|
|
In English
SuomeksiTesting and Use
All of the memory kits were tested using the same setup.
Hardware:
- Processor: Intel Core 2 E8500
- Motherboard: Foxconn Black Ops (Intel X48)
- Memory: Being tested...
- Video: Sapphire HD4850 Toxic
- Sound: Onboard HD Audio
- Storage: Western Digital 200 GB
- Power Supply: Xigmatek NRP-HC1201
- Chassis: Foxconn Black Ops Testbench
- Cooling: Corsair Nautilus
Software:
- Operating system: Windows Vista Ultimate x64 SP1
- Drivers: Catalyst 8.8
The memory kits were all run at their specified speeds after which I ran them at 1600 MHz, 1800 MHz and 2000 MHz with the tightest timings I found stable. If the kit wouldn't run at 2000 MHz I tested them at the highest speed I was able run them at. I used the 1:2 divider for the memory in every run, running the memories at the FSB speed. The CPU multiplier on the otherhand was always at 7, so you have to remember that when the memory speeds increase the CPU runs faster as well. This means that you really can only compare the results that have been run at the same memory clockspeed.
Note: Not every motherboard can or will run memory kits at frequencies this high. Officially most motherboards support memory frequencies up 1600 MHz, although not all will even go this far. It is also important to remember that the overclocking results that we got in our tests may or may not be replicated with other kits.
G.Skill F3-14400CL8D-2GBPI
This kit is specified at DDR3-1800 with 8-8-8-21 timings, which worked through out the testing without a hitch. At 1600 MHz I was able to tighten the settings to 6-7-7-18 with the stock voltages. Going even lower might be possible, but I didn't get them working stable enough without increasing to voltage too much.
After getting these two settings benched, I turned my attention to getting this kit to run at 2000 MHz, unfortunately only 1900M Hz was stable enough for the benchmarks even after loosening the timings a little bit.
CSX Diablo3-2000-2GB-Kit
The CSX Diablo kit had the tightest timings specified of any of these kits, but they come at a cost of a higher voltage. This kit actually has three different "max"-settings for three different chipsets. The first one is for the Intel X38 platform, and the kit is rated at 7-6-5-14 T1 at 1600 MHz with just 1.85 V. The second rated settings are for the Intel X48 platform and this time around the timings are 8-7-6-15 with the T1 command rate at 1866 MHz. The higher speeds come again with the cost of higher voltages, 1.9 V. The third and last setting is for the X48 and Nvidia nF790i chipsets. This setting has the memory running at 2000 MHz which may or may not run that easily with the X48-chipset. The nF790i-chipset should however be able to run the memory at this speed easier as the FSB doesn't have to be as high as the memory speeds. The timings for the 2000 MHz speed are blazing fast at 8-7-6-19 and a command rate of T1.
As this kit had the recommended timings included for just about every speed, I ran them accordingly. I even tried to tighten the settings and to run the memory at higher speeds, but was unable to do so. Even adding more voltage didn't help.
Crucial Ballistix PC3-16000 CL9 EPP
As the second and last kit that's specified at DDR3-2000 MHz is the least radical looking of the roundup. The small heatsinks do fool many people, as this kit manages to run at 2000MHz at a lower voltage of 1.9 V when compared with the CSX kit. However the timings on this kit are considerably looser, being rated at 2000 MHz with 9-9-9-28 timings. Atleast some Crucial kits however are based on the Micron D9GTS chips that should do better.
Getting this kit to run at the specified speed went as easily as I could hope. The first limiting factor in getting the memory to run at 2000 MHz was to get the settings for the motherboard in order, but as soon as those were sorted things ran smoothly. After running the kit at the specified 2000 MHz speed and timings I tried to tighten the settings without too much luck. More voltage might have done the trick though. At 1800 MHz the timings were tightened to 8-8-8-28 which ran fine with the same voltage. When lowering the speed to 1600MHz this kit actually managed to hit the same timings as the CSX kit.
 | Pages: 1 2 3 4 5 6 7 8 9 10 |  |
Content in english!
Sisältö suomeksi!
