Bandwidth, latency, and 2D benchmarks
| Sisoft Sandra 2009 SP1 - float buffered | |
|---|---|
| Crucial 6GB DDR3-1,066MHz CL7 | Qimonda DDR3-1,066MHz CL7 3GB |
| 18.24 | 16.88 |
The larger-capacity Crucial set does well in Sisoft Sandra's float benchmark, producing a result which is almost eight per cent higher.
| Sciencemark 2 memory latency | |
|---|---|
| Crucial 6GB DDR3-1,066MHz CL7 | Qimonda DDR3-1,066MHz CL7 3GB |
| 40.12 | 40.52 |
Latency, though, is practically identical between the two.
| HEXUS.PiFast - calculation to 10m places | |
|---|---|
| Crucial 6GB DDR3-1,066MHz CL7 | Qimonda DDR3-1,066MHz CL7 3GB |
| 33.19 | 33.34 |
| wPrime multithreaded - 1024M calculation | |
|---|---|
| Crucial 6GB DDR3-1,066MHz CL7 | Qimonda DDR3-1,066MHz CL7 3GB |
| 290.55 | 290.85 |
Running two processor-intensive benchmarks - single-threaded HEXUS.PiFast and multi-core-threaded wPrime - fails to show any meaningful difference in performance between the two.
We would expect as much really, because the benchmarks are run in isolation and don't have a huge memory footprint.