Navigate: Advice Beginners BIOS Guide CPUs Links Mainboards Memory Network Storage Video/Sound Cards Contact Forum SiteMap Sponsors WebNews Home	.	.
Prices: CPU Intel P4 2.4C-800 P4 2.6C-800 P4 2.8C-800 P4 3.0-800 P4 3.2-800 AMD AthlonXP XP 1700+ XP 2000+ XP 2400+ XP 2500+ XP 2700+ XP 3000+ XP 3200+ Athlon64 Athlon64 3200+ Athlon64 FX-51 Opteron Opteron 240 Opteron 242 Opteron 244 Opteron 246 Memory Corsair Crucial Kingston Mushkin OCZ Search Prices:

IDE Performance

Storage media are the slowest part of any PC at this moment and while they are playing a subordinate role in gaming, they are the bottleneck in a variety of applications like video editing or real-time scientific analysis using imaging techniques that need to store an immense flood of data in real time.

"Insufficiencies" of the PCI / ATA interface as implemented in the VIA chipsets have brought this aspect of the currently best selling core logic for the SocketA platform under a cross fire, we'll reserve our comments on this for a later time but the entire scenario stresses the fact that every chain is only as strong as its weakest link. In this case, we are talking about system performance and IDE transfer rates.

HDTach

HDTach is a free download for Win9x systems however does require registration / purchase of a license for Win NT-based operating systems. One caveat about HDTach is that its extended use can lead to drive corruption, in fact, I would guestimate that at least 80 % of all failures even of the IBM GXP75 series here occurred after using HDTach. There is also some controversy about the read-outs of HDTach which directly relates to the just said. To go back to the very begining:

• HDTach is meant as a tool to measure performance of a given HDD in a given system
• Most of the advanced options like "Write" performance require a blank drive without any data
• Any partition borders as well as sectors filled with data (e.g. operating system etc.) will show up as depressions in the sustained bandwidth graph.
• The access-time measurements are only accurate on a fresh, blank drive. With any drive containing data, we have found differences between 11.5 ns and 16 ns in repeated trials. In most cases, the access latencies will get shorter in repeated runs. We interpret this to show that the program apparently remembers which sectors can be used for the test and does not waste time trying to access drive sectors that contain other data.
• These effects are very easy to show, after a low level format, most drives will show a smooth line for sustained performance.
• Partitioning the drive will result in most cases in inverted spikes at the partition borders.
• Defragmenting a given partition will result in worse scores within the particular part of the drive than if a non-defragmented partition is monitored.
  

  Disregard some of the inverted spikes since they line up with partition borders. Otherwise, empty logical drives show up without jitters, wherever data are, the performance takes a nose dive. The important thing is, though, that the overall performance is right on target.

  Winstones

  We ran both Business Winstone2001 and Content Creation Winstone 2002. In both cases, the results were obtained with the integrated graphics that may have caused some performance hit because of the tapping of the integrated graphics into the overall memory bandwidth

  

  Business Winstone 2001 and Content Creation Winstone2002. Using the integrated graphics, the scores are not bad at all, for example, in Business Winstone 2001, the AMD Thoroughbred XP2200+ achieved 68.5 points on a KT333 system. Somewhat surprising came the performance drop in Content Creation with the V5. However, these results are plausible in view of the IDE (including burst rate) dependence of Winstone results in general and specifically Content Creation 2002. That means that any PCI data transfers, including the graphics adapter will impinge on the IDE performance and cause a performance hit of some kind, even though the one we see here is rather large.

  next page:    => Gaming Performance =>

  If you enjoyed reading this article and found it useful, please consider making a small donation to LostCircuits.
  Thank you!

  ---

  General disclaimer: This page only reflects the author's personal opinion and assumes no responsibility whatsoever regarding any of the contents or any damages that may occur explicitly or implicitly from reading the contents of this site. All names and trademarks mentioned in this review are the exclusive property of the respective parent companies.
  All contents of this site are protected by international copyright laws. Reproduction of the contents even in parts is not allowed