We saw how PC Mark 2005's memory bandwidth benchmark mostly reflected the overclocking capabilities of each module set, so let's drive the point home with SiSoftware's Sandra memory benchmark.
Sandra's results again are mostly due to memory speed, though Super Talent's 920 MHz memory clock somehow outperformed Wintec's 930 MHz memory clock, which was likely due to timings other than the four we set manually.
Of course the real reason why overclockers should consider DDR3 is to remove any memory speed limitations that might impede CPU overclocking. Given the small difference in memory performance-per-clock, overclockers should consider the fastest memory that fits within their budget.
- The New Mainstream Standard?
- Tomorrow's Bandwidth...Today!
- Speed Vs. Latency: Myths And Facts
- 10 Kits For Your Consideration
- Kingston ValueRAM PC3-10600
- Kingston HyperX PC3-11000
- Mushkin Enhanced EM3-10666
- OCZ PC3-10666 Platinum Edition
- OCZ PC3-10666 ReaperX HPC Enhanced Bandwidth
- PDP Patriot Extreme Performance PC3-10666 Low Latency
- Super Talent PC3-10600 CL8
- Wintec AMPX PC3-10600
- SPD Timing Comparison
- Test Settings: Overclocking Comparison
- Overclocking Results
- Overclocking Results, Continued
- Boot Straps, I.e., Intel's "Wrench In The Works"
- Test Settings: Lowest Stable Latencies
- Test Settings: Lowest Stable Latencies, Continued
- Lowest Latency Test Results
- Lowest Latency Benchmark Results
- Audio Conversion
- Synthetics, Continued