The Intel Optane SSD 900p 480GB Review: Diving Deeper Into 3D XPoint
by Billy Tallis on December 15, 2017 12:15 PM ESTConclusion
As expected, the 480GB Optane SSD 900p performs about the same as the 280GB model. That makes it one of the overall fastest SSDs money can buy, but the Optane SSDs don't win in every test.
Higher performance is often an important selling point for higher capacity SSDs—and sub-par performance can be a major reason to avoid the smallest model in most product lines. This doesn't really apply to the Optane SSDs, so consumers are faced with the simpler question of how much fast storage they really want to pay for. As the most expensive "consumer" SSDs on a per-GB basis, the Optane SSDs force potential buyers to consider just how much blazing fast storage they actually need. I'm currently using an Optane SSD in one of my machines as a cache in front of a RAID array of hard drives. For this use case, even the 280GB model is larger than necessary. But as a primary storage device, the 480GB model would definitely feel less crowded.
Given the high price per GB of the Optane SSDs so far, the upcoming 960 GB and 1.5 TB models of the Optane SSD are going to be an even tougher sell: The market for $1200+ SSDs is pretty small, and very few users actually need a full TB of data within ten microsecond's reach.
Our first round of power measurements of the Optane SSD 900p showed what we expected: the Optane SSD 900p requires far more power than M.2 NVMe SSDs, and usually ends up being less efficient than a good M.2 SSD in spite of the great performance of the Optane SSD. It's hard to score well on efficiency with an idle power draw of over 3.5W. The Optane SSD 900p did score a clear efficiency win for random reads at low queue depths, where its performance advantage over flash-based SSDs is greatest.
Don't hold your breath for a M.2 version of the 900p, or anything with performance close to the 900p. Future Optane products will require different controllers in order to offer significantly different performance characteristics. Higher sequential performance to compete against the top flash-based SSDs will require a higher channel count, making for a more expensive drive with an even larger and more power-hungry controller. Lower power consumption will require serious performance compromises. In the near term, we're much more likely to see a new controller that's a step up from the Optane Memory M.2's single channel, but not large enough to rule out using the M.2 form factor. A three or four channel controller should be able to fit within a M.2 card's physical, electrical and thermal limits, but would offer much lower performance than this Optane SSD 900p.
| 250-280 GB | 480-512 GB | 1TB | 2TB | |
| Samsung 960 EVO | $127.99 (51¢/GB) | $240.00 (48¢/GB) | $449.99 (45¢/GB) | |
| Samsung 960 Pro | $289.99 (57¢/GB) | $619.00 (60¢/GB) | $1227.00 (60¢/GB) | |
| Intel Optane SSD 900p AIC | $389.99 (139¢/GB) | $599.99 (125¢/GB) | ||
| Intel Optane SSD 900p U.2 | $369.99 (132¢/GB) |
For the most part, the Optane SSDs are holding to their MSRPs, leaving them more than twice as expensive per GB as the fastest NAND flash based SSDs. They're a niche product in the same vein as the extreme capacity models like Samsung's 2TB 960 PRO and 4TB 850 EVO. But where the benefits of expanded capacity are easy to assess, the performance benefits of the Optane SSD are more subtle. For most ordinary and even relatively heavy desktop workloads, high-end flash storage is fast enough that further improvements are barely noticeable.
Facebook
Twitter
RSS
69 Comments
View All Comments
Notmyusualid - Sunday, December 17, 2017 - link
So, when you are at gun point, in a corner, you finally concede defeat?I think you need professional help.
tuxRoller - Friday, December 15, 2017 - link
If you are staying with a single thread submission model Windows may we'll have a decent sized advantage with both iocp and rio. Linux kernel aio is just such a crap shoot that it's really only useful if you run big databases and you set it up properly.IntelUser2000 - Friday, December 15, 2017 - link
"Lower power consumption will require serious performance compromises.Don't hold your breath for a M.2 version of the 900p, or anything with performance close to the 900p. Future Optane products will require different controllers in order to offer significantly different performance characteristics"
Not necessarily. Optane Memory devices show the random performance is on par with the 900P. It's the sequential throughput that limits top-end performance.
While its plausible the load power consumption might be impacted by performance, not always true for idle. The power consumption in idle can be cut significantly(to 10's of mW levels) by using a new controller. It's reasonable to assume the 900P uses the controller derived from the 750, which is also power hungry.
p1esk - Friday, December 15, 2017 - link
Wait, I don't get it: the operation is much simpler than flash (no garbage collection, no caching, etc), so the controller should be simpler. Then why does it consume more power?IntelUser2000 - Friday, December 15, 2017 - link
You are still confusing load power consumption with idle power consumption. What you said makes sense for load, when its active. Not for idle.Optane Memory devices having 1/3rd the idle power demonstrates its due to the controller. They likely wanted something with short TTM, so they chose whatever controller they had and retrofitted it.
rahvin - Friday, December 15, 2017 - link
Optane's very nature as a heat based phase change material is always going to result in higher power use than NAND because it's always going to take more energy to heat a material up than it would to create a magnetic or electric field.tuxRoller - Saturday, December 16, 2017 - link
That same nature also means that it will require less energy per reset as the process node shrinks (roughly e~1/F).In general, pcm is a much more amenable to process scaling than nand.
CheapSushi - Friday, December 15, 2017 - link
Keep in mind a big part of the sequential throughput limit is the fact that the Optane M.2s are x2 PCIe lanes. This AIC is x4. Most NAND M.2 sticks are x4 as well.twotwotwo - Friday, December 15, 2017 - link
I'm curious whether it's possible to get more IOPS doing random 512B reads, since that's the sector size this advertises.When the description of the memory tech itself came out, bit addressability--not having to read any minimum block size--was a selling point. But it may be that the controller isn't actually capable of reading any more 512B blocks/s than 4KB ones, even if the memory and the bus could handle it.
I don't think any additional IOPS you get from smaller reads would help most existing apps, but if you were, say, writing a database you wanted to run well on this stuff, it'd be interesting to know that small reads help.
tuxRoller - Friday, December 15, 2017 - link
Those latencies seem pretty high. Was this with Linux or Windows? The table on page one indicates both were used.Can you run a few of these tests against a loop mounted ram block device? I'm curious to see what both the min, average and standard deviation values of latency look like when the block layer is involved.