Chic, <At 1.5 Ghz, AMD on DDR, Intel on DRDRAM, how much better do you think Willy will be on a percentage basis (average for a wide range of benchmarks)?>
I think the best AMD can hope for is to stay within 10% on certain benchmarks, most notably the office suite benchmarks. I expect up to a 30% advantage in some other benchmarks for Willamette, particularly the high-end workstation benchmarks like SPEC2000 or 3D Studio MAX. And that's assuming that the clock speeds will be equal, which itself is another thing that AMD has to hope for.
<How much of this performance gain will be due to DRDRAM?>
I don't know, but I'm sure you can guess how much PC100 SDRAM hurt the performance of Athlon back when it was released. Athlon's EV6 bus had 1.6 GB/sec of bandwidth, yet the 0.8 GB/sec PC100 channel hampered performance. Athlon got a small boost with KX133 chipset and PC133 SDRAM, and Athlon is expected to benefit even more when the DDR chipsets are out.
Take the possible performance difference between Athlon on KX133 and Athlon on DDR, and you'll have a pretty good idea of how much the dual-RDRAM channels will benefit Willamette (over a single-RDRAM or single-DDR channel, for example). Unlike Athlon which started low and is being pushed up higher, Willamette will start high and be pushed down. An interesting strategy, no doubt.
Tenchusatsu |
