The Silicon Motion SM2262EN SSD Controller Preview: Good And Bad Amplified
by Billy Tallis on August 1, 2018 11:45 AM ESTLooking Forward
It's great to see that Silicon Motion is continuing to improve on the greatest performance weakness of flash-based SSDs: random read latency. The SM2262 provided a surprising jump in random read performance, surpassing even MLC SSDs. The SM2262EN takes things a bit further, and Intel's Optane SSDs now lead by only a factor of 6 or so. This kind of improvement has the greatest potential to improve real-world performance as perceived by end users, and random reads are the workload where flash-based SSDs are furthest from saturating the SATA or PCIe host interfaces. (The Intel/Micron 64L 3D TLC NAND probably deserves some credit for enabling these improvements, too.)
The SM2262EN is well-tuned to turn in stellar scores on many other benchmarks, and it is near the top of the charts for almost all of our synthetic benchmarks. However, it is clear that Silicon Motion is making some serious tradeoffs to attain these scores. The SM2262EN's performance degrades severely when the drive is full, and this effect seems to be even stronger than for the SM2262 drives we've tested. Silicon Motion has developed the fastest SLC write cache currently available, but once it's full, the situation isn't pretty.
For a 2TB drive like our review sample, I'm not too worried about these performance hits: it's pretty hard in practice to fill a really large SLC cache, and this isn't the easiest drive to fill up to 100% with real data. Our toughest benchmarks are well beyond the range of normal usage patterns, so abysmal scores there don't necessarily mean the drive is bad or unsuitable for power users. What it means is that Silicon Motion needs to be careful not to go too far with their optimization efforts. With the lower capacity drives that will account for the bulk of SM2262EN sales, it will be much easier to bump against the limits of which scenarios the controller can handle well. It may be necessary to use a much higher overprovisioning ratio on the lower capacity models, in order to ensure that they can maintain high performance even under heavy workloads. Products using the SM2262EN will live or die based on how well it can avoid falling into low-performing pits when presented with real-world workloads.
Idle power management seems to be broken on our SM2262EN sample, which is a disappointment because the SM2262 drives currently on the market offer the best trouble-free power management situation out of all the NVMe drives we've tested. Silicon Motion needs to ensure that this bug doesn't make it into retail products.
Consumer SSD prices have been declining for most of the year, in large part spurred by the adoption of 64L 3D TLC by the many brands that do not manufacture their own flash. Last year saw 64L 3D NAND hit the market in a limited number of products from the biggest players in the market, but now that NAND is in use everywhere and in plentiful supply. The downward trend is likely to continue in the near future, so even though the SM2262EN may command some premium over the base SM2262, the product lines that adopt the updated controller will probably do so at or below the current prices.
Looking further forward, we don't have much information about the rest of Silicon Motion's roadmap. We know they're working on PCIe 4.0 support, but this won't be needed in the consumer market anytime soon. They have also developed a new generation of more robust LDPC error correction intended to better support QLC NAND, and this will likely show up in the next generation of controllers from Silicon Motion. A full refresh of Silicon Motion's NVMe controller lineup is probably still at least a year away, so the SM2262EN may need to stay competitive for quite a while. It will probably be paired with 96L 3D NAND as soon as it is available, which may provide further incremental increases to performance and power efficiency.
The SM2262EN may also end up paired with QLC NAND for a very fast and cheap bulk storage drive. Such a drive would definitely benefit from the fast and aggressive SLC caching strategy used by the SM2262EN.
Facebook
Twitter
RSS
28 Comments
View All Comments
Death666Angel - Thursday, August 2, 2018 - link
I don't read it that way, but okay. :) I don't have a definitve cost breakdown of an SSD. But my best guess is NAND is still the factor #1 and goes up with capacity. #2 would be the controller or the RAM, depending on size of the SSD, which usually correlates with the size of the RAM. But controllers can cost a few dollars or a few tens of dollars, so that is still a relevant number in pricing of an SSD. Samsung and WD price their drives that way. because they can, so far.FunBunny2 - Friday, August 3, 2018 - link
well, here's the problem. if you're an economist, then marginal cost is the driver of price in a competitive market. whether that's true for SSD/SSDparts is murky. for the accountant/MBA types, average cost drives price, regardless of market.now, the crunchy aspect of correlating cost to price is the production process. in the 19th century, labor was a significant component of cost and thus price. demand slackens, fire people to keep both costs more or less stable. demand increases, hire for the same effect.
in the 21st century, with SSD/SSDparts, there's virtually no labor in direct production, so marginal cost is near 0; ergo the econ types say to drop price to move more product. the accountant/MBA types recognize that most of average cost, while higher than marginal, is mostly amortization of R&D and capital equipment (all those new fabs AT has been reporting on, recall?). even they understand that the decision is the same as the econ, a very rare event: the only way to make money is to move more product and drive down average cost. but they can only do this is demand increases. and that can only happen if end-user product vendors can 1) more ways to use the parts, and 2) people have more money to buy the end-user product.
1) is largely a substitution exercise; i.e. a zero-sum game among end-user product vendors. there's no growth in aggregate demand for end-user product, thus none for SSD/SSDparts. nobody wins.
2) is a purely macro-economic phenomenon, and thus dependent on the 'middle class' having more moolah to spend on more bling. you can see where this is going? with right-wing governments driving income concentration, aggregate demand eventually collapses. this is exactly what created the Great Recession.
end-user product vendors can't directly move 2), all they can do is encourage their governments to spread the wealth so that aggregate demand can grow, and they can sell more product. on the whole, they haven't shown the smarts to see where their bread is buttered. as labor cost diminishes, just firing bodies gains you less and less until it gains you nothing. growth in highly capitalized production economies of the 21st century doesn't work as it did in the primitive 19th.
greggm2000 - Thursday, August 2, 2018 - link
What I'd really like to see are SSD tests done on an (user) encrypted drive. Would performance be equivalent to a fully filled drive? I imagine this would be a fairly common use case?Billy Tallis - Thursday, August 2, 2018 - link
Software encryption does technically leak information if it uses TRIM commands or otherwise signals to the drive what data is and isn't valid. It also imposes performance overhead from doing the encryption on the CPU . There aren't many reasons to justify using software full-drive encryption on a SSD when self-encrypting SSDs are so common (Samsung, Crucial MX, etc).Icehawk - Saturday, August 4, 2018 - link
Is Opal effected by this? What performance cost is there? We’ve got whopping laptop at woro with it enabled buy I’d like to push us in a more secure direction. Would probably help our PCI score too.Chaser - Sunday, August 5, 2018 - link
I wish someone would build a review site that includes SSDs that writes reviews based upon a an average PC gamer's performance perspective. I myself have tested the Evo 860, the 970 EVO, Optane 900, the XPG SX8200, and the Patriot Hellfire. Like many other revealing Youtube videos that compare these drives most often the Evo 860 is either faster at loading a game, the same or very slightly slower. While I understand that Anandtech has readers that are looking at higher usage scenarios, I'd venture to say MOST of their readers are in the former category.As it stands today with most similar sites we see chart after chart of benchmarks on multiple pages. We read about accolades on random and sequential performance. Some sites rank the drives from 1-10. But in the end, the user experience differences prove to be negligible for most users and a simple article like that probably would entice site visits to read through the hairsplitting benchmarks.
KAlmquist - Sunday, August 5, 2018 - link
I'll repeat something Billy Tallis stated in a comment and probably should incorporate into the text of the review: “I did run the Heavy and Light tests on this drive with it 80% full and the results were similar to the 100% full case.”When I partition an SSD, I've always left a bit of space unused in order to effectively increase the spare area to 20% or so. That improved performance consistency with older SSD designs. With the SM2262EN, it might still reduce write amplification, but not enough to substantially affect performance.
kensiko - Wednesday, January 9, 2019 - link
I'm hesitating between the AData XPG SX8200 (SM2262) and the pro one (SM2262EN), 50 CAD$ difference. Any opinion ?