Valkyrie Website
Storage

Intel Optane SSD 900p 280GB PCIe NVMe SSD Review

Testing & Methodology


In a bid to make our testing as thorough and as accurate as possible, we have devised a testing methodology that will give us the maximum amount of data. While this means that testing each drive will take considerably longer – easily up 36 hours – the overall picture that we can gather is bigger too. It helps both you as the reader and us as the reviewer’s to gain a clearer view of the drive’s overall performance, strengths, and weaknesses. It also provides a more accurate analysis on how each drive performs under different conditions.

Data Fillage

There are many ways to test a storage drive, and we have chosen to include close to every test possible. We will start with a typical benchmark analysis of the formatted drive with various village scenarios. However, depending on how a drive is built, the performance may differ depending on how much data it is already storing. Testing this, the drive will be tested empty as well as with 25%, 50%, and 75% space usage. For these tests, we will be using Anvil’s Storage Utilities, AS SSD Benchmark, ATTO Benchmark, and CrystalDiskMark.

Application and Gaming

Besides these synthetic fill tests, the drive is tested with PCMark 8’s storage benchmark to get real-world performance results.

Plain Drive Performance

Following the fillage tests, the partition gets removed, and we check the drive’s raw performance. These tests will determine the hardware’s capabilities while eliminating any interference from the set file system. AIDA64’s storage benchmark and IOmeter are both great tools for this, and they help us paint an even clearer picture of the drive. The IOmeter tests I run are quite different from the rest as they will give a final result after a longer performance test rather than a maximum or average score; this gives us yet another view of the tested drive over the regular tests.

Drive Conditioning

Durability is a vital factor when it comes to NAND-based drives as the chips only can be rewritten so many times before they’ll stop working properly. Modern SSDs come with wear-level algorithms to increase the lifetime and reduce this factor as much as possible. We also take this into consideration in our reviews even tho it would take years to simulate this under real-world conditions. Luckily, Anvil’s Storage Utilities comes with an endurance test that can perform this action much faster, but it is still a time-consuming process. process. Depending on the hardware, this can easily take several days, and we write thousands and thousands of small files in various sizes and compressibility to drive over and over.

Following this conditioning process, all of the previous benchmarks will be run again in the same scenarios and the same order. This will give us a great view on how a drive will perform after prolonged usage, and that is a very vital factor when you pick out your storage drive.

The Preparation and In Between

The drive will be formatted as NTFS with default settings, as most users will do. The quick format option isn’t used as that might affect the performance the first time data is written.

Modern operating systems all take care of the TRIM command in the background and it isn’t something to worry about for regular users. However, for these reviews, I will add a waiting period between each benchmark to make sure that the TRIM command has been fully executed and finished. After all, we want a clear picture of the drive’s performance with minimal interference from the operating system.

On each page that follows with the benchmark results, I have inserted result screenshots from the benchmarks, created drive analysis chart for the fill-level performance, and added drive comparison charts where possible.

Storage Bench System June 2017

Hardware

Software

  • AIDA64 Storage Benchmark
  • Anvil’s Storage Utilities
  • AS SSD Benchmark
  • ATTO Benchmark
  • CrystalDiskMark
  • IOMeter
  • PCMark 8 Storage Benchmark
  • Windows 10
Previous page 1 2 3 4 5 6 7 8 9 10 11Next page
Gigabyte Black Friday
Gigabyte AERO X16
Endorfy Arx 700 Air
Klevv Urbane V RGB
SAMA Power Supplies
Gigabyte AERO X16
SAMA Power Supplies
INNO3D RTX 50 Series
INNO3D RTX 50 Series

Related Articles

6 Comments

  1. sorry but what are the advantages in the real world if you do not report these tests:
    – windows 10 start up time
    – photoshop applications, premiere, etc., startup up time
    – h264 video file conversion, ecc.
    your test without this information is not complete …

    1. Hi Pauls

      When we test storage drives, we test them separate from the OS drive. Testing from an OS drive won’t show the full potential of a drive and might create interference in the results. As such, a boot time isn’t available. For SSDs, no matter which one, the boot time is so fast anyway that there is no reason to measure it – in my opinion. It really doesn’t matter if it takes you 4.5 or 4.7 seconds to boot windows, or does it? It’s something else for drives with caching technology such as SSHDs or when you use an Optane caching module. In those cases, boot times are important and will be compared.

      App and game loading testing is available through the PCMark 8 Storage test and they are easily compared to all the previous drives we’ve tested. As the apps themselves don’t show or log their start time, there is no real way to measure it with comparable figures. Using a stopwatch, for example, would be way too inaccurate. There’s also the matter of costs of the software, the Adobe Suite isn’t exactly cheap and we do not work with pirated software.

      If you got any applicable test ideas for the video conversion test, let us know and we’ll see if that’s something worth incorporating into future reviews.

      1. Hi there,

        in my opinion it is very important to test the drive as a system drive and measure OS boot time and real world applications performance. Most of us (if not all of us) are going to use this drive as a system drive. Optane 900p is either too expensive or having not enough capacity for all other applications.
        Intels last “consumer” SSD 750 has a huge performance drop in terms of OS booting time (I’m not talking about 4.5 or 4.7 seconds, it was > 20 seconds). So it would be very useful, if you could extend your test and measure at least this value.

        Sorry for my bad English.

  2. Thank you for the reply
    unfortunately you have not tested a samsung 960 so I can not make a comparison,
    but by reading the pc mark 8 data it seems that in the real world the differences are minimal even though it has absolute values in 4k that no other ssd has at this time.
    For video conversion tests you could try it with handbrake that is free

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button
Close

Adblock Detected

Please consider supporting us by disabling your ad blocker!   eTeknix prides itself on supplying the most accurate and informative PC and tech related news and reviews and this is made possible by advertisements but be rest assured that we will never serve pop ups, self playing audio ads or any form of ad that tracks your information as your data security is as important to us as it is to you.   If you want to help support us further you can over on our Patreon!   Thank you for visiting eTeknix