Tech Support banner

Status
Not open for further replies.
1 - 8 of 8 Posts

·
Registered
Joined
·
61 Posts
Discussion Starter #1
Hi again all,
I have a Maxtor 120GB ATA/133 7200rpm HD connected as master. Also have a Maxtor 40GB ATA/100 5400rpm HD connected as slave. The slave is only used for photo storage. The machine seems to be running OK, but I have a question. Due to the difference in architecture and rpm between the two HDs, would the older, slower drive affect overall system performance? The system is running on a P4 3.06GHz processor w/1GB PC3200 RAM, Windows XP Home SP2.
 

·
Registered
Joined
·
25,116 Posts
Welcome to TSF:

I believe the IDE controller will adjust to the slowest drive on the cable / thats why its not a good idea to mix hard drives and dvd drives on the primary channel.

This is not gospel mind you / but until someone who is "for sure" contradicts me / I am fairly certain this is the case. I am further inclined to believe this spped adjustment as factual due to the fact I had a similar set-up and noticed a big speed adjustment when I ditched a ATA-66 drive that I used to keep for file storage. I doubt you would see any speed lost using a ATA 100 drive though / but the 5400 rpm's has got to cost you somewhere.

As cheap as drives are on ebay / a slow drive is hardly worth Coddling any more / you can buy a NEW western digital 80 gig 7200 rpm - 8meg cache drive for under $40.00 / why suffer any performance lag whatsoever for that ??????


regards

joe
 

·
Registered
Joined
·
227 Posts
the slowest drive on a chain takes the chain speed down, having a lower RPM drive isn't going to cost you any performance on your main drive, the only performance slow down here is by the fact you have a ATA-100 drive with an ATA133. What your'e basically doing is running An ATA 133 at 100 transfer rates. But the fact of the matter, the slowdown isn't going to be noticable. The fact of the matter, is most common Hard drives really only take advantage of about 50-70% of the transfer speed of the cables. The physical drive itself isn't built fast enough to take advantage of full ATA, unless you're running like 2 SATA II 10,000 RPM drives in Raid 0 (which i've ALWAYS wanted to do, but don't have the moola). Even at that, each single drive is still limited by the physical limitations of the drive itself.

Couple an ATA 100/133 drive with an ATA66 HD or DVR (yes some DVR's are ATA66/UDMA4, including mine), it is going to be more noticable. Couple a 100/133 drive with a 33 device, and it will be painfully noticable
 

·
Registered
Joined
·
7,747 Posts
Typical maximum sustain transfer rate of a ATA 100/133 is usually 50-60 MB/s. The only time you would see 100 or 133MB/s is through bursts when data is writing/reading from the cache of the HDD controller on the HDD. But as both have stated the maximum transfer speed of the channel is of the slowest on the channel.
 

·
Registered
Joined
·
1,188 Posts
@ EJayJay
You could try to test using HDTach on your HDDs with and without attaching the slower HDD. Let us know what you find :sayyes:
 

·
Registered
Joined
·
61 Posts
Discussion Starter #6
and the winner is.......

First, thanks for the HD Tach tip, please. As this was not a matter of dead and alive HDs I may not have executed the tests exactly as instructed. :smile: However, the results gave me enough data to make my desision. I am going to purchase a 80GB ATA/133 HD, not because of the results of these tests,
but because of what Joe said in a previous post in the following statement: As cheap as drives are on ebay / a slow drive is hardly worth Coddling any more / you can buy a NEW western digital 80 gig 7200 rpm - 8meg cache drive for under $40.00 / why suffer any performance lag whatsoever for that ??????
The test results showed only slightly marginal gains between having the ata/100 connected or disconnected. And test results have a CPU Utilization margin of error of (+/- 2%). For example, the largest difference I recorded was a gain of .5ms in Random Access time and .2mb Average Read. Burst Speed differed only by .1mb/s and Sequential Read speed was consistant except for a couple of spikes showing up on the graph in differant times. These variations in speed are rather inconsequential to me. I really could not tell the slightest bit of difference in overall system performance.
Of course the test for the individual 40gb ata/100 average read time was much slower than that of the 120gb ata/133,(29.9mb/s to 48.0mb/s), but it made almost no difference in the Average Read speed of the ATA/133 no matter if the slave was connected or not as stated above! :4-dontkno
So what I gather from this is that reading a file from the slower drive will be slower because of THAT drive but will have no effect on the faster drive. So I would simply expect faster read speeds (as noted above) on an ata/133 slave drive but no other performance improvements of note on the master. Hope this makes some sense! -razz:
 

·
Registered
Joined
·
25,116 Posts
Ejay

I appreciate your taking the time to share your findings of the hard drive tach reports / I was surprised the impact on the IDE channel as a whole did not suffer more perfromance loss than that ??? I was expecting see to see a larger lag gap than that ~~ hmmm

thanks for the info

joe
 

·
Registered
Joined
·
227 Posts
I'm not suprised it did not suffer performance loss. Keep in mind, ATA 100/133 only uses a little over half of the transfer speed of the cable, so, transferring 133 at 100 rates, will not make much difference, after all, the maximum transfer rate you can peak is, as stated above by joe, 50-60 MBPS
 
1 - 8 of 8 Posts
Status
Not open for further replies.
Top