SATA 3 Performance

[seekermeister]

For the benefit of the naysayers, who said that connecting a SATA 3 drive to a SATA 3 controller wouldn't have any real effect, I'm attaching screenshots. The first is with the drive connected to the SATA 2 controller of my old motherboard, and the second is with the same drive connected to the SATA 3 controller of the new motherboard.

I realize that those of you who are accustomed to SSD performance may consider this trivial, but I do not.

 

[pparks1]

From a naysayer viewpoint, I think there might be more going on that the simple port that you plug into. Most drives with HDTune look like the second chart, starting fast and dwindling down as you near 100%. However, your first test isn't showing that at all. It really looks like with your old mobo, something was really limiting hard drive speed to 80MB and no more. That isn't necessarily normal with SATA 3.0Gbps. I've had speeds of 120MB+ starting on my SATA 3.0Gbps on my GA-EP45-UD3R mobo.

I'd be interested to see a speed difference of that SAME hard drive between the SATA3 port on your NEW mobo and a SATA2 port on the NEW mobo. I'd bet you wouldn't see much difference there.

For reference, here is my 2TB Samsung 5400RPM storage drive on SATA 3.0Gbps

My 1TB caviar black in same machine;

And a 1TB Seagate 7200.12 running in my old AMD rig (AMD Athlon 64 X2 4400+) with SATA 3.0Gpbs ports

 

[seekermeister]

Sorry, no can do. All of the SATA ports are SATA 3 on this rig. The only way that I could do SATA 2 is to install a controller card that I have, but I'm not going to. I did test this same drive on that controller card, as well as directly connected to the motherboard of the old rig, and there was little, if any difference.

 

[pparks1]

I understand....cannot test what you don't have.

It's good to see you get a performance boost, I'm not sure why your old mobo was really holding you back that much. But just don't assume that everybody coming from a SATA 3.0Gbps to SATA 6.0Gbps is going to see this level of performance increase that you are seeing. The change in my case might only be marginal.

 

[GeneO]

There is something seriously wrong with your SATA 2 port. The plot should not be flat like that at 89 MB/s. You should see the variation in performance from one end of the platter to the other. SATA 2 can do about 260 MB/s and you should easily be able to perform at the rate your are seeing on the SATA 3 on the SATA 2 port.

Agree with pparks1

 

[seekermeister]

It may be that there was a problem with the old rig, other than just a difference in port types, but I just ran a couple of more benches on some SATA2 drives, and they also show a marked increase in performance from before.

 

[pparks1]

Perhaps the benchmarks take advantage of increases in the SATA 6.0Gbps interface. I'd be more interested to see what difference is noted in actual real world performance. It's there that I always say that it's pointless to upgrade your mobo and hard drives to simply get mechanical 6.0Gbps drives. The increase in performance simply won't justify the cost.

 

[seekermeister]

I'm not sure how one would go about accessing the "real world performance", unless you mean something like file transfers from one location to another. I was thinking of trying that with some video files, but I'm having a problem accessing them, because of a problem with my external storage that I posted about in another thread. However, I did copy/paste a 8+GB data file from one partition to another of the same drive, and it was a lot faster than it would have been before. My statement is somewhat subjective, but I don't know how to be more precise about it.

 

[pparks1]

For me, I real world test by doing a handful of things that I normally would and timing them. For example,
1). From power button press to login screen: in seconds
2). Time to copy a file from 1 folder to another folder on the same drive
3). Time to take a pile of files in a folder and compress them into a zip file
4). Time to take a video file and re-encode it with something like handbrake.
5). Time to rip a DVD onto the hard drive and then encode the video into another format.

Stuff like that generally. If it's faster, great. If it's about the same....then "who cares"?

 

[GeneO]

It may be that there was a problem with the old rig, other than just a difference in port types, but I just ran a couple of more benches on some SATA2 drives, and they also show a marked increase in performance from before.

It looks to me like your SATA 2 port was running at SATA 1 speed.

 

[pparks1]

^ Yeah, I tend to agree.

 

[seekermeister]

For me, I real world test by doing a handful of things that I normally would and timing them. For example,
1). From power button press to login screen: in seconds
2). Time to copy a file from 1 folder to another folder on the same drive
3). Time to take a pile of files in a folder and compress them into a zip file
4). Time to take a video file and re-encode it with something like handbrake.
5). Time to rip a DVD onto the hard drive and then encode the video into another format.

Stuff like that generally. If it's faster, great. If it's about the same....then "who cares"?

Without actually timing all of these, I have done the first two, and I can say that it is quite noticeably faster. Even initiating a program like Opera, took about a minute previously, and it is almost instantaneous now.

The difference may be due to other factors, because there is a lot more different to this rig than just the SATA 3 controller. Whatever the reason, I'm not complaining.

 

[seekermeister]

As a matter of curiosity, I just encoded the movie The Westerner (1:39:40) with HandBrake, and it finished in ~ 35 minutes. Previously, I would have expected my old rig to have taken perhaps 3 -4 hours to do the job. However, I do not believe that the difference is due primarily to the SATA 3 performance, but more with the new processor. This is the primary reason that I decided to upgrade, because it was taking forever to update my movie library. So, to this extent, I'm quite pleased.

 

[Sardonicus]

Glad you noticed a speed increase. However, the Wd1502faex is a very fast drive and even has a 64Mb cache, so it's shocking to see the low HD Tune numbers you put up. The numbers should be closer to what a lot of us are getting today with our SATA III builds. That's the reason for the skepticism. Something was holding your drive back. Hopefully you didn't have a jumper on the drive, giving you SATA I speed. If you take a look at some of the HD Tune SATA II scores over the internet for that drive, you'll see what we mean.

Still.....glad you're now seeing the speed that you should have previously achieved.

 

[seekermeister]

There has never been a jumper on the drive.

 

[stormy13]

Three little words for what was likely holding the drive back; Nforce 4 chipset. You even have a quick and easy way to test seeing as your remaining NF4 board is virtually identical to the one you replaced, just connect that drive up and run the test on it there and see what kind of numbers it gets.

 

[seekermeister]

I understand your comment so far as nForce 4 chipset goes, but as far as I understand, this motherboard uses an AMD chipset, and I haven't read anything relating to NF4, so I don't understand how you consider it identical to my old motherboard, and since I have tested the drive on this motherboard, I'm not sure what test that you mean?

 

[stormy13]

Not the new board is the same as the new, the board you used to have was the same BF4 chipset as your second computer in your specs.

I assume that flatline result in your first post is the same drive only one the old NF4 computer. If it is put that drive in your other NF4 computer and run the same test on it and see what results you get.

 

[seekermeister]

I don't need to test it in the HTPC to know that it would get similar results. I don't know that it would flat line, but it would be much lower than it now is.

 

[tw33k]

The most important stat on those benches is the access time and it's better on the SATA II port. Funny results.