Page 2: Stability, Testing, Conclusion
Stability:
I always have a section on how ASE Labs handles stability and this review is no different. I shall do this for every RAM review that is done. At ASE Labs, RAM is only stable if it passes Memtest86+ and a plethora of memory intensive applications for a few hours like mprime and doing continuous compiling of specific programs. Just because the memory POSTs and runs doesn't mean that it is stable. Linux users may find memory errors by having errors in compiling (very common) to errors in tar as well as segfaults. Windows users can expect program faults and BSODs. Memory stability is very important. When we say that a specific specification of RAM is stable, we mean it!
The Kingston KHX14400D3K2/2GX performed without incident at its rated timings and voltage. It passed all tests at 8-8-8-24 (1T) with 1.9V running at 1800MHz set in the BIOS. The Asus P5E3 Deluxe provided the platform for the testing. After the initial stability checks and the more involved compilation testing were completed, the XMP settings of the modules were tested.
Testing:
I have already reviewed similar kits to this and based on what I have shown repeatedly, benchmarking RAM is an exercise in futility due to the myriad of differences in timings and speed. If you have two kits that are rated the same, there should be a very little difference (if any) in the perceived speed. RAM testing comes down to stability, features, and overclocking. We will skip the benchmarks this time around.
Using the Asus P5E3 Deluxe, the board was set to use XMP settings from the modules. The XMP settings incorrectly registered Profile #1 as 8-8-8-21-1T-1.9V-1800MHz. I don't know why the RAM had the timings a bit off, but they ran perfectly stable at that speed as well. I guess I'm the lucky one that got a faster kit or something. The rest of the settings were set to auto so the board could handle everything as would any person using XMP. A Core 2 Quad Q6600 was used for this XMP test. The Asus P5E3 Deluxe refused to POST when everything was set to auto. I was a bit fluster due to the lack of experience with XMP so I asked a few fellow tech sites for additional information and got in contact with Kingston and Asus in the process. It turns out that Asus itself had a problem running a Q6600 with XMP at 1800MHz. If I would guess, it is because the board is not setting the correct multiplier for the speed. They mentioned that the Q6600 could handle 1800MHz which I find hard to believe when running at 6x.
My Q6600 goes past 3GHz. When I forced the CPU multiplier to 6x, and left all the other settings on auto, the board properly POSTed and proceeded to boot into Linux. I put the blame solely on Asus for this XMP problem. As I understand it, XMP settings should be handle by the board in a way that the CPU is minimally overclocked. The correct way would be to drop the multiplier to the lowest it can to run at 1800MHz which the board didn't bother doing. Kingston's modules had the correct information, so they can't be blamed for this. This is something to think about and is another reason why I usually don't bother running XMP settings.
I wouldn't assume that you would get modules with XMP settings that are 8-8-8-21-1T-1.9V-1800MHz. If you purchased these modules, you should get the rated speed. Like I said, I must be special.
Overclocking:
I decided that overclocking this kit should focus on speed over timings so leaving the voltage at the stable 1.9V, the highest speed that was totally stable was 1880MHz which is a fairly good increase for modules that are already running at high speeds. 1.9V is pretty high for DDR3 already so putting more voltage to get a few additional MHz seemed a bit lacking to me. I would be surprised if most boards could even handle these speeds. Needless to say, this kit will give you more than enough headroom for your CPU to handle.
Conclusion:
The KHX14400D2K2/2GX kit is going for about $240 at the time of writing. This is an impressive drop since a few months ago when kits like this were commanding about double that. DDR3 is really coming to the price point when the true laid back enthusiast can try it out without breaking the bank. Even though I couldn't get XMP working properly on the Asus P5E3 Deluxe, it was not the fault of Kingston. Even so, I rarely use XMP settings or any other "easy" tweaks.
Kingston has delivered a worthy upgrade to the original »KHX14400 kit that did not include XMP support. Not only does it have better features, but it performed better as well. If your system can handle 1800MHz RAM, this kits delivers performance you would expect from a company like Kingston. The price is pretty much right at this point in time.
I'd like to thank Dave from Kingston for the continued support.
I always have a section on how ASE Labs handles stability and this review is no different. I shall do this for every RAM review that is done. At ASE Labs, RAM is only stable if it passes Memtest86+ and a plethora of memory intensive applications for a few hours like mprime and doing continuous compiling of specific programs. Just because the memory POSTs and runs doesn't mean that it is stable. Linux users may find memory errors by having errors in compiling (very common) to errors in tar as well as segfaults. Windows users can expect program faults and BSODs. Memory stability is very important. When we say that a specific specification of RAM is stable, we mean it!
The Kingston KHX14400D3K2/2GX performed without incident at its rated timings and voltage. It passed all tests at 8-8-8-24 (1T) with 1.9V running at 1800MHz set in the BIOS. The Asus P5E3 Deluxe provided the platform for the testing. After the initial stability checks and the more involved compilation testing were completed, the XMP settings of the modules were tested.
Testing:
I have already reviewed similar kits to this and based on what I have shown repeatedly, benchmarking RAM is an exercise in futility due to the myriad of differences in timings and speed. If you have two kits that are rated the same, there should be a very little difference (if any) in the perceived speed. RAM testing comes down to stability, features, and overclocking. We will skip the benchmarks this time around.
Using the Asus P5E3 Deluxe, the board was set to use XMP settings from the modules. The XMP settings incorrectly registered Profile #1 as 8-8-8-21-1T-1.9V-1800MHz. I don't know why the RAM had the timings a bit off, but they ran perfectly stable at that speed as well. I guess I'm the lucky one that got a faster kit or something. The rest of the settings were set to auto so the board could handle everything as would any person using XMP. A Core 2 Quad Q6600 was used for this XMP test. The Asus P5E3 Deluxe refused to POST when everything was set to auto. I was a bit fluster due to the lack of experience with XMP so I asked a few fellow tech sites for additional information and got in contact with Kingston and Asus in the process. It turns out that Asus itself had a problem running a Q6600 with XMP at 1800MHz. If I would guess, it is because the board is not setting the correct multiplier for the speed. They mentioned that the Q6600 could handle 1800MHz which I find hard to believe when running at 6x.
My Q6600 goes past 3GHz. When I forced the CPU multiplier to 6x, and left all the other settings on auto, the board properly POSTed and proceeded to boot into Linux. I put the blame solely on Asus for this XMP problem. As I understand it, XMP settings should be handle by the board in a way that the CPU is minimally overclocked. The correct way would be to drop the multiplier to the lowest it can to run at 1800MHz which the board didn't bother doing. Kingston's modules had the correct information, so they can't be blamed for this. This is something to think about and is another reason why I usually don't bother running XMP settings.
I wouldn't assume that you would get modules with XMP settings that are 8-8-8-21-1T-1.9V-1800MHz. If you purchased these modules, you should get the rated speed. Like I said, I must be special.
Overclocking:
I decided that overclocking this kit should focus on speed over timings so leaving the voltage at the stable 1.9V, the highest speed that was totally stable was 1880MHz which is a fairly good increase for modules that are already running at high speeds. 1.9V is pretty high for DDR3 already so putting more voltage to get a few additional MHz seemed a bit lacking to me. I would be surprised if most boards could even handle these speeds. Needless to say, this kit will give you more than enough headroom for your CPU to handle.
Conclusion:
The KHX14400D2K2/2GX kit is going for about $240 at the time of writing. This is an impressive drop since a few months ago when kits like this were commanding about double that. DDR3 is really coming to the price point when the true laid back enthusiast can try it out without breaking the bank. Even though I couldn't get XMP working properly on the Asus P5E3 Deluxe, it was not the fault of Kingston. Even so, I rarely use XMP settings or any other "easy" tweaks.
Kingston has delivered a worthy upgrade to the original »KHX14400 kit that did not include XMP support. Not only does it have better features, but it performed better as well. If your system can handle 1800MHz RAM, this kits delivers performance you would expect from a company like Kingston. The price is pretty much right at this point in time.
I'd like to thank Dave from Kingston for the continued support.