Articles: Memory
 

Bookmark and Share

(7) 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 ]

Testbed and Methods

We assembled the following system for our performance tests of Corsair TWIN2X1024-8000UL DDR2 overclocker memory:

  • Intel Pentium 4 520J CPU (2.8GHz, 1 MB L2, Prescott D0);
  • ASUS P5WD2 Premium mainboard (LGA775, Intel 955X);
  • Corsair TWIN2X1024-8000UL memory;
  • PowerColor RADEON X800 XT (PCI-E x16) graphics card;
  • Maxtor MaXLine III 250GB HDD (SATA150);
  • Microsoft Windows XP Professional SP2 OS.

The mainboard BIOS contained the following settings:

  • Performance Mode: Auto;
  • HyperPath3: Enabled.

This is exactly the same system as the one we used to test other overclocker memory modules. Therefore, we will be able to compare the results we will obtain today with those numbers we got earlier during our previous test session.

Performance

We tested the stability of the memory subsystem in two stages. First we ran the Memtest86+ version 1.60 to see if there were no failures during the test. Then we double checked the result in Windows XP with the S&M utility version 1.7.3 and Prime95 version 24.13. This “double-testing” allowed us to make sure that the results were credible.

Since Corsair TWIN2X1024-8000UL works at the nominal 2.2V, we carried out all our tests at this particular voltage setting, even though it seems much higher than the official JEDEC approved voltage and is not supported by all mainboards.

Our test results demonstrated that Corsair TWIN2X1024-8000UL memory modules can be regarded as highly universal. They worked just fine at 667MHz frequency with the minimal timing settings, and could easily reach 1GHz bar with higher timings. Is it surprising? Not at all! If you remember, we said exactly the same thing about all other overclocker DDR2 SDRAM based on Micron D9DQT memory chips. This is exactly the reason why all overclocker DDR2 SDRAM modules claimed to support 1GHz working frequency are so similar in action.

Nevertheless, we can still point out a few differences in the functioning of memory modules by different manufacturers. Firstly, different memory makers use different PCBs for their DIMMs, which can be better or worse protected against EMI. Secondly, the heat dissipation might be different. And thirdly, the manufacturers may have different criteria for memory chips selection. For example, Corsair offers not only their high-end TWIN2X1024-8000UL overclocker kit, but also slower overclocker modules and memory solutions for common mainstream users. This allows the company to make sure that only the best of the best DDR2 chips go into their TWIN2X1024-8000UL DIMMs.

As a result Corsair TWIN2X1024-8000UL memory modules perform very well with “average” timings of 4-4-4-12. In this workmode the memory remains stable up to 964GHz frequency. And if we increase CAS# Latency to 5, it will go beyond 1GHz. So, we have every right to state that the practical features of Corsair TWIN2X1024-8000UL not only correspond to the claimed specification, but even boast some ”reserved potential”.

If the timings are set to less aggressive 5-5-5-15, the potential of Corsair memory modules doesn’t grow that much. In this case they can only remain stable at 1028MHz maximum. Unfortunately, Corsair TWIN2X1024-8000UL failed to get beyond the psychologically important limit of 1066MHz, which would automatically qualify this memory for use in synchronous mode in systems based on top Pentium 4 Extreme Edition CPUs. Moreover, they do not manage to conquer this frequency barrier even if the voltage is set to 2.3V.

 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 ]

Discussion

Comments currently: 7
Discussion started: 09/22/05 05:21:44 AM
Latest comment: 05/10/06 11:38:58 AM

View comments

Add your Comment