Sequential Read Performance

Our first test of sequential read performance uses short bursts of 128MB, issued as 128kB operations with no queuing. The test averages performance across eight bursts for a total of 1GB of data transferred from a drive containing 16GB of data. Between each burst the drive is given enough idle time to keep the overall duty cycle at 20%.

Burst 128kB Sequential Read (Queue Depth 1)

The burst sequential read speed of the MyDigitalSSD SBX is clearly faster than what SATA drives offer, but the Samsung drives and even the larger capacities of the Intel SSD 760p are in an entirely different league. The Phison E7 drives with planar MLC offer about the same performance as the SBX.

Our test of sustained sequential reads uses queue depths from 1 to 32, with the performance and power scores computed as the average of QD1, QD2 and QD4. Each queue depth is tested for up to one minute or 32GB transferred, from a drive containing 64GB of data.

Sustained 128kB Sequential Read

The flash is clearly much more of a bottleneck than the PCIe x2 interface on the longer sequential read test, where the internal fragmentation left over from the random write test prevents the SBX from delivering data any faster than the SATA SSDs. The Intel SSD 760p performs even worse at 128GB, but beats the SBX at 512GB.

Sustained 128kB Sequential Read (Power Efficiency)
Power Efficiency in MB/s/W Average Power in W

Power efficiency continues to be a downside to the MyDigitalSSD SBX even though its power draw in absolute terms is substantially lower than the other NVMe drives. The Intel 760p is clearly worse off this time with much lower efficiency at every capacity, while Samsung's high-end drives are fast enough to offer efficiency as good or better than SATA drives.

The sequential read performance of the SBX improves a bit between QD2 and QD8, allowing it to eventually surpass the Intel 760p that is generally better at QD1 but doesn't scale up with higher queue depths.

Sequential Write Performance

Our test of sequential write burst performance is structured identically to the sequential read burst performance test save for the direction of the data transfer. Each burst writes 128MB as 128kB operations issued at QD1, for a total of 1GB of data written to a drive containing 16GB of data.

Burst 128kB Sequential Write (Queue Depth 1)

The burst sequential write speed of the MyDigitalSSD SBX exceeds that of SATA SSDs, though only barely in the case of the 128GB SBX. The 128GB Intel 760p fails to keep up with mainstream SATA drives, while the larger capacities have a clear lead over the SBX.

Our test of sustained sequential writes is structured identically to our sustained sequential read test, save for the direction of the data transfers. Queue depths range from 1 to 32 and each queue depth is tested for up to one minute or 32GB, followed by up to one minute of idle time for the drive to cool off and perform garbage collection. The test is confined to a 64GB span of the drive.

Sustained 128kB Sequential Write

On the longer sequential write test, only the 512GB SBX is able to stay ahead of mainstream SATA SSDs. The Intel 760p has a minimal performance lead over the SBX at 128GB, growing to a 22% advantage at 512GB.

Sustained 128kB Sequential Write (Power Efficiency)
Power Efficiency in MB/s/W Average Power in W

The MyDitigalSSD SBX's power efficiency during sequential writes is better than Intel's drives, but not much else. The SBX is paying the power cost of PCIe (albeit only two lanes) but delivering SATA performance.

The MyDigitalSSD SBX delivers essentially the same sequential write performance at all tested queue depths, and power usage is also flat. The results from the SBX are most comparable to mainstream SATA drives, while the high-end NVMe drives are generally much faster and more power hungry.

Random Performance Mixed Read/Write Performance
Comments Locked

46 Comments

View All Comments

  • dgingeri - Tuesday, May 1, 2018 - link

    For $53 for a 128GB one, with a 5 year warranty? That's now the boot drive of my server.
  • dgingeri - Monday, May 7, 2018 - link

    It has worked remarkably well as a server boot drive. I highly recommend it.
  • vailr - Tuesday, May 1, 2018 - link

    An external USB 3.0 connected PCIe M.2 type NVMe adapter would be faster than any USB thumb drive, and would be ideal for a bootable external "Windows to Go". Is such a device available yet?
    Something like this: https://www.amazon.com/StarTech-com-M-2-SATA-SSD-E... but compatible with PCIe NVMe M.2 80mm drives, such as this MyDigitalSSD, or the Samsung 960 NVMe, for example.
  • MajGenRelativity - Tuesday, May 1, 2018 - link

    The only ones I've seen are Thunderbolt adapters, which require a Thunderbolt port. They all come populated with an SSD too. The "cheapest" I've seen is the TekQ Rapide, which while priced below others and with decent performance, isn't exactly cheap at $250
  • PeachNCream - Tuesday, May 1, 2018 - link

    Since NVMe SSDs still command a price premium even with low cost drives like the SBX out there, it may just not make a lot of sense to build NVMe-to-USB drive enclosures. After all, SATA 3.0 is rated to 6 Gbit/s and USB 3.0 is rated at 5 Gbit/s which means you're already going to be at the saturation point of USB 3.0 with a SATA SSD in a USB enclosure at a relatively low cost for a removable boot drive. I've been doing something like that with a 2.5 inch SATA to USB 3.0 enclosure and a spare 120GB Patriot Torch. Ubuntu happily boots from it and I can't really discern much difference (responsiveness, performance, read/write speed, etc.) between using the drive in the external enclosure and using that same drive on my laptop's internal SATA connector.
  • Death666Angel - Tuesday, May 1, 2018 - link

    http://www.microsatacables.com/m-2-ngff-pcie-ssd-t...
    This explicitly states PCIe (and is out of stock), all others just state SATA M.2. But as Peach described, USB 3.0 is already saturated by SATA 3.0.
    You could frankenstein something. Get one of those PCIe slot to USB things the mining community uses, then a PCIe to M.2 NVME adapter and then hope it somehow works. :D Not pretty though. ;)
  • MajGenRelativity - Tuesday, May 1, 2018 - link

    The PCIe slots to USB cable just repurpose the pins on the connector to carry PCIe signals. They do NOT follow USB communication protocols
  • Death666Angel - Tuesday, May 1, 2018 - link

    Thanks for that info and sorry for my misinformation. :)
  • MajGenRelativity - Thursday, May 3, 2018 - link

    No problem. I also checked out the drive you linked, and it only supports the one Samsung OEM drive that uses PCIe with the AHCI protocol, not NVMe. Not sure why it doesn't support NVMe, but it says it doesn't, so good idea to keep an eye on that.
  • dgingeri - Tuesday, May 1, 2018 - link

    That would presume that there is a USB to PCIe adapter chip, which there isn't. Thunderbolt, as previously mentioned, is available, but that is because Thunderbolt is based on PCIe anyway. So, no bridge chip is required.

Log in

Don't have an account? Sign up now