Articles: Memory

Bookmark and Share

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 ]

Testbed Configuration

We are going to test GeIL memory kits in an LGA 2011 system built on Asus Rampage IV Formula mainboard, which is currently one of the best platforms for CPU overclocking, according to our tests. Since overclocking memory modules are primarily purchased by enthusiasts, we also used Intel Core i7-3960K processor overclocked to 4.5 GHz. We selected this particular frequency because it may be reached in two different ways: with the base clock generator frequency set at 100 MHz (4.5 GHz = 45 x 100 MHz) and 125 MHz (4.5 GHz = 36 x 125 MHz). This gives us additional flexibility with memory frequencies. Besides the regular DDR3-1600, DDR3-1866 and DDR3-2133, we also worked in additional modes: DDR3-1666 and DDR3-2000.

As a result, the complete list of hardware and software components in our testbed looked as follows:

  • CPU: Core i7-3960X, overclocked to 4.5 GHz (Sandy Bridge-E, 6 cores, 1.5 MB L2 + 12 MB L3);
  • Mainboard: ASUS Rampage IV Formula (LGA2011, Intel X79 Express);
  • Memory:
    • Black Dragon PC3-12800 CL9 (4 x 4 GB, DDR3-1600, 9-9-9-28);
    • Enhance Corsa PC3-12800 CL9 (4 x 4 GB, DDR3-1600, 9-9-9-28);
    • EVO Corsa PC3-14900 CL10 (4 x 4 GB, DDR3-1867, 10-10-10-32);
    • EVO Corsa PC3-14900 CL9 (4 x 4 GB, DDR3-1867, 9-10-9-28);
    • EVO TWO PC3-14900 CL9 (4 x 4 GB, DDR3-1867, 9-10-9-28).
  • Graphics card: EVGA GeForce GTX 580 Classified 3 GB (03G-P3-1588-AR);
  • Drive: Crucial m4 256 GB (CT256M4SSD2);
  • Operating system: Microsoft Windows 7 SP1 Ultimate x64;
  • Drivers:
    • Intel Chipset Driver;
    • Intel Management Engine Driver;
    • Intel Rapid Storage Technology;
    • NVIDIA GeForce 285.62 Driver.


As we wrote in our special review deicated to the memory controller of Sandy Bridge-E processors, it is the clock rate that has the biggest effect on the performance of a quad-channel memory kit. Therefore, it is important for overclocker-friendly kits to be able to work at higher frequencies and thus ensure performance benefits.

 The GeIL kits we've collected for our today's test look promising in this respect. This brand is known for its reliability and, which is very important, each of these kits works at a voltage of 1.5 volts whereas LGA2011 platforms let you easily increase it to 1.65 volts without any problems. That's why we have high hopes of overclocking these products above their default 1600 and 1867 MHz.

Here is our testing algorithm:

  1. Memory voltage is increased from the default 1.5 to 1.65 volts;
  2. Failsafe timings of 11-13-13-31 are selected to find out the highest frequency the memory kit is stable at;
  3. Using the maximum frequency we’ve found, we then look for the most aggressive timings at which the memory kits is still stable.

The stability of the memory subsystem was verified by 10 runs of LinX 0.6.4 AVX Edition using the whole memory amount and by an additional 1-hour-long check with Memtest86+ v4.20.

GeIL Black Dragon PC3-12800 CL9

This Black Dragon kit looks splendid but doesn’t speed up well. We couldn’t even make it work at 1867 MHz, although its chips are marked in the same away as the chips on the memory modules rated for 1867 MHz.

Thus, we had to limit ourselves to DDR3-1666 mode and a 4% increase in frequency. We also gained something by replacing the default timings of 9-9-9-28-2T with 8-9-8-27-1T without any stability issues.

GeIL Enhance Corsa PC3-12800 CL9

This is one of the cheapest offers among the products covered in this review. However, the marking and some other details indicate that its chips and PCB are the same as in the more expensive EVO Corsa kits. This must be the reason why, despite its modest specs, the Enhance Corsa PC3-12800 CL9 kit can boast good overclocking potential.

After setting its voltage at 1.65 volts, we made this quad-channel 16GB memory kit stable in DDR3-2133 mode with rather good timings of 10-12-10-30-1T.

GeIL EVO Corsa PC3-14900 CL10

Rated for 1867 MHz, this memory kit could be overclocked as successfully as the Enhance Corsa PC3-12800 CL9. It was stable in DDR3-2133 mode with 10-12-10-30-1T timings but couldn't conquer a higher frequency.

GeIL EVO Corsa PC3-14900 CL9

Having the same PCB, chips and heatsinks as the EVO Corsa PC3-14900 CL10 and Enhance Corsa PC3-12800 CL9, this memory kit also notched 2133 MHz at 10-12-10-30-1T timings.

By the way, if you run your system at a base clock rate of 125 MHz, you can use any of these three memory kits (EVO Corsa PC3-14900 CL9, EVO Corsa PC3-14900 CL10 and Enhance Corsa PC3-12800 CL9) in DDR3-2000 mode with 9-11-10-30-1T timings.

GeIL EVO TWO PC3-14900 CL9

Identical to the EVO Corsa PC3-14900 CL9 in its specs, the quad-channel 16GB EVO TWO PC3-14900 CL9 kit behaved differently in our overclocking test. Despite its impressive heatsinks, it could not work in DDR3-2133 mode.

The highest frequency we achieved with this memory was 2000 MHz. The best timings in this mode were 9-11-10-30-1T.

The results of this test are summarized in the following table:

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 ]


Comments currently: 23
Discussion started: 03/30/12 06:41:17 PM
Latest comment: 09/20/16 01:24:44 PM

View comments

Add your Comment