Date:
Tuesday , September 23, 2003
Author:
Kyle Bennett
Editor:
Sean Quinn
Google +1

Athlon 64 Vs. Pentium 4

AMD's Athlon 64 and Athlon 64 FX Vs. Intel's Pentium 4 Extreme Edition. We showcase the latest CPUs from AMD and Intel and take a look at what they will deliver.

continued...

Test Systems

For our testing, we were sent a full test system from AMD. We generally like to do things on our own, so we grabbed the CPU and the BIOS from the supplied mainboard and set off with our own retail parts.

AthlonFX 51 2.2GHz - Asus SK8N (nForce3 Pro) retail mainboard (using supplied AMD BIOS), (2X512MB) 1GB Legacy CAS 2.5 DDR400 (2.5,3,3,6), ATI Radeon 9700 Pro w/ ATI Catalyst 3.7 drivers, 40GB Maxtor ATA133 HDD, Windows XP w/SP1 and DX9B.

Athlon64 3200+ 2.0GHz - MSI K8T Neo (VIA K8T800) & Shuttle AN50R (nForce3 150), (2X512MB) 1GB Corsair CAS 2.0 DDR400 (2,3,2,5) ATI Radeon 9700 Pro w/ ATI Catalyst 3.7 drivers, 40GB Maxtor ATA133 HDD, Windows XP w/SP1 and DX9B.

Intel Pentium 4 Extreme Edition & Non-EE 3.2GHz/800MHz - Intel Bonanza Desktop (i875P), (2X512MB) 1GB Corsair CAS 2.0 DDR400 (2,3,2,5), ATI Radeon 9700 Pro w/ ATI Catalyst 3.7 drivers, 40GB Maxtor ATA133 HDD, Windows XP w/SP1 and DX9B.

AthlonXP 3200+ - Asus A7N8X (nForce2 Ultra 400), (2X512MB) 1GB Corsair CAS 2.0 DDR400 (2,3,2,5) ATI Radeon 9700 Pro w/ ATI Catalyst 3.7 drivers, 40GB Maxtor ATA133 HDD, Windows XP w/SP1 and DX9B.

What we've laid out here for you is a comparison using only retail components. There are no reference boards or funky hardware that can't be purchased through retail channels currently, or at least very soon.

We've done some extra testing with varied memory bus frequencies as well. Some of you looking to upgrade may not want to go out and purchase the DDR400 that's currently on the market, so we've tested the two new AMD CPUs with both DDR333 and DDR400.

Music Benchmark

We used dBPower Amp to convert a Frank Zappa 100MB .wav file to MP3 using LAME.

Article Image

Obviously from the results, a benchmark of this nature relies little on CPU cache or other subsystem components as long as there isn't an obvious bottleneck there. The very high clock of the Pentium4 here makes handy work of the rest of the field. It's interesting to note that while the P4 did win, it didn't win by the delta that would be suggested on a clock-to-clock comparison with the Athlons. No doubt the Athlon64s are doing more work per CPU clock cycle in comparison to the Pentium 4s and the AthlonXP.

Movie Benchmark

For this set of data we used the widely popular "DVD2AVI" program along with the v 5.0.5 DivX CODEC. A VOB file of the movie "8mm" was ripped that was approximately 3.5GB in size. We then compressed the video stream (no audio) using our program and CODEC.

Article Image

We're seeing much different results here than we did with our MP3 conversion. Both flavors of the Athlon64 CPUs clearly lead the category over the Pentium 4s. The AthlonFX finished approximately 9 minutes faster than the P4, and if you're ripping a lot of movies back-to-back this could certainly make a difference.

What's probably most noteworthy about this graph is not the fact that the Athlon64s outpaced the P4, but that the AthlonXP 3200+ got outpaced as well by the Athlon64 3200+. Also consider that the P4 and the P4EE came in at a dead tie, which suggests that the extra CPU cache is not a factor in this benchmark. If cache is not a factor here, then it would make sense that pipeline enhancements made to the new K8 core certainly did impact instructions per clock as can be seen when comparing the two 3200+ rated CPUs side by side.