The Intel Optane SSD 900P 280GB Review
by Billy Tallis on October 27, 2017 9:30 AM ESTRandom Read Performance
Our first test of random read performance uses very short bursts of operations issued one at a time with no queuing. The drives are given enough idle time between bursts to yield an overall duty cycle of 20%, so thermal throttling is impossible. Each burst consists of a total of 32MB of 4kB random reads, from a 16GB span of the disk. The total data read is 1GB.
The Optane SSD 900P doesn't break the record for QD1 random reads, but only because we've also tested the 32GB Optane Memory M.2, which is about two microseconds faster on average for each 4kB read. The Optane SSD 900P is still about 7 times faster than any flash-based SSD.
Our sustained random read performance is similar to the random read test from our 2015 test suite: queue depths from 1 to 32 are tested, and the average performance and power efficiency across QD1, QD2 and QD4 are reported as the primary scores. Each queue depth is tested for one minute or 32GB of data transferred, whichever is shorter. After each queue depth is tested, the drive is given up to one minute to cool off so that the higher queue depths are unlikely to be affected by accumulated heat build-up. The individual read operations are again 4kB, and cover a 64GB span of the drive.
When longer transfers and higher queue depths come into play, the Optane SSD 900P passes the Optane Memory M.2 and remains more than 6 times faster for random reads than any flash-based SSD.
 |
|||||||||
Both Optane devices more or less level off at queue depths of 8 or higher. The Optane SSD 900P saturates at about 1800 MB/s while the Optane Memory tops out around 1300 MB/s. The Samsung 960 PRO 2TB hasn't caught up by QD32, and doesn't surpass the QD1 random read performance of the Optane SSD until the Samsung reaches a queue depth of about 8.
Random Write Performance
Our test of random write burst performance is structured similarly to the random read burst test, but each burst is only 4MB and the total test length is 128MB. The 4kB random write operations are distributed over a 16GB span of the drive, and the operations are issued one at a time with no queuing.
The burst random write performance of the Optane SSD 900P is slightly higher than the Intel SSD 750 1.2TB, and about 14% faster than Samsung's fastest.
As with the sustained random read test, our sustained 4kB random write test runs for up to one minute or 32GB per queue depth, covering a 64GB span of the drive and giving the drive up to 1 minute of idle time between queue depths to allow for write caches to be flushed and for the drive to cool down.
With higher queue depths in play, the Optane SSD 900P scales up faster than the Intel SSD 750 1.2TB, leaving the Optane SSD with a 7-10% lead over the Samsung 960s and Intel 750.
 |
|||||||||
Samsung's 960 PROs and the larger 960 EVO all trail slightly behind the Optane SSD's random write performance for queue depths 1 to 4, then the Samsung drives level off and leave the Optane SSD with a substantial performance advantage at high queue depths. The Intel 750 is slightly faster at QD1 and QD2, but saturates at an even lower performance level than the Samsung 960s.
Facebook
Twitter
RSS
205 Comments
View All Comments
ddriver - Friday, October 27, 2017 - link
I'd reserve judgement in regard with endurance until I see how many TBs written before it cr@ps out.Laptops - that activates my hilarity unit. It can only shine in enterprise workloads, laptops are inherently underpowered and targeted at completely different workloads.
We need to get it into smart watches, now that will be a game changer. Imagine the possibilities ;)
BrokenCrayons - Friday, October 27, 2017 - link
"Laptops - that activates my hilarity unit. It can only shine in enterprise workloads, laptops are inherently underpowered and targeted at completely different workloads."The millions of customers that purchase mobile workstations find your misunderstand laughable.
ddriver - Friday, October 27, 2017 - link
I highly doubt mobile workstations sell in the millions. They are a rather niche market.And I know that mobile workstations are used for very different workloads from the enterprise. Clearly, you lack insights into "workstation" and "enterprise" ;)
BrokenCrayons - Friday, October 27, 2017 - link
If, by "enterprise" you're trying to imply server workloads, then sure, I'd agree that mobile workstations fill a different computing role in the workplace. However, there's nothing preventing a storage-intensive workload from ending up on any sort of laptop, even a consumer-oriented system.Why intentionally close the door to storage technologies? Is it a brand loyalty hangup? Do you really care that much about what company logo is on a product? Its sort of an inferior and defective mental state you've got there over something as meaningless as what company developed a particular product. Or are you close-minded over the idea that the performance claims weren't met? That's almost as irrational as brand loyalty (or dis-loyalty in your case...which is equally silly, by the way). Decrying a reasonably priced product that performs well because it didn't live up to a marketing claim merely means you haven't the capacity to put aside your emotions and see a product for its usage past the crap you get fed by meaningless product announcements. That's just as dumb as falling prey to marketing to begin with. Either way, someone with the ability to control their own emotions and think rationally wouldn't be in the situation you're in right now.
ddriver - Friday, October 27, 2017 - link
There is no such thing as a "server workstation". There is this thing called "server", and this thing called "workstation". There is some very minimal possible overlap in specs, but the two are used for completely different purposes.Hypetane's advantage is mostly due to the controller rather than the storage medium. And that performance obviously comes at a high power cost. Which is why I am not seeing this make its way into laptops, where it will offer next to no real world performance benefits while being a significant power drain.
You love wasting money to have worse battery life or something?
Emotions? U crazy? This is electronics, it's just stuff, inanimate matter. Who would see it in an emotional context, aside from braindead fanboys?
One of the many things I do is also to play the guitar. People are always like "you don't love your guitars enough" - coming from people who collect guitars for decoration and barely even play and naturally suck at it. To which I reply - "to me the guitar is not a fashion accessory, it is an instrument, I cannot love it any more than I can love a hammer or a wrench or an electric drill".
It is the same for hardware. Some may buy it for bragging rights or self esteem. I buy what I can afford as long as it can serve the purpose I buy it for. Surely I didn't like buying intel CPUs the last 10 years, but when AMD's got garbage, there is nothing I can do about it.
What drives my criticism of intel is their shitty act. They pretend as if they are the drive of innovation, but they have actually held technology a hostage through their monopoly, and impeded competition and progress tremendously, causing irreparable damage to humanity. You don't have the capacity to understand how much better things could have been had progress not been hostage of the greedy and corrupt but I do. And they keep on doing it even after being caught red-handed. As they say, "a leopard cannot change its spots".
The real problem however is "people" like you being reduced so low, not only do you NOT have a problem with the degradation and exploitation of your own species, you go forth to applaud it. Honestly, what does intel have to do to get criticism from you? Run over your dog? Burn your house down? Chop your hands off?
It is your fault that intel's act is so shitty. People like you, who let it slide and even applaud it. It is your fault amd wasn't competitive for a decade. It is your fault that they ask 10000 dollars for a piece of silicone that costs 50 dollars to make. AT servers will run out of storage space before I list all the shit "people" with your mindset and responsible for.
BrokenCrayons - Friday, October 27, 2017 - link
Yup, easily manipulated.lmcd - Friday, October 27, 2017 - link
Please help me understand how GloFo was ever going to catch up enough for AMD to realistically have competed? They're only finally competing with the slowdown post-Moore's Law. Blame whatever you want but Intel's process advantage brought them massive performance advantages and design advantages. Add in the importance of yields -- as core count increases, yield becomes massively more important. AMD never was going to compete with Intel until the barrier to improvement became physics itself, and not merely our tools of manipulation.utmode - Saturday, October 28, 2017 - link
How come Moore's Law is law when it is not a law anymore or never been. ItAndrewJacksonZA - Friday, October 27, 2017 - link
"Intel has almost taken all the fun out of testing a SSD.":-)
jjj - Friday, October 27, 2017 - link
The drive has very high power consumption and the power meter fails? That's very very suspicious and you should have delayed the review.To make it worse, the results will be published only in bench and the review not updated?
I can't trust you anymore, ever.