Testing Methods
We strive to get the most accurate and apple-to-apple test bed we can configure in every new round of tests we run and this round brings us closer than ever to having that. Only the memory and few internal differences separate these platforms now. All tests are ran three times and balanced for accuracy, and the system was burnt in using Burn-In Mark Pro for three consecutive days. No restarts were required during the burn-in time and the systems remained stable throughout the entire process. We always use the latest available drivers with all the current OS updates installed prior to testing. We also reformat and reinitialize the system from scratch before beginning each round of tests. Each system was used daily for over ten days with notes taken on how well each system functioned during various operations.
Performance Summary
In many of the single-threaded applications the AMD FX-55 CPU just plain kicks ass and is faster than Intel's much higher clocked 670. This also gives AMD the needed headroom to in the next round to seriously kill Intel as they are only at 2.8GHz a whole GHz less than Intel. AMD can now let their technology tactics in CPU design work for them and still increase the clock speed of their CPUs, further increasing their effectiveness and ability. The real-world simulation test Sysmark, has in general always had AMD as the underdog in its scoring, AMD has turned that around in a complete 360 and is now showing scores better than Intel. Both systems were completely stable throughout the test sequences and in our day-to-day usage of them and subsequently were never turned off for over ten days. Neither CPU showed any problems regarding heat issues, although the Intel 670 CPU did run hotter than the AMD FX-57 by about ten degrees Celsius.
Performance Scores
