Synthetic Benchmarks
First, we take a brief look at Futuremark’s popular synthetic benchmarks, 3DMark05 and 3DMark06. We run both at their default settings.
3DMark05 was released in September 2004, before dual-core desktop processors were launched, and just weeks prior to the Athlon 64 4000+’s release. So this was a top-of-the-line gaming processor at the time.
3DMark05 does not support multi-threading in the overall performance tests and as a result we see the higher clock speeds and larger L2 cache of the A64 4000+ allowing it to beat out the X2 4200+. The high-clocked X2 5600+, on the other hand, leaves both behind by more than 2,000 points. Unlike in its holistic tests, 3DMark05 does have a multi-threaded CPU test, and we see the X2 4200+ surpass the higher-clocked A64 4000+ in the CPU score.
3DMark06 was released in January 2006, which was over six months after Intel and AMD had dual-core CPUs on the market. Multi-threading support was now implemented in not only the CPU tests, but in the overall tests as well.
Looking at 3DMark06, we see very different results, as this time the X2 4200+ more than makes up for its lower clock speeds and pulls far ahead of the single-core A64 4000+. As a matter of fact, it is able to pull halfway up to the X2 5600+. In the CPU tests we see the A64 4000+ far behind the dual cores and unable to reach a four digit score.
Now that we see a growing advantage for the dual-core CPUs in these Futuremark synthetic benchmarks, it is time to move onto some games to see if these indications of future performance equate to better performance in games.