Interesting test results. Something very interesting to see is that handling of small packet sizes of data transfer is quite bad across all types of connection and device.
Definitely shows that USB3 is allowing much more bandwidth for sustainable sequential and large packet transfer which is great. I was surprised by the maximum read and write speeds for the SATAII drive, even on USB3. SATAII drives have a theoretical transfer maximum of 300MB/s, with the drive mechanisms able to transfer about 100-130MB/s for current hardware, so with a max for the sequential speed of 72 for read and 73 for write still falls quite short of that target.
To see hoe your results compare with internal HDDs connected via SATA I just ran a couple of benchmarks using the same tool.
SATAII port on motherboard, SATAII Samsung F3 Ecogreen 2TB 32MB cache HDD
SamsungF3EcoGreen5400RPM32MBCache.jpg
As you know these Samsung Ecogreen drives run at a fairly slow 5400RPM, so you can never expect amazing performance, but even so, it shows much faster sequential speed, and the 512K test was only beaten by your USB3 test which is unsurprising and shows that solidstate memory storage is faster overall.
SATAII port on motherboard, SATAII Samsung F3 1TB 7200RPM 32MB cache HDD
SamsungF17200RPM32MBCache.jpg
Interesting result this one. Its as expected faster than the Ecogreen drive in the sequential test buy a long way. But drops off a lot in the 512K test and then more so in smaller write tests. This is the system drive so maybe that did effect smaller sizes during the test.
SATAII port on motherboard, SATAII Seagate Barracuda 7200.10 16MB cache 500GB HDD
Seagate9800.97200RPM16MBCACHE.jpg
It's also worth mentioning that I also ran tests on this older Seagate Barracuda SATAII 500GB HDD and its speed was much lower, showing how these newer Samsung drives are really delivering better performance. Especially surprising how much better the results were for this Ecogreen drive considering the Seagate is 7200RPM. Just goes to show that RPM isn't the most important factor.
note: I also just realised I ran these tests with the virus checker enabled, which could effect it quite a lot. I will run it again next week when i have more time.
[Reply]
My drives are not empty so that could effect it again, especially for the smaller file tests. I think a virus checker would effect it as it will test every read/write to the drive for suspicious activity, slowing it down.
[Reply]