running a 32TB 8-bay RAID box (times 3 for local backup and off-site backup), populated with 4TB WD Red drives. Total cost per RAID box was around $3000 CDN. Net storage is roughly 25 TB, since my RAID config reduces capacity by 1/8. So at a cost parity per net GB of storage, a 60TB SSD could be sold at $7200 CDN to approximately break even vs. my spinning platter config.

I don't expect to see a $7200 price point for an SSD like this, let alone $2K, any time soon. Building 60TB of of Samsung 850's would run over $30,000. Dense storage solutions (especially ones targeted first at an enterprise market, rather than a consumer market) often command a price premium for some period of time after launch. I wouldn't be surprised to see the Seagate or Toshiba units with a launch price tag of more than $30,000 USD.

Backblaze recently posted a back-of-the-envelope assessment of their interest in the 60TB Seagate SSD, and their reception to it was somewhat lukewarm to say the least. They based their assessment on a predicted launch price of $20,000 USD, which I feel is quite optimistic. Based on $20K per SSD, their assessment is that 8 storage pods of Seagate 8TB hard drives would cost them $160,000 to deploy, while one storage pod of Seagate 60TB SSD's providing roughly the same raw capacity would cost $1.2M. Not a great cost-benefit argument there, and of course a higher price for the SSD would only make it worse.

If a massive storage consumer like Backblaze doesn't buy the concept, I feel it's going to take a big chunk of time before the technology is relevant to purchasers like us.
https://www.backblaze.com/blog/seagate- ... -pod-next/

P.S. Samsung is actually shipping a 16TB SSD, and has been for awhile. At the time of its announced shipping in March of this year, no official price was given but various speculations put it between $5000 - $8000 USD. The only actual price I've seen quoted is from CDW, and it's $10,300. Not exactly a bargain yet...
https://www.backblaze.com/blog/a-16tb-ssd-omg/

Applying Moore's Law, if today the cost is $500/TB (16TB for $8K US from Samsung), then today the price for the 60TB would be $30K in 2018 it would be $15K, in 2020 it would be $7.5K, in 2022 it would be $3.75K and finally in 2024 it would be under $2K....

So 8 years from now we can afford to buy a large SSD that would be price comparable with large spinning platter HD's today. The more interesting question to me is where do the two price-performance curves begin to get close enough to each other to see a seismic shift in the storage technology base?

Right now and for the immediate future, it's still heavily dominated by spinning platter technology (perhaps augmented by some on-board SSD in the case of hybrid drives). Has anybody done something as basic as plot the curves of price-per-GB of HD's vs. SSD's over a time axis, to see if they are converging or not, to date? Anecdotally, it seems to me that SSD's have been in a holding pattern of around 10X the cost per GB for quite some time. These new high capacity SSD's have not changed that equation yet.

Affordability is determined in a couple of ways: absolute, and relative. Even if SSD's come down in price, as long as HD's continue to be viable and increasing along their own capacity per $ curve, they will remain more affordable for mass storage needs. A few people are more like you & me, E.J., in evaluating things more on their absolute affordability. E.g. I have a $ figure in my head that, as soon as 2+ TB SSD's reach that point, I will start buying them, regardless of what's going on with HD pricing.

But mass market shifts typically happen based on relative affordability (or some other significant differentiation) compared to the other common alternatives. For average purchasers it won't matter how expensive or cheap SSD's are, as long as HD's are still cheaper for equivalent capacity and "good enough" performance. As long as the mass market shift still doesn't trigger, I think it's going to take longer than we want to see for SSD's to accelerate ahead in price-performance.

I did find this graph:
http://www.jcmit.com/disk2015.htm

It shows that spinning platter storage technology has a capacity per $ curve that has been decreasing at a long-term rate at least as aggressively as solid state technology including volatile memory as well as flash memory and SSD's. Of course this kind of retrospective look can't be guaranteed to project arbitrarily into the future. Different technologies face different relative and absolute limits on price-performance, and new technologies may enter the scene with their own performance enhancement curve. And of course, no price-performance curve can decrease indefinitely, it has to plateau out at some point at the bare minimum profitable cost of producing and supporting any product at all.

But it's interesting to note that, whatever data this chart is based on, solid state memory has not closed the price-per-capacity gap with hard drives. If anything, hard drives have maintained, and in some periods widened, their advantage steadily since the 1980's.

I'm poking around to see if there's anything more systematically documented on this...

I was just looking at this old thread again lately. I'm spec'ing out a new workstation build, and probably looking at a new laptop in the near ish future. Storage is a big consideration for both systems, so I've been poking around to see what the current state of affairs is with SSD prices and capacities. The short answer is, not much has changed in the past year.

My local system builder is still selling the popular Samsung 850 EVO (not even the Pro model or the newer 960 series) 2TB SSD for $1099 CDN, and the 4TB size for $2349 CDN. These prices have moved maybe $50 down since a year ago. Samsung is a bellwether player in the SSD storage market. You can find faster units, more expensive units, larger units, etc. But bang for buck, Samsung seems hard to beat.

The 16TB Samsung SSD model I referred to above in the old discussion (which is a SAS enterprise model, not a consumer SATA model) is still listed on CDW for just over $10,000 USD.

There has been some movement on SSD storage tech. Samsung now has the 960 series NVMe high performance SSD, for example. I don't believe those were out in any significant way last year. Still, a 960 PRO NVMe 2TB unit lists at my system builder for $1899 CDN. By comparison, a 10TB Seagate desktop hard drive can be had from them for $699 CDN. In terms of price vs. storage capacity, that's a difference of about 13.5X. (Setting aside performance, even though it is worth something.)

With little movement from Samsung on price-vs-capacity in a full year, especially on the affordability side of the equation, this tells me we're probably further out from truly affordable high capacity solid state storage than I may have thought last year.

Has anybody seen anything indicating a significant change that I'm missing?

$699 is expensive even with the exchange rate. I bout a 10GB Enterprise drive recently for $349 USD. Can't wait for large capacity SSD pricing to come down - it isn't following Moore's Law at the moment.