OCZ PC2-8500 SLI Ready Edition

Take the risk out of knowing your sticks will hit the mark.

It’s interesting all the cross branding that’s going on at the moment in the industry. Case in point: These new sticks from OCZ. We would have loved to have seen exactly how the deal was brokered to so blatantly plaster NVIDIA’s logo all over the product in question. On the upside, at least we can be 100 percent sure the sticks are EPP-capable.

There’s also a pair of ATI sticks out there with the CrossFire brand attached to it – although Lord knows what real benefit this has.

The black heat spreader is somewhat attractive, even if the perforated look is the norm for OCZ these days, and the 8500 rating means it clocks in at 1066MHz and does so at the 5-5-5-15 level, so it’s no slouch on the speed front even at its defaults.

Putting both 1GB sticks into dual channel in our EVGA 680i SLI board netted us 7457MB/s memory read, 4869MB/s memory write, 5424MB/s memory copy and 59.1ns memory latency under Everest Ultimate Edition 3.5.

While the sticks wouldn’t run 1066MHz at 1T or with tighter timings and remain Orthos stable, the MHz themselves certainly did go up, right to the 1180MHz mark when using the FSB with a 1:1 ratio, giving 8114MB/s read, 5372MB/s write, 6031MB/s copy and 54.2ns latency. Unfortunately relaxing the timings didn’t allow us to hit 1200MHz stable. It’s going to be interesting to see when we’ll hit the 1333MHz mark – place your bets on whether DDR2 or DDR3 will reach it first.

Bumping down to 800MHz using unlinked mode, we were able to hit 3-3-3-3 at 1T, giving 7937MB/s read, 4851MB/s write, 5477MB/s copy and 54.7ns latency – so better to go for the high speeds here rather than the low latency – or better yet, push your FSB and drop your CPU multiplier if you’re able, and stick with the low latency.

If you have to have everything branded by NVIDIA or just want some decently fast sticks at an almost reasonable price, then these 8500s are a pretty good deal. It’s not quite as insane as the Flex XLC series, but then we’re pretty sure nothing is.