Overclocking and System Setup

This content was originally featured on Amdmb.com and has been converted to PC Perspective’s website. Some color changes and flaws may appear.

Our last overclocking journey was a little more trouble than I expected, so this time we are going to do everything right.

While I was at Quakecon 2002 over the weekend, I had not yet overclocked the Athlon XP 2600+ processor sample I had received. The timing on the release forced me to only spend a day and half with the chip prior to leaving for the show and only a day and half after returning from the show. I spent some time with Kyle Bennett of HardOCP while there and we were discussing the new processor with the folks from AMD that were in attendance. Kyle had already been messing with the processor for its overclocking potential and by using only front-side bus increases, was able to get a 2.485 GHz clock speed out of the chip. As you might guess, this piqued my interest and was one of the first things I did upon returning home.

Try as I might, I had no luck unlocking this new Thoroughbred core. I tried the L1 and L3 bridges; I tried different boards and bios’; I tried the same things I used on the 2200+ and others I had seen on the web for the 2200+, but I just couldn’t get it to budge. Every time I changed the multiplier in the bios, the machine wouldn’t post until I reset the bios. Do I think AMD actually locked the processors permanently? Nope, I just wasn’t able to discover the trick to it – but I am sure someone out there will very soon.

That left me with front-side bus overclocking only. That didn’t give me a good feeling at first with the somewhat limited success of FSB overclocking on the first 2200+ Thoroughbreds. But, AMD assured me that the problems with the core had been worked out with the additional layer that was designed into “Rev B” and they were very much correct. I was able to push the Athlon XP 2600+ processor up to a very fast 2.43 GHz and be 100% stable. I used the Thermalright AX-7 heatsink with a Vantec Tornado 80mm fan (Damn they are loud!) to keep the processor at below 115 degrees Fahrenheit at that speed as well. That is 152 MHz FSB and the default 16x multiplier for the 2.43 GHz speed.

What is even better is that I was able to post at speeds up to 2.576 (161 MHz x 16) but it wasn’t stable in Windows at all at that kind of speed. At 155 MHz, we could get through about 2 runs of 3D Mark 2001 but it would still eventually crash.

Here I have included a couple graphs of the overclocked Athlon XP 2600+ processor at various speeds. I included the SiSoft Sandra tests, as they are the ones that easily showed off the extra power you got with the extra speed, as well as some good old Quake III.

AMD Athlon XP 2600+ Processor Review - Processors 25

The FSB increases give the memory results a nice jump! At the highest speed, we saw a 17% jump in FPU results and an 18% jump ALU results.

AMD Athlon XP 2600+ Processor Review - Processors 26

Again we are seeing nice increases with the overclocking of this processor. 14% in the integer testing (ALU) and about the same for the floating-point testing (FPU).

AMD Athlon XP 2600+ Processor Review - Processors 27

Ah, yes! Breaking the 300 frames per second mark! 🙂 Here we are seeing absolutely screaming speeds on Quake III with the 2.43 GHz Athlon XP beating out the 2600+ stock 2.13 GHz by about 10% in the 16-bit mode and 8% in the 32-bit mode.

As you can see, the overclocking power of the newer “Rev B” Thoroughbred core seems to be very impressive. It will remain to be seen if other people have the same success here as we did, but being that I have spoken to a couple other editors on this issue already, I think we can expect this core to be a bit more forgiving for clock speed increases.

The system setup for this review consists of the test setup that AMD delivered to us which was basically an Athlon XP 2600+ processor, an Epox 8K3A+ motherboard based on the KT333 chipset, and 512 MB of TwinMOS PC2700 memory. I switched it out for Corsair XMS3200 memory instead to try and get the best memory settings possible. I then added our MSI Ti4400 video card and used an Antec TruePower 430 watt power supply to keep things stable.

AMD Test System Setup
CPU 1 x 2.13 GHz AMD Athlon XP 2600+ Processor
1 x 1.80 GHz AMD Athlon XP 2200+ Processor
Motherboard Epox 8K3A+ KT333
Memory 1 x 512MB Corsair Micro XMS3200 DDR DRAM
Hard Drive 80 GB 7200 RPM IBM EIDE
Video Card MSI GeForce 4 Ti 4400
Video Drivers Detonator 27.40
Operating System Windows XP

This is the first time here at Amdmb.com that will be testing the fastest processor of the AMD line up against the fastest processor of the Intel lineup. That’s right, there was an Intel 2.53 GHz processor on the Amdmb.com test bench! 🙂 I used an Intel 845E chipset motherboard that supported DDR memory. Unfortunately, I wasn’t able to get my hands on a faster performing motherboard with a P4X333 chipset to run against, so the Intel benchmarks may run a slight bit low when compared to other benchmarks you see of the 2.53 GHz P4. Cest la vie. Perhaps next time we will have a proper Intel setup to compare it with.

Intel Test System Setup
CPU 1 x 2.53 GHz Pentium 4 Processor
Motherboard Intel D845EBT i845E
Memory 1 x 512MB Corsair Micro XMS3200 DDR DRAM
Hard Drive 80 GB 7200 RPM IBM EIDE
Video Card MSI GeForce 4 Ti 4400
Video Drivers Detonator 27.40
Operating System Windows XP

Tests:

Quake III: Arena
3DMark 2001
PC Mark 200
Wolfenstein MP
Jedi Knight II
DronezMark
SiSoft Sandra Memory Bench
SiSoft Sandra CPU Bench
Content Creation Winstone 2001
Business Winstone 2001
4 different SPEC view perf tests
Cachemem
ScienceMark

« PreviousNext »