The Intel Optane SSD 900P 280GB Review
by Billy Tallis on October 27, 2017 9:30 AM ESTDrive Features
Intel's strategy for developing the Optane SSD 900P is the same they used for the SSD 750 in 2015. Both products are consumer-oriented derivatives of the flagship enterprise SSD of their time. As with the SSD 750, this has resulted in the Optane SSD 900P being a physically large and power-hungry drive that can't come close to fitting in the M.2 form factor used by most other consumer PCIe SSDs. The controller alone on the Optane SSD 900P is too large to fit on a 22mm-wide M.2 module, and its idle power draw of 5W is more than most M.2 PCIe SSDs draw under load.
The Optane SSD 900P shares the same platform as the enterprise Optane SSD DC P4800X. It uses the same Intel controller with a PCIe 3.0 x4 link to the host system and a seven channel interface to the 3D XPoint memory. The P4800X is available as a 375GB drive that uses four dies per channel for a raw capacity of about 448GB. The 900P uses three or five dies per channel and similar overprovisioning ratios to offer usable capacities of 280GB and 480GB from raw capacities of about 336GB and 560GB. Because 3D XPoint memory supports direct read and in-place write operations with bit granularity instead of being constrained by the page and block structure of NAND flash memory, the Optane SSD 900P isn't using the extra capacity as spare area for any garbage collection process and is likely instead devoting most of it to error correction overhead and metadata.
The odd capacities and channel counts of the Optane SSD 900P aren't a requirement for using 3D XPoint memory so much as they are a consequence of Intel's performance and power targets for the drive. However, the use of 3D XPoint memory does lead to a few striking hardware design differences from flash-based SSDs. The Optane SSD 900P does not have any DRAM cache. On a NAND flash-based SSD, this would be a major red flag indicating reduced performance, even when mitigated by the NVMe Host Memory Buffer feature. Flash-based SSDs need quick access to the mapping tables that track which logical blocks are stored in which physical pages. 3D XPoint memory, on the other hand, is much simpler to manage, with no need for garbage collection and little or no need for wear leveling, and the storage medium itself is far faster. This makes the large DRAM buffer unnecessary.
The lack of buffers seems to also extend to the individual 3D XPoint chips themselves. NAND flash dies include page-sized RAM buffers as intermediate storage between the controller and the flash memory array. This allows the controller to transfer the payload data for a write command to the NAND die and then move on to issuing other commands to other dies on the same memory channel (or other planes on the same die) while the relatively long NAND program operation is happening. This multiplexing of operations within a single memory channel is a large part of why flash SSDs tend to be faster at higher capacities even within the same product line with the same SSD controller. The Optane SSD 900P accesses its 3D XPoint memory more directly and read and write operations are completed synchronously without an intermediate buffering stage. This means that the 280GB and 480GB models have essentially the same performance, and this same level of performance could in theory extend down to a one die per channel configuration that would offer 112GB raw capacity and 96GB usable capacity (and could be profitably sold for under $180).
While the lack of RAM buffers on the Optane SSD doesn't bring any major performance problems, it does have one minor downside. Flash-based SSDs are able to aggressively take advantage of their buffers to accept write operations almost as fast as the PCIe link allows for, so the Optane SSD 900P doesn't have much room to improve peak write speeds beyond what the 3D XPoint memory itself can sustain. However, the Optane SSD can continue providing those write speeds long after a flash-based SSD would have stalled to let the flash catch up to the controller.
The last major implication of the DRAMless design of the Optane SSD 900P is that large banks of capacitors are unnecessary to provide power loss protection. The Intel SSD 750 was one of the few consumer SSDs to offer enterprise-grade power loss protection, but those capacitors added to an already high BOM and were probably only kept to save the effort of more significant firmware changes relative to the enterprise SSDs it was based on. With the Optane SSD 900P, Intel gets to retain power loss protection on the consumer version more or less for free. This allows the Optane SSD 900P to avoid the consequences of Microsoft's stupid NVMe driver write cache policy defaults, which severely curtail the write performance of consumer NVMe drives that have volatile write caches. (Our client SSD test protocol uses the more reasonable but non-default write buffer policy for all Windows-based tests, so this advantage doesn't show up in out performance charts.)
When an SSD vendor makes consumer and enterprise drives based on the same controller platform, there are usually major firmware differences to optimize for different performance requirements. Consumer SSDs tend to use SLC write caching to offer better burst write performance while enterprise SSDs tend to avoid it to offer more consistent performance. There are few opportunities with the Optane SSD to tweak performance between the consumer and enterprise versions, but there are several features from the P4800X that have been omitted from the 900P. The only one that has any performance impact is the ability to reformat the P4800X to use a 4kB logical block size instead of 512 bytes; the 900P only supports 512B sectors and doesn't support end-to-end data protection. The 900P also omits the SMBus interface for out of band management and doesn't include the diagnostic LEDs.
The most significant functional difference between the P4800X and the 900P is the endurance rating, which has been reduced from 30 drive writes per day to 10 DWPD. That's still a very high figure for a consumer SSD, but Intel's policy of putting the drive into read-only mode when the rated endurance is reached can be a nuisance, and it'll be encountered sooner on the 900P. The difference in write endurance is proportional to the price difference between the 900P and the P4800X, so the de-rating isn't too unfair. Overall, the Optane SSD 900P is still the closest a consumer SSD gets to offering enterprise SSD performance and features, and the 900P will probably cannibalize some P4800X sales.
Note that our usual set of power measurements is unavailable due to an equipment failure. The Optane SSD 900P will be re-tested once we have a new power meter, and the results will be posted to the SSD Bench database. In the meantime, rest assured that the Optane SSD 900P won't win any awards for efficiency.
AnandTech 2017 SSD Testbed | |
CPU | Intel Xeon E3 1240 v5 |
Motherboard | ASRock Fatal1ty E3V5 Performance Gaming/OC |
Chipset | Intel C232 |
Memory | 4x 8GB G.SKILL Ripjaws DDR4-2400 CL15 |
Graphics | AMD Radeon HD 5450, 1920x1200@60Hz |
Software | Windows 10 x64, version 1703 |
Linux kernel version 4.12, fio version 2.21 |
- Thanks to Intel for the Xeon E3 1240 v5 CPU
- Thanks to ASRock for the E3V5 Performance Gaming/OC
- Thanks to G.SKILL for the Ripjaws DDR4-2400 RAM
- Thanks to Corsair for the RM750 power supply, Carbide 200R case, and Hydro H60 CPU cooler
205 Comments
View All Comments
ddriver - Friday, October 27, 2017 - link
I got myself a bucked of salt. The necessary requirement to swallow that Houdini "2.7x better" claim from the launch PR.I've been rendering stuff since the days of 3d max for frigging DOS. And I am yet to experience a scenario where CPU load is not in the 99% range.
Having a rendering job that cannot feed the CPU to above 10% load with the insanely fast 960 pro has got to be an unprecedented case of cooked-up benchmark in human history.
extide - Friday, October 27, 2017 - link
Did you read the article? It pretty clearly explains how they got that result, and it makes sense.ddriver - Friday, October 27, 2017 - link
Oh yeah, I get it. Hypetane is a synthetic beast. Which allows to showcase said advantage as long as you focus on it in a carefully devised and completely detached from real-world usage workload.Don't get me wrong. It is good that hypetane is now available in capacities that actually allow to use it. And if endurance turns out to be tangibly better than nand, I might actually buy it. Low queue depth performance is good, especially random read, which may not be of that much practical use to most of the people out there, but I could make good use of that.
But it will remain "hypetane" even after I go and buy it. Because intel said "1000 times better", and it is not even 10 times better. A zero on its own might be nothing, but two zeroes after a positive number make quite a lot of difference.
ddriver - Friday, October 27, 2017 - link
"no other alternative nonvolatile memory technology is close to being ready to challenge 3D XPoint"Except for SLC, which was so good it was immediately abandoned once inferior and more profit friendly NAND implementations were available.
A SLC based product coupled with MRAM cache will easily humiliate hypetane in its few strong aspects.
Too bad NAND drives are now moving to TLC and QLC, even MLC is heading in the "luxury item" category. Too bad because 3D SLC has tremendous potential. Let's see if it gets realized.
extide - Friday, October 27, 2017 - link
How would that work. SLC is slower than Optane, can't be written at a block level, needs trash collection, etc. Then you cache it with a technology similar to Optane? Why not just build a drive with all MRAM, oh yeah, too expensive. Looks like Optane wins.ddriver - Friday, October 27, 2017 - link
Nope, SLC is actually faster. Look it up.And what it cannot do is write at the bit level. Which is not really a big deal. Even CPUs cannot address RAM at bellow a byte, if you want single bit operations, you have to use bitwise operators. Writing at a higher level is actually very efficient, because it reduces overhead. If single bit addressing was important, that's who computers would work.
Furthermore, single bit writes produce a significant challenge when tracking wear levels. Hypetane still wears out, you know... It will be tremendously harder to accurately track wear at bit level, and I am abot 99.999999% sure it is not how intel does it, meaning that a lot of that supposed extra endurance will be forfeited by managing wear at a coarsely grained level. They won't be managing that at bit level, the overhead will be tremendous and will completely diminish potential advantages.
The MRAM cache will reduce a lot of write amplification and garbage collection.
It also looks like 3d SLC has about 3 times the density of the chips intel is currently using for hypetane.
"Why not just build a drive with all MRAM" - density is too low. Which is also why we use RAM for working memory, I mean volatility can easily be solved by say adding a RTG battery to a DRAM drive, giving it effectively about a century of continuous, uninterrupted power. It is doable, but then again, redundant, and while it is true that the industry does a lot of pointless things nowadays, the only ones that qualify are those with a desirable usability to profitability ratio, and a RTG DRAM drive is simply too good to offer...
"Looks like Optane wins" - anyone can win when running unopposed. The moment someone makes a SLC/MRAM hybrid and it loses to hypetane, I will retract my statement and admit I was wrong. I have zero problem with that ;)
vanilla_gorilla - Friday, October 27, 2017 - link
So you're saying Optane sucks because it would be slower than a drive that doesn't exist?ddriver - Friday, October 27, 2017 - link
No, I am saying it "sucks" because for all intents and purposes, it is not any faster than a 2 year old drive that it was supposed to beat by a 1000 times.And the reason I put it "sucks" is because I never said it does suck. I give it a very realistic valuation. What sucks is how far that realistic valuation is from what intel promised. Which is entirely on them.
name99 - Friday, October 27, 2017 - link
He's saying two distinct things.(a) This costs too much for what it delivers. IF Samsung wanted to compete with it, they could do so with a suite of existing technologies. But they probably won't do so because there is little demand for a product like this; honestly it only exists so that Intel can say "see, 3D-XPoint is too, real".
(b) The place where 3D-XPoint ACTUALLY makes sense is, more or less, what AnandTech says --- as a slower (but much larger) RAM replacement. That's what plays to the technology's strengths (simple controller, byte-level access). But Intel STILL are not shipping that --- which makes one wonder WTF not?
It IS reasonable to point out that Intel has been lying about this product since the day it was announced, and that the only reason they're shipping these SSD drives is to throw up more smoke to hide the fact that the actually sensible use case remains (for some reason) impossible.
Being a fanboy isn't about always praising your company, it's about refusing to criticize your company even when they're clearly in the wrong. Intel is clearly in the wrong here, in the sense that nothing that they promised about Optane is actually reality even today, two years after the announcement.
If you think that's reasonable behavior, ask yourself how you would react if your favorite villainous company did the same.
Would you be impressed if AMD announced that they're going to ship a GPU 1000x faster than the competition, and two years later all they have is something 2.7x as fast (under very specialized circumstances)?
Would you let Apple off the hook if they said that the Apple car was going to have 1000x the range of a Tesla, then they shipped two years later, a car with 2.7x the range of a Tesla?
Drumsticks - Saturday, October 28, 2017 - link
Re: AMD example: if AMD claimed a product would be 100x or 1000x faster than Nvidia, but only delivered something 6-10x faster in the majority of cases, and on par in the rest, for only 2-3x more money, I'd still be pretty satisfied.