Comparing an Intel Core 2 Duo E8500 3.16 GHz desktop processor and an AMD Phenom X4 9350e 2.0 GHz high-efficiency quad core CPU for HTPCs or similar solutions isn’t really an apples to apples shootout. Given that, it’s important not to take this article as a general recommendation against purchasing a Phenom X4, or as a recommendation for the Core 2 Duo. Personal preferences aside, we found the comparison important, as we’re certainly not the only people wondering if a low-power quad core processor at average clock speeds might actually be a better choice than a fast dual core processor. Using an Intel low-power quad core for comparison would have been great, but such a product does not (yet) exist, hence the recommendation for supplying a low-power 65 W quad core processor go to AMD.
Fair or Unfair?
We could have used a faster Phenom X4 processor, as you can get either the 2.0 GHz 9350e or a 2.4 GHz Phenom X4 9750 for the same $200. The latter would have the Phenom win the majority of the benchmarks, and if you’re looking for performance, this is probably the best thing you can do—the quad core Core 2 processors are clearly more expensive. Yet we wanted to use a quad core that comes as close as possible to the currently best dual core processor. The efficiency that remains with the 9350e, and the 65 W power cap in particular, were lost if we had used a mainstream Phenom.
Core 2 Duo Wins
The Core 2 Duo Wolfdale at 3.16 GHz proved that its architecture is capable of competing with a quad core processor that runs average clock speeds, and it proved that it offers far better power efficiency. The Core 2 Duo matched or surpassed the performance of the Phenom X4 9350e, but this is the time to stop and consider another factor. At this point it would be unfair to declare Core 2 Duo the undisputed winner, as the test methodology shows a constraint that isn’t AMD’s fault: the Phenom X4 9350e and other quad core processors are handicapped by the software landscape, which still isn’t really optimized for four or even more cores. Some applications are, but apart from that, it’s the synthetic benchmarks that provide proof of quad cores actually being faster. Still, that advantage does not translate into our everyday lives.
Software Reworks Required!
The example of WinRAR (thread-optimized) versus WinZIP (single threaded compression process) makes very obvious which software vendors do and do not tweak their applications to take advantage of more than two cores. AVG Anti Virus, Fritz 11 Chess, the Mainconcept 1.5.1 H.264 encoder, SiSoft Sandra XII and Supreme Commander are all examples showing that a 2.0 GHz quad core can certainly beat a sophisticated 3.16 GHz dual core.
Of course, there will always be applications that do not scale well with an increasing core count, and AMD has a long way to go if it wants to get back on top and compete with Intel’s processors. Core 2 is clearly superior in apples to apples comparisons, and it also beats the Phenom X4 in this comparison for the reasons mentioned above. One fact remains clear above all: our comparison has shown that the time for quad core processors just hasn’t arrived yet.