Tom's CPU Architecture Shootout: 16 CPUs, One Core Each, And 3 GHz

Benchmark Results: Professional Applications

Professional graphics rendering performance is largely dependent on the number of available cores, which means that results from one core are more synthetic than anything, given today's predominantly multi-core landscape. If we were running this test with all cores enabled, we'd naturally see something completely different.

Once again, Intel delivers better performance per clock from its most modern products. Like we said, though, putting a six-core AMD chip against a quad- or dual-core Intel processor at the same price point will drastically change the performance picture. Our exploration here is experimental by nature; we want to test per-clock performance of just one core.

Photoshop is one of the few applications where Intel's x86 architectural efforts don't just end up in a faster product than AMD’s cores; they're substantially faster. It’s not hard to imagine what happens if you scale the performance difference across additional processor cores. At the same time, at any given price point, it's also easier to get more of AMD's cores than Intel's.

Differences are similarly notable in Adobe’s Premiere. The results clearly demonstrate that you shouldn’t use the early Athlon 64 X2 processors or especially Intel’s NetBurst-based products if you can help it.

AMD’s cores look much better in Blender, which is also an image rendering application. The program seems to benefit a lot from the Sandy Bridge architecture.