AMD AM2 3500+ Vs. AMD 939 3500+

Testing

For testing these two processors the only difference between the systems will be the motherboard, CPU, and RAM; the rest of the parts, along with the drivers, are identical. When initially building this machine, I attempted to use a Gigabyte motherboard. After fighting with it for a three days, I could not get the computer stable at all. When I switched to the Asrock motherboard, I was able to achieve 100% stability. So it would seem that there are still some bugs to work out of this new architecture, but most likely I would say it is a BIOS issue. Specs are as follows:

AM2

• AM2 3500+
• Asrock AM2NF4G
• 2X512MB 667Mhz 5-5-5-10
• 80Gig Maxtor SATA
• Gigabyte 6600GT 128MB
• Thermaltake 680 Watt Power Supply
• Windows XP Pro Service Pack 2
• NVIDIA Driver 84.21

939

• AMD 3500+
• MSI K8N Neo 4 Platinum
• 2X512MB Kingston 2-2-2-5
• 80Gig Maxtor SATA
• Gigabyte 6600GT 128MB
• Thermaltake 680 Watt Power Supply
• Windows XP Pro Service Pack 2
• NVIDIA Driver 84.21

For testing, we will be following the standard array of OCC CPU tests, though we will also be throwing in a set of game benchmarks to see if the socket type can indeed affect game FPS.

At the head of the tests we have Apophysis, a unique piece of fractal rendering software. We have a pre defined image to render, and tested how long it took to render on each machine. Remember, lower is better.

As we can see form the graph, the AM2 lost by over 5 minutes. While this is not a huge loss, it is a disappointment. Following Apophysis we ran the Sciencemark suite of benchmarks.

Again AM2 is lagging behind, though only slightly.

Next on the list is Cinebench, which is a subsection of Cinema4D, another piece of fractal rendering software. There are two CPU tests, though we only chose to run the first of the two.

With a 5 point win this time, the AM2 seems to be a smidge ahead of its 939 predecessor.