Testing Optane Memory

For this review, Intel provided a fully-assembled desktop system with Windows 10 pre-installed and Optane Memory caching configured and enabled. The system was assembled by Intel's Demo Depot Build Center as the equivalent of a typical low to mid-range retail desktop with an i5-7400 processor, a B250 motherboard and 16GB of RAM. Storage is a 1TB 7200RPM WD Black hard drive plus the Optane Memory 32GB module.

Intel Optane Memory Review System
CPU Intel Core i5-7400
Motherboard ASUS B250-PLUS
Chipset Intel B250
Memory 2x 8GB Kingston DDR4-2400 CL17
Case In Win C583
Power Supply Cooler Master G550M
OS Windows 10 64-bit, version 1607
Drivers Intel Optane Memory version 15.5.0.1051

In addition, we tested the Optane Memory's performance and power consumption as a standalone SSD using our own testbed. This allowed us to compare against the Optane SSD DC P4800X and to verify Intel's performance specifications for the Optane Memory.

Unfortunately, this review includes only an abbreviated set of benchmarks, for two reasons: the Optane Memory review system arrived less than a week ago, as I was trying to finish up the P4800X review, and the Optane Memory module did not survive testing. After about a day of benchmarking the Optane Memory review system locked up, and after rebooting the Optane Memory module was not detected and the OS installation was corrupted beyond repair. The drive is not completely dead: Linux can detect it as a NVMe device but cannot use it for storage or even retrieve the drive's error log. In communicating with Intel over the weekend, we were not able to figure out what went wrong, and the replacement module could not be delivered before the publication of this review.

The fact that the Optane Memory module died should not be taken as any serious evidence against the product's reliability. I kill review units once every few months during the course of ordinary testing, and I was due for another failure (ed: it's a bona fide AnandTech tradition). What we call ordinary testing is of course not something that anybody would mistake for just the intended use of the product, and no SSD brand has been entirely free from this kind of problem. However, the fact remains that we don't have as much data to present as we wish, and we don't have enough experience with the product to make final conclusions about it.

For comparison with the Optane Memory caching configuration, we selected the Crucial MX300 525GB and the Samsung 960 EVO 250GB. Both of these are available at retail for slightly less than the price of the Optane Memory 32GB module and the 1TB hard drive. They represent different capacity/performance tradeoffs within the same overall storage budget and are reasonable alternatives to consider when building a system like this Optane Memory review system.

For testing of the Optane Memory caching performance and power consumption, we have SYSmark 2014 SE results. Our synthetic tests of the Optane Memory as a standalone SSD are abbreviated forms of the tests we used for the Optane SSD DC P4800X, with only queue depths up to 16 considered here. Since those tests were originally for an enterprise review, the drives are preconditioned to steady state by filling them twice over with random writes. Our follow-up testing will consider the consumer drives in more ordinary workloads consisting of short bursts of I/O on drives that are not full.

Intel's Caching History SYSmark 2014 SE
Comments Locked

110 Comments

View All Comments

  • evilpaul666 - Thursday, April 27, 2017 - link

    Everyone presumes that technology will improve over time. Talking up 1000x improvements, making people wait for a year or more, and then releasing a stupid expensive small drive for the Enterprise segment, and a not particularly useful tiny drive for whoever is running a Core i3 7000 series or better CPU with a mechanical hard drive, for some reason, is slightly disappointing.

    We wanted better stuff now after a year of waiting not at some point in the future which was where we've always been.
  • Lehti - Tuesday, April 25, 2017 - link

    Hmm... And how does this compare to regular SSD caching using Smart Response? So far I can't see why anyone would want an Optane cache as opposed to that or, even better, a boot SSD paired with a storage hard drive.
  • Calin - Tuesday, April 25, 2017 - link

    Did you brought the WD Caviar to steady state by filling it twice with random data in random files? Performance of magnetic media varies greatly based on drive fragmentation
  • Billy Tallis - Wednesday, April 26, 2017 - link

    I didn't pre-condition any of the drives for SYSmark, just for the synthetic tests (which the hard drive wasn't included in). For the SYSmark test runs, the drives were all secure erased then imaged with Windows.
  • MrSpadge - Tuesday, April 25, 2017 - link

    "Queue Depth > 1

    When testing sequential writes at varying queue depths, the Intel SSD DC P3700's performance was highly erratic. We did not have sufficient time to determine what was going wrong, so its results have been excluded from the graphs and analysis below."

    Yes, the DC P3700 is definitely excluded from these graphs.. and the other ones ;)
  • Billy Tallis - Wednesday, April 26, 2017 - link

    Oops. I copied a little too much from the P4800X review...
  • MrSpadge - Tuesday, April 25, 2017 - link

    Billy, why is the 960 Evo performing so badly under Sysmark 2014, when it wins almost all synthetic benchmarks against the MX300? Sure, it's got fewer dies.. but that applies to the low level measurements as well.
  • Billy Tallis - Wednesday, April 26, 2017 - link

    I don't know for sure yet. I'll be re-doing the SYSmark tests with a fresh install of Windows 10 Creators Update, and I'll experiment with NVMe drivers and settings. My suspicion is that the 960 EVO was being held back by Microsoft's horrific NVMe driver default behavior, while the synthetic tests in this review were run on Linux.
  • MrSpadge - Wednesday, April 26, 2017 - link

    That makes sense, thanks for answering!
  • Valantar - Tuesday, April 25, 2017 - link

    Is there any reason why one couldn't stick this in any old NVMe-compatible motherboard regardless of paltform and use a software caching system like PrimoCache on it? It identifies to the system as a standard NVMe drive, no? Or does it somehow have the system identify itself on POST and refuse to communicate if it provides the "wrong" identifier?

Log in

Don't have an account? Sign up now