Variations in apparent maximum measured throughput for SATA III and SATA II?

[lovecakeboy]

Hello all,

I've been noticing some interesting variations between people's results, where the speeds seem to be maxxing out the SATA link.

There are a bunch of results around 543 MB/s, but although they are often within 0.1 MB/s of each other within the same set of results, ie. on the same drive / motherboard combination, between systems the actual speed varies from about 542 to 544 or so. I've seen one at 546!

So, is this an actual speed difference being measured for different chipset / drive firmware combinations, or do we think that the accuracy of ReadSpeed's results are somehow dependent on the motherboard or other factors?

There is a similar variation around what looks like it might be the SATA II limit, around 273 MB/s.

 

[lovecakeboy]

Allow me to quote @PHoganDive's suggestions from this thread:

https://forums.grc.com/threads/results-from-a-couple-computers.349/post-3348

PHoganDive said:


SATA II = 300MB/sec, SATA III = 600MB/sec, but there is a certain amount of overhead involved, so maximum actual data speeds are about 90% of that, so subtract about 30MB/sec for SATA II and 60 MB/sec for SATA III.

 

[AlanD]

We saw during development that the accuracy figure varied between machines, although it is primarily driven by clock speed. A faster CPU will give a more accurate reading. There may be further variations introduced by differing chipset drivers and motherboards.

 

[ColbyBouma]

The highest result I've seen is 576.1: https://forums.grc.com/threads/readspeed-results-2-ssds-spinner.352/

 

[lovecakeboy]

Ooo, well spotted @Dagannoth!

Thanks for the inside info, @AlanD.