Hi folks,
I've just finished and uploaded the review of the CSX Diablo DDR3-2000 2GB Dual Channel Memory Kit review.
You can read some quotes from the review here:
Originally Posted by BenchZone's ReviewCSX decided to contact us at... wait a sec! CSX who ?
CSX is the "short name" ( initials ) for "CompuStocX" a german hardware manufacturer.
CompuStocX isn't new in the game of RAM, they've been the biggest upgrade RAM supplier for Apple in Europe for the last 10 years or so, and they've just entered the enthuasiasts PC market with their Diablo memory series.
Back on topic now, if you're not familiar with DDR3 memory, it's the right time to get to know them, because Intel's new baby ( CPU ) is expected to be launched soon ( this November ) and you will have to use DDR3 with those CPUs, whether you like it or not, as the new "Nehalem" family of CPUs are now featuring a IMC ( Integrated Memory Controller ) [ the memory controller used to be in the chipset, the NorthBridge to be precise ], which is a DDR3 only controller.
Get ready to see some new Triple Channel memory kits soon as well, in 3GB and 6GB total capacities ( 3*1GB kits and 3*2GB kits ) from various memory manufacturers ( including CompuStocX more than likely ) as the new LGA1366 "Nehalem" processors feature a Triple Channel DDR3 memory controller instead of the 'typical' Dual Channel memory controller that has been the standard for both AMD and Intel desktop platforms since the Athlon XP and Pentium IV days.
DDR3 memory used to have a lot of fans ( the benchmarkers mainly ) and a lot of haters ( the average joe ) due to its high price, and no real benefit in real-life applications ( most of them ) [ there are some exceptions ].
Now DDR3 has reached new low price limits finally, and you can go out and get some nice 2GB DDR3-1600 kits at NewEgg and other retailers, priced at 120$ to 140$ ( 110 to 130 euros ) , and some nice 4GB kits at 230$ to 300$ ( 220 to 290 euros ).
The name "Diablo" reminds me ( and surely does the same for all the exotic cars fans ) of that devilish looks and blazingly fast 90's car, the Lamborghini Diablo SV, if CSX had that car in mind when they were looking for a name for their overclocking memory series, they've got to be fast.
Will the CSX Diablo DDR3-2000 2GB Dual Channel memory kit stand up to the history of the "Diablo" ? We'll find out soon in this review!Originally Posted by BenchZone's ReviewBelieve it or not, the testing for this review lasted 8 days. I'm talking about hours and hours of testing per day.
As you can see from the Test System information page I've tested this memory kit on 5 different motherboards, including the sort of frustrating Striker II Extreme and the strap-sensitive P5E64 WS Evolution.During these 8 days I've run about 1500 loops of the 5th test from memtest86+ v2.01. In case you're wondering that's more than enough to "burn" your LCD or plasma monitor with the memtest86+ interface :-p
I think I should get back on topic now. Oh well, here's how I tested the memory's performance for this review:
I could've used the Asus Striker II Extreme nFORCE 790i Ultra SLI motherboard to benchmark this kit at various timings-frequencies sets whilst keeping the FSB & CPU frequencies intact. Unfortunately the Striker II Extreme ( and the other 790i Ultra SLI motherboards that I gave a quick try-out ) couldn't overclock the RAM well at 6-5-5-16 timings, to be precise, the maximum stable frequency with those timings on the Striker II Extreme was DDR3-1600 ( 800MHz ) while on the P5E64 WS Evolution they stopped at DDR3-1713 ( 856.5MHz ).
So I turned to the board that gave me the best results ( well, on most of them at least ), the Asus P5E64 WS Evolution.
The P5E64 WS Evolution is using the latest high-end LGA775 chipset from Intel, the X48, which is a brilliant chipset to say the least.
Unfortunately, like every Intel LGA775 chipset you can't set the memory frequency "independently" from the FSB. There's are some different dividers ( FSB/DRAM Ratios, or Memory Multipliers like I like to call them ) that you can use to adjust the frequency of your RAM accordingly to the FSB though.
The 3/5 FSB/DRAM ratio for example, or the 5:3 as I call it, allows you to run your memory at DDR3-1110 ( 550MHz ) with your FSB at 333MHz.
There are various dividers, 3/5, 2/3, 1/2, etc. To calculate the resulting memory frequency you need to do the following...
...when I finished the tests another thing popped in my head:
Ok, you have some numbers Bill, but how are you going to present this review since you only have this memory kit to test ?
Our DDR3 Memory Round-up ( done in February ) can't be used for benchmarking comparisons since the testing platform is different.
We can use the maximum stable clocks of each memory kit tested back then, but not the numbers from the benchmarks.
So I came up with an idea.
I tested the CSX Diablo DDR3 kit at three different timings-frequency sets, whilst ( trying to ) keeping the CPU frequency the same, to check if there's any difference in performance between those configurations.
I did my best, and the results were pretty good I must admit. The two configurations are really close to each other, 3712MHz and 3745MHz, while the third configuration slipped to 3800MHz unfortunately.Read the full review now!Originally Posted by BenchZone's ReviewI know that I'm supposed to give you my last thoughts on the CSX Diablo DDR3-2000 2GB Dual Channel Memory Kit right now, but please excuse me for a sec, need to catch up with the garbage truck outside, I need to toss a memory kit in!
Nah, I'm just kidding, this kit is too good to get dumped in the trash.
No matter how hard I try, I can't find a bad thing to say about this memory kit.
It's reasonably priced ( actually carries a very good price tag ). I tested it on 5 different motherboards and didn't face any compatibility issues. Overclocking the kit was a great experience and the kit exceeded my expectations without breaking a sweat. It comes with a lifetime warranty. CSX's customer support look ok as well.
How the h*ll am I supposed to say something bad about it ?
Oh, there might be something bad... maybe not.
The heatspreaders didn't get any warm during our tests, they were just a tad higher than the room temp. This could be a good indication or a bad one. This means that... those heatspreaders are damn good or... plainly suck. No matter what, good or bad, the memory was stable and worked flawlessly overclocked or not, so... who cares if the heatspreaders are good or bad.
Just out of curiosity I will try to slip away from work for a second and place some thermal probes inside them and check the temperatures :-p
Be advised that during overclocking, if you're going to use voltages higher than 1.8V you REALLY NEED to use a fan to cool the RAM down to keep it stable and healthy ( unless you're fine with RMAing your RAM four times each month, just kidding, that's not so nice ).
I'm always using the Corsair Dominator AirFlow RAM Cooler that I received with my DDR2 Corsair Dominator memory kit back in 2007.
If you're a DIY guy, you can use a 80mm fan or a 120mm fan. And you don't need a high RPM/CFM ( AirFlow ) fan, you just need a nice and silent fan to move the air around.
Bookmarks