Slashdot Mirror
With HDDs On The Ropes, Samsung Predicts SSD Price Collisions As NVMe Takes Over (tomshardware.com)
At its Global SSD Summit, Samsung shared its vision of the current state of SSD market and also outlined the future trends. The company noted that SSDs are steadily displacing HDDs in more applications, but NVMe is shaping up to be the dark horse that may put the venerable HDD to rest. From an article on Tom's Hardware: Samsung loves Google, and not just because it probably buys plenty of its SSDs. Samsung outlined its rather intense focus on Google Analytics for marketing purposes last year, and this year it pointed out that recent Google searches for "SSD upgrades" outweighed searches for "CPU upgrades." The historical trend indicates that this wasn't always the case (of course), but with 40 million searches for SSD upgrades this year, it is clear that SSDs are on the move. Performance stagnation in the CPU market is probably to blame here, as well, and we routinely advise readers to spend their hard-earned dollars on GPU and SSD upgrades before the CPU. The cellphone industry has long served as the prime example of an explosive growth market; it grew 19.1% in the last five years alone. SSDs, by contrast, grew 54%, and the steady downward pricing slope is a key factor. The all-important price-per-GB fell from $1.17 in 2012 to a mere $0.36 in 2016 (69% reduction). This is an average value, you can find SSDs for even less on the retail market. The SSD market grew 6x (to 130,000,000) from 2012 to 2016. Samsung's NAND shipments benefit from both the smartphone and SSD industries, and the company presented a chart that highlighted the changing NAND shipment mix. A higher percentage of flash heads into the SSD and Mobile segments every year as the percentage of UFD (USB Flash Drive), cards, and "others" decline.
after the floods of 2011 in Thailand, the HDD market raised prices, consolidated companies to fix prices higher, and has been milking them ever since. Some HDD prices per GB today are almost as low as they were before the rains in 2011...
Yes, but that is what the XPoint technology is trying to address. The NVMe technology is not designed to operate like ram and the latencies are still very high. Nominal NVMe latency for a random access is 15-30uS. The performance (1.5-3.0 GBytes/sec for normal and 5 GBytes/sec+ for high-end NVMe devices, reading) comes from the multi-queue design allowing many requests to be queued at the same time.
Very few workloads would be able to attain the required request concurrency to actually max-out a NVMe device. You have to have something like 64-128 random requests outstanding to max-out the bandwidth (fewer for sequential). Server-side services have no problem doing this, but very few consumer apps can take full advantage of it.
The NVMe design is thus more akin to being a fast storage controller and should not be considered similar to a dynamic ram controller in terms of performance capability.
Because of the request concurrency required to actually attain the high read capability of a NVMe device, people shouldn't throw away their SATA SSDs just yet. Most SATA SSDs will actually have higher write bandwidth than low-end NVMe devices (particularly small form factor NVMe devices). And for a lot of (particularly consumer) workloads, the NVMe SSD will not be a whole lot faster.
That said, I really love NVMe, particularly when configured as swap and/or a swap-based disk cache. And I love it even more as a primary filesystem. It's so fast that I've had to redesign numerous code paths in DragonFlyBSD to be able to take full advantage of it. For example, the buffer cache and VM page queue (pageout demon) code was never designed for a data read rate of 5 GBytes/sec. Think about what 5+ GBytes/sec of new file-backed VM pages being instantiated per second does to normal VM page queue algorithms which normally only keep a few hundred megabytes of completely free pages in PG_FREE. The pageout demon couldn't recycle pages fast enough to keep up!
Its a nice problem to have :-)
-Matt