Corsair TWINX2048-4400PRO 2GB Memory Kit Review

Memory by KeithSuppe @ 2006-06-18

What can possibly be said about Corsair which hasn?t been said before. Their products are known and respected worldwide, as is their reputation for producing the highest quality memory. Their memory is the most consistent ever tested with every product to date either meeting or exceeding specifications. Today we test a product from their XMS Series, in particular the TWINX2048-4400PRO kit.

  • prev
  • next

Bandwidth / Latency Benchmarks

The many Faces of DFI


Madshrimps (c)


If I were asked which motherboard BIOS featured the most extensive set of memory settings, that answer would be the DFI LANParty UT NF4. The settings which are so numerous as to appear superfluous, they’re not. Changing any one of these can result in a no-post or break that MHz barrier once unattainable. My research of these timings nearly three years ago left me with a newfound respect for manufacture’s SPD timings. While it's prima facie the memory maker’s SPD will produce the most stable environment, many believe these are often too conservative. This is changing as the number of Overclockers and PC-Enthusiasts grows steadily.

Conservatism aside there's only so much which can be done with certain DRAM IC's and who knows better the idiosyncrasies of these chips then the manufacturer whom purchased them in bulk. The bottom line with Corsair is that they usually buy top shelf or "top bin" ICs. This may not always give you more to work with in so far as a performance ceiling; however, what it does give you is a much more reliable, stable platform. I did push the memory seeking the tightest timings at the fastest speed. These particular IC's react adversely to voltages above 2.85V ~ 2.90V and suggest the "tweakers" out there try lower rather then higher voltages when overclocking.

Below I've listed the combination of CPU multiplier/FSB speed for each series of benchmarks. I've also included highest speed reached at DDR580 (10x290FSB=2.9GHz) this was stable through 90% of the benchmarks; however I didn't include it in the charts for that reason.

Benchmarks

As indicated on the previous page in Thumbnails, benchmarks were run under the following system settings:

- XMS 4400 / 11x200FSB / DDR400 / CL3-4-4-8
- 4000PT / 11x200FSB / DDR400 / CL3-4-4-8
- XMS 4400 / 11x240FSB / DDR480 / CL2.5-3-3-6
- XMS 4400 / 11x240FSB / DDR480 / CL3-4-4-8
- XMS 4400 / 10x250FSB / DDR500 / CL3-4-4-8
- 4000PT / 10x250FSB / DDR500 / CL3-4-4-8
- XMS 4400 / 9x275FSB / DDR550 / CL3-4-4-8
- 4000PT / 9x275FSB / DDR550 / CL3-4-4-8
- XMS 4400 / 9x285FSB / DDR570 / CL3-4-4-8
- XMS 4400 / 10x290FSB / DDR580 / CL3-4-4-8


Madshrimps (c)


Madshrimps (c)


Madshrimps (c)


Madshrimps (c)


Madshrimps (c)


Next we move onto Synthetic 3DMarks and Far Cry and DOOM 3 ->
  • prev
  • next