ATTO Disk Benchmark

In the graph above is a graphical representation of the data from the ATTO Disk Benchmark Read test across all drives.  To see the specific numbers, please refer to the screenshot above this graph.  What we see from this graph is that the Mushkin Reactor does have a lower read speed across the board to the end.  The new TeamGroup T-Force Vulcan SSD actually has a slightly higher consistent read speed from 64KB to the end over the SAMSUNG 860 EVO, which is impressive.  The TeamGroup SSD got up to 526 MB/s in the read speed test.    

In this graph, we are looking at the write speed in graph form.  It is clear that the SAMSUNG 860 EVO does have a higher consistent write speed, but in reality, the results very close together.  What’s not together is the Mushkin Reactor write speeds, it’s much slower around 400 MB/s writes.  The new TeamGroup T-Force Vulcan SSD smashes the write speed of the Mushkin Reactor SSD and comes very close to the SAMSUNG 860 EVO, so this is impressive.

AJA System Test

We are introducing a new test to our benchmark suite for SSDs, AJA System Test.  This one is unique; it tests actual high-quality video and its ability to stream from the SSD.  For our test we are using the 3840×2160 (4K) Ultra HD video setting with a 4GB file size and 10-bit YUV color.  We found this test to be very unique in testing performance.

According to this benchmark, the TeamGroup T-Force Vulcan SSD is on top in read performance at 520 MB/s.  There might be some caching here since this does exceed the maximum 510 MB/s read speed, though we did select the option to disable any caches, and we tested and re-tested this.  The SAMSUNG 860 EVO and the MUSHKIN REACTOR both came in with 517 MB/s of read. 

On the write speed, the SAMSUNG 860 EVO is slightly faster at 463 MB/s write, which tracks with other results we have done.  The TeamGroup SSD follows it very closely at 450 MB/s.  The Mushkin Reactor falls way behind both, so the write tests are in-line with other benchmarks. 

Recent Posts

Brent Justice

Brent Justice has been reviewing computer components for 20+ years, educated in the art and method of the computer hardware review he brings experience, knowledge, and hands-on testing with a gamer oriented...

Join the Conversation

4 Comments

  1. "There is no reason to have a mechanical drive anymore with prices like this. 500GB is plenty for a boot drive, or as your only drive."

    Uhh, not with 50+GB game downloads it’s not. Boot drive / laptop upgrade maybe.

    I have a $100 Toshiba 5TB hdd for my game drive / steam library. I look forward to that price for a huge SSD, 5+ years from now. But by the time that happens, the $100 hdd will probably be a 10+TB one.

  2. Nice to see a large amount of test programs used in the review. TeamGroup seems to have some of the better options for price/perf these days for sata drives. Happy so far with mine.

    I agree with Burt though, 500gb is my standard now for a boot drive and maybe 1-2 games. Everything else on dedicated ssd’s or hdd. 250GB is workable but getting too small and for quality brands it seems worth it to just jump to 500gb at $50-65 over a $30-40 250gb.

  3. Oops, missed this review. Adding some extra detail.

    Very good drive – basically a Crucial MX500 clone. The MX500 often comes with 96L flash now (which is also denser, but has twice the planes). Team’s coding here is clearly from Intel/Micron with the "01T" meaning 1Tb packages (x4 = 512GiB). This is in contrast to the similar L5 Lite 3D and drives like it – SU800, X3, S280, etc – which use inferior flash. Much like the L5 Lite 3D, it has static SLC – ~6GB at this capacity – which makes for uniform performance (can see this in HD Tune); the MX500 has a significantly larger, dynamic cache that is slower outside SLC at this capacity. Warranty is of course lackluster, but if you want a top-tier SATA drive this is among them in terms of performance.

    The firmware revision matches typical SM2258 encoding which usually refers to the type of flash it’s using. For example you common see Qx6/7xxA in the cheaper drives using second-tier Samsung TLC, Hynix’s 72L TLC (which isn’t great), inferior BiCS (SanDisk/Kioxia), older Intel/Micron 32L TLC, etc. Older revisions were for the 32L/384Gb IMFT TLC. In any case, R/S is for the newer and better flash products, e.g. B16A/B17A which is 64L Intel/Micron TLC (B17A is double the density but also double the planes, so same interleaving).

    My only advice for reviewing is that FIO is a strong option for benchmarking, whether on Linux or Windows, although ultimately you would require Linux for full utility. I’ve posted on Reddit about using FIO and also making a dedicated Linux benchmarking USB drive, I haven’t posted about getting these tools working on Windows – it is possible, I have it working in Visual Studio (direct) or Cygwin (indirect), but it’s fairly limited and a pain to work with in contrast to nvme-cli for example.

  4. I should add that I’ve seen reviews of it like TPU’s showing MLC and in fact I had someone with one show me one with MLC – 48L from Samsung, to be specific. And TPU also had 64L TLC from Samsung I believe. That type of older/inferior flash is often found in the L5 Lite 3D and other budget drives as I mentioned above – my theory is that the early Vulcans came basically as L5 Lite 3Ds. It’s my understanding that currently they do have Micron flash though. Just wanted to correct that as people don’t understand that second-tier Samsung can actually be inferior, even the MLC, although TPU’s results were still good performance-wise. (also as I mentioned, firmware revision can hint at flash – the early Samsung variants match the L5 Lite 3D’s firmware)
Leave a comment