OCZ ReaperX HPC 4GB PC2-6400

[phoronix]

Phoronix: OCZ ReaperX HPC 4GB PC2-6400

Last July we had looked at the Reaper HPC DDR2 memory from OCZ Technology, which not only was clocked well for its time with a memory frequency of DDR2-1066MHz, supported NVIDIA's EPP (Enhanced Performance Profiles), and boasted OCZ's admirable lifetime warranty with EVP (Extended Voltage Protection), but it was the first model to bring forth their next-generation patent-pending memory heatspreaders. The Reaper HPC (Heat Pipe Conduit) cooling solution is made up of normal heatspreaders attached to the ICs, but bonded to the top of that is a single copper heatpipe in a loop that passes through an aluminum fin array. We were pleased by both the cooling and memory performance with the Reaper HPC, but OCZ Technology didn't stop there in their quest to deliver 110% customer satisfaction. OCZ has taken their Reaper HPC design to the next level by strapping on two copper heatpipes, ensuring that each memory IC is in direct contact with one of these thermo-conductive pipes, and greatly increasing the size of the aluminum fin array. This newest memory cooling solution is known as ReaperX HPC and can be found on the PC2-6400 ReaperX HPC EB 4GB dual channel kit, which we are reviewing today.

http://www.phoronix.com/vr.php?view=12014

[oblivious_maximus]

Ever since you guys cut open a few heatpipes that one time, I find myself hoping, every time I read a review(at any site) with a heatpipe involved, that the reviewer will end the review by cutting open the heatpipe(s). Whenever it doesn't happen (which is every time), I feel a little more empty inside, and I wonder to myself whether that heatpipe I just read about is empty inside too.

[n0nsense]

I own this kit for few month.
I have ASUS P5N32-E SLI M/B, and as part of experiment, i had no problem to install 4 sticks or this two side by side on separate channels.
as for performance test, i did it with compilation.
with the old G.Skill 2x1GB 800MHz 4-4-3-5 memory, it took ~6minutes to compile firefox. with this beauty, i took only 4.
On the other hand if you want to see extensive memory usage, it will be better to compile some C++ code like boost or rb_libtorrent libraries.
BTW, compilation test will tell us much more of the performance then ETQW fps.

[Ex-Cyber]

That LAME result looks a little odd. How does a module running at 800MHz/5-5-5-15 outperform a module running at 1066MHz/5-5-5-15?

[n0nsense]

That LAME result looks a little odd. How does a module running at 800MHz/5-5-5-15 outperform a module running at 1066MHz/5-5-5-15?

It probably was 800MHz/4-4-3-15 1T.
It will definitely beat 1066 5-5-5-15 2T

[3xch4ng3]

Nice review, I have this exact kit but haven't had a chance to overclock it at all. Plus, my board isn't the best in the world, so I'm most likely going to wait till i get a new one.