Re: So its more like 13:1 in favor of Athlon 64s (754) even with limitations of low end graphics controllers.
Peter, of the Athlon 64 chips being reviewed, one was a DTR version (81.5W) and the other was a fully fledged desktop part (89W). Your previous comment claimed that Dothan would have a tough time competing against Athlon 64 - even at the same power levels. But on the contrary, it does a decent job competing, even against fully optimized desktop chips with 20% greater memory bandwidth.
With the exception of the synthetic benchmarks you listed, here is how far Dothan is behind a CPU with >4x higher TDP and 400MHz greater clock speed advantage:
-2%, Unreal Tournament 2004 - 1024 x 768 - Convoy
-1%, Unreal Tournament 2004 - 1024 x 768 - Colossus
-3%, Halo : Combat Evolved - 1024 x 768
-4%, Adobe Photoshop CS Filter Test (wins against cheaper one)
-32%, Alias/Wavefront Maya 6.0 Zoo Render Benchmark
-17%, Windows Media Encoder 9 MPEG Encoding Test
+7%, LAME 3.96 MP3 Encoding Test
A 1-5% performance difference is not recognizable to the end user. This means that to most mobile users, they could buy a 9+ pound monstrocity with an Athlon 64, or get nearly identical performance with a Dothan chip with one quarter the power dissipation and the largest performance penalty being rendering tasks, which are hardly ever performed by mobile users (only notable exception being mobile workstations).
Face it, Pete: Dothan kills the Athlon 64 in the mobile market. Even your skewed benchmarks prove this. The ones from Anandtech show an Athlon 64 more likely to be in mid-sized laptop, and Dothan shows a more decisive performance victory. Given that thin & light laptops is the fastest growing segment, AMD is going to need 90nm just to compete, and by this time, Dothan will be a stronger product with 60% greater memory bandwidth and likely higher clock speeds.
News 
Market Data 
Discover 
