difference between ATA storage drivers and DMA

[chorro]

does anyone know what the difference is in performance between using the Intel ATA storage driver/accelerator versus just using and enabling DMA in Win2k? I have the latest Intel ATA storage drivers installed and have not noticed any performance increase. My 40gb harddrive is rated at ATA100 and my 12g is rated at ATA66.

------------------
P3 700
Tyan S2060 815E
384mb PC100
40gb 5400 Maxtor
Geforce2 MX400 64mb
SB Live 5.1 Surround
48x Lite-on cdrom
24x10x40 Lite-on cdburner
Win2k Professional

[ecrosson]

If you have two drives on a system, and they are not the same speed, or rating, then the system will automatically default to the lower of the two drives.
In this case you will not see any difference because they both are performing like ATA66 devices.
If you match both drives to be the same stats, or remove the lower rated one from your system, then you should see the performance difference.
This phenomena is much like what happens with RAM when PC133, and PC100 are used in the same system at the same time. It will cause all the RAM to run at the PC100 speed overall.

Hope this information helps out some.

------------------
Tech talk to me !!! PLEASE???

[chorro]

so you can also mix PC133 and PC100 ram together? I thought you couldn't do that but you could use PC133 in a PC100 mobo since it's backwards compatible...

------------------
P3 700
Tyan S2060 815E
384mb PC100
40gb 5400 Maxtor
Geforce2 MX400 64mb
SB Live 5.1 Surround
48x Lite-on cdrom
24x10x40 Lite-on cdburner
Win2k Professional

[NoBoB]

Originally posted by ecrosson:
If you have two drives on a system, and they are not the same speed, or rating, then the system will automatically default to the lower of the two drives.

That is completely untrue.

Modern busmastering IDE controllers have long had the ability to run each device at the fastest possible speed. The only time a "slowing" occurs is when a slow device and a fast device are actively exchanging data, but that would be true no matter where the devices were connected.

It is likely that the reason for not noticing a performance difference is that the drive is not capable of going any faster. Virtually no drive available is capable of exceeding the ATA/66 design speed for sustained transfers, quite a few don't even approach ATA/33. They'll do burst transfers pretty quickly, but after the drive buffer is exhausted, the sustained speed comes into play. Today's best drives barely reach 50MB/sec sustained, so the argument of ATA/66 vs. ATA/100 is pretty much moot for sustained data transfer.

[ecrosson]

If I gave erroneous information I do apologize. However, I must say that if you do run two drives of differing speeds, then the maximum transfer rate you will achieve for them at one time is the rate of the lower speed drive, this I do know to be true, and it is also stated in different words by NoBob above. In this respect it is quite similar to them defaulting to the lowest speeds as I mentioned.

With regard to the RAM, yes, they are backward compatible, but you would not want to run a P133 BUS board with P100 RAM. You can run a P100 BUS board with either P100, or P133 RAM and it should/would be just fine.

I do not use the more technical terminology in most cases, so did not go into BUS mastering details, and perhaps in this case I should have done so.

I'm sorry chorro, if I have caused you any hardship, or confusion. We are all just trying to help in the best ways we are able.

(Hope this is of benefit).

------------------
Tech talk to me !!! PLEASE???