21-01-2005, 03:58 PM
|
#1 (permalink)
|
| Administrator
Join Date: Jul 22 2004 Location: Brisbane, Australia
Posts: 22,924
| [review] 2x256MB Corsair TwinX XMS3200C2 rev4.2 - Samsung TCC5 @ xtremeresources.com Introduction
Corsair is a well established company with a good reputation for making high performance memory, and is o*ne of the leading RAM manufacturers for the overclocking community. While waiting o*n my BH-5 memory order, I bought 3 pairs of 2 x 256Mb Corsair XMS TwinX512-3200C2 Rev. 4.2 the week before. To my surprise, they arrived manufactured with ICs that are practically unknown to most overclockers, Samsung TCC5. Like most overclockers I don't know much about these chips, and I've never read much about them around the web. I've tested them a little bit which is what encouraged me to produce a review about these fairly different ICs utilized o*n this ram. Testing Methodology I will run Memtest Test #5 for 20 loops for each frequency/timing combination that will test be tested. Normally, after 20 loops of Test #5 without errors, the RAM can be stated as 100% stable. Screens of CPU-Z will be shown first at the best frequency/timing combinations I could establish, in order to illustrate the different combinations in detail. Benchmark screens will be shown for each combination identified, so you can clearly see the differences in performance.The popular benchmarking utility SiSoft Sandra known as the Memory Bandwidth Benchmark, will be used also.. I will NEVER run Sandra buffered since the buffered results are very inaccurate. Sandra will always be set to unbuffered before each test I run. Next up will be Everest, a benchmarking program formerly known as Aida32. It's benchmarking Memory Read and Write speeds separately provide a nice view of the increase in memory speed. All tests are run 5 times and the average results are the chosen screenshot and reference. http://www.xtremeresources.com/modul...wcontent&id=11
Not sure that 20 loops of test #5 is enough but nice review
|
|
Last edited by eva2000; 21-01-2005 at 04:01 PM.
|
| |