Verify SATA 3.0Gbps operation

[racolvin]

Hi folks...

There isn't anything in my BIOS which tells me that my new drives are operating in either 1.5Gbps mode or in 3.0Gbps mode. There's nothing I can see in Device Manager that indicates this either...

Does anyone know of a way to verify this sort of thing?

R

 

[tcsenter]

Provided the SATA controller supports 3Gbps, you can use HD Tach as a quick-and-dirty way to measure the burst speed (not sequential). Under Vista, you'll have to run HD Tach in XP compatibility mode.

As long as your single-disk burst speed is higher than the theoretical max for SATA 1.5Gbps (~150MB/sec), it is consistent with SATA 3Gbps.

 

[racolvin]

thanx, I'll give that a try when I get home

 

[DSF]

However, just because your burst speed is lower doesn't necessarily mean that you aren't running on SATA 3.0Gbps.

 

[racolvin]

oh? esplain please

 

[tcsenter]

Originally posted by: DSF
However, just because your burst speed is lower doesn't necessarily mean that you aren't running on SATA 3.0Gbps.

No, but it would surely indicate a problem somewhere, if even with the measurement tool (HD Tach). Burst speed measurement should be weighted to characterize performance of the interface, proper, while being relatively agnostic to the raw capability of the device residing behind it.

If your SATA 3Gbps drive + controller cannot manage burst speeds in excess of 150MB/sec, something ain't right.

 

[Old Hippie]

Here's the burst speeds on a few drives.

Not all SATA 300 drives give a burst speed above 150MB/s.

Edit. I can't get the expanded list, that shows all tested drives, to link. Just hit the link for "Show all Drives", and you'll see what I'm talking about.

 

[racolvin]

Interesting.

When I ran the HD TACH thing on my system last night, my WD 160GB SATA300 drive (don't have the exact spec here at work) hit a burst speed of 190. My WD 500GB SATA300 drive hit a burst of 221.

Wonder what the issue is ....

 

[Old Hippie]

Originally posted by: racolvin
Interesting.

When I ran the HD TACH thing on my system last night, my WD 160GB SATA300 drive (don't have the exact spec here at work) hit a burst speed of 190. My WD 500GB SATA300 drive hit a burst of 221.

Wonder what the issue is ....

What issue? Too fast for ya? :laugh:

Those are great specs. :thumbsup:

 

[racolvin]

lol .. well, based on the other link you provided my drives shouldn't be performing that well ;-)

but the other questions would be:
1) why aren't they hitting their potential max at all
2) why is the 160 so much slower than the 500? The 160 is my C: drive, the 500 is my D:

 

[Old Hippie]

lol .. well, based on the other link you provided my drives shouldn't be performing that well ;-)

You're right, your scores look a little......shall we say.....optimistic.

The drives are hitting their max. But they're not hitting the potential max of the MB's electronics.

No single mechanical HDD will.

These are burst speeds. They mean very little, if any, to real world performance. I always figured they were more dependent on the drive's cache anyway.

 

[tcsenter]

Originally posted by: Old Hippie
Not all SATA 300 drives give a burst speed above 150MB/s.

Some of those results are very peculiar if not suspect. Look at the two ATA/133 Hitachi T7K500's in the 180MB/sec range (surrounded by SATA/300 drives), while the two SATA/300 Hitachi T7K500's are suspiciously in the 80MB/sec range (surrounded by PATA drives):

Hitachi GST Deskstar T7K500
HDT725040VLAT80, ATA/133, 400GB, 8MB
189.50MB/sec

Hitachi GST Deskstar T7K500
HDT725050VLAT80, ATA/133, 500GB, 8MB
188.80MB/sec

Hitachi GST Deskstar T7K500
HDT725040VLA360, SATA/300, 400GB, 16MB
88.40MB/sec

Hitachi GST Deskstar T7K500
HDT725050VLA360, SATA/300, 500GB, 16MB
87.70MB/sec

Those results are impossible for ATA/133, so someone had to have mixed these drives up when keying in the results. Also notice that nearly all SATA/300 drives listed between 148MB/sec and 155MB/sec are NCQ enabled, while very few above that are NCQ enabled. In fact, there are a total of eight "SATA/300-NCQ" entries between 148MB and 155MB, but only three entries above 155MB.

This observation isn't surprising since we know NCQ is likely to negatively influence measurement of interface throughput, not actual interface throughput. Most of those drives would be scoring above their current levels with NCQ disabled.

But yes, it is true that measured SATA 3Gbps interface performance can be lower than 150MB/sec. My point is that, by definition, it shouldn't be. The interface can only operate at the data rates defined by the technology, no more and no less.

Typically, the reason we see such huge differences in measured interface performance have to do with coding decisions owing to benchmarking philosophy, which determine testing metrics. The interface tests of different utilities are more or less inclusive of other factors such as seek performance.

IMO, an interface test should be precisely that - an interface test. While usefulness or meaning would be extremely limited, it is supposed to be. That's why its called an "interface test", not an "interface plus some other non-interface things test".

The reason I recommended HD Tach is because its burst speed test is designed to be more specific to interface performance, proper. Thus, it is usually a reliable way to confirm the active operating mode of the interface. Of course, HD Tach results cannot be interpreted in the context of scores from any other utility, nor vice versa.

It would be valid to say "Here are some HD Tach results that shows your HD Tach score could be this or that". It is not valid to say "Here are some results from a completely different utility that shows your HD Tach score might be this or that."

 

[racolvin]

Well at this stage I'm confident that my drives are actually operating in the 3Gbps mode, which is all I was really trying to find out originally.

Why they have some performance discrepancies as displayed by my running of HD Tach I don't know but I'm willing to bet that there isn't anything I could do about it even if I did know

So I'll just go with the idea that so far I've been happy with their performance and leave it alone

 

[Old Hippie]

So from your post, I can get two valid assumptions...

1) Tom's hardware is wrong in their testing
2) HDDs designed for ATA 133 can not surpass SATA 150 specs

Is that a valid assumption, tcsenter?

 

[Old Hippie]

Well at this stage I'm confident that my drives are actually operating in the 3Gbps mode, which is all I was really trying to find out originally.

'Tis true, You are open to the SATA 3 specs......whether we can validate it or not. :laugh:

 

[tcsenter]

Originally posted by: Old Hippie
1) Tom's hardware is wrong in their testing

No, Tom's Hardware inadvertently mixed up the Hitachi SATA and PATA drives that I mentioned. Failing to disclose the test configuration is not good form, either.

2) HDDs designed for ATA 133 can not surpass SATA 150 specs

No, the ATA/133 interface is incapable of exceeding 133MB/sec.

 

[Old Hippie]

Originally posted by: tcsenter

Originally posted by: Old Hippie
1) Tom's hardware is wrong in their testing

No, Tom's Hardware inadvertently mixed up the Hitachi SATA and PATA drives that I mentioned. Failing to disclose the test configuration is not good form, either.

2) HDDs designed for ATA 133 can not surpass SATA 150 specs

No, the ATA/133 interface is incapable of exceeding 133MB/sec.

Ok.

Whatever you say.

I see now that everybody is wrong......except you.

 

[tcsenter]

Originally posted by: Old Hippie
Whatever you say.

I see now that everybody is wrong......except you.

Everybody as in...? And wrong about...?

Parting sarcasm is a poor substitute for substance.

 

[Old Hippie]

Parting sarcasm is a poor substitute for substance.

Kinda like all those links you posted to qualify your answers? :roll:

 

[tcsenter]

What are you confused about that you need links to explain?

The maximum theoretical performance of the ATA/133 interface is 133MB/sec. There is no such thing as maximum theoretical performance plus more.