Far cry benchmark

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain about 3.5fps
Turion 64 at 2.4Ghz, 1x registered DDR 480 stick, CR1 ( Re: Battle of the platforms: socket 754! )

Quake 4 benchmark

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain about 4 fps
Turion 64 at 2.4Ghz, 1x registered DDR 480 stick, CR1 ( Re: Battle of the platforms: socket 754! )

FEAR built-in benchmark. All details set to maximum except for soft shadows which nd22 disabled. FSAA 4x, Anisotropic 16x. Enabled HW audio mixing and EAX 2.0 since we have Audigy 2 ZS.

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain 2 fps
Turion 64 at 2.4Ghz, 1x registered DDR 480 stick, CR1 ( Re: Battle of the platforms: socket 754! )

Crysis built-in 32bit benchmark. All details set to HIGH except for anti-aliasing quality which was set to 4x. First run is ignored (warmup), the remaining 3 are averaged.

• 1600x1200 - min 11.74, avg 30.78, max 46.98

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain 1 fps
Turion 64 at 2.4Ghz, 1x registered DDR 480 stick, CR1 ( Re: Battle of the platforms: socket 754! )

Serious Sam 2 (2005) benchmark with Turion 64 at 2400Mhz and GTX 275:

The game needs to be patched to 2.070. Benchmarked using Serious Sam 2 Benchmark tool, 1st benchmark.

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain about 2.5 fps
Turion 64 at 2.4Ghz, 1x registered DDR 480 stick, CR1 ( Re: Battle of the platforms: socket 754! )

FEAR Perseus Mandate (2007), version 1.0. It falls into Windows Vista era (2007-2009).

Maximum details, FSAA 4x, Anisotropic 16x, SS off. Enabled HW audio mixing and EAX 2.0 since we have Audigy 2 ZS.

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain about 3 fps
Turion 64 at 2.4Ghz, 1x registered DDR 480 stick, CR1 ( Re: Battle of the platforms: socket 754! )

Prey (2006) benchmark with Turion 64 at 2400Mhz and GTX 275:

The game needs to be patched to 1.4.0. Benchmarked using Prey Benchmark tool.

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain about 6-7 fps
Turion 64 at 2.4Ghz, 1x registered DDR 480 stick, CR1 ( Re: Battle of the platforms: socket 754! )

Splinter Cell Chaos Theory (2005) benchmark with Turion 64 at 2400Mhz and GTX 275:

The game needs to be patched to 1.0.5. Benchmarked using Splinter Cell Chaos Theory Benchmark tool.

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain about 5 fps
Turion 64 at 2.4Ghz, 1x registered DDR 480 stick, CR1 ( Re: Battle of the platforms: socket 754! )

Company of Heroes Complete Edition (2009) benchmark with Turion 64 at 2400Mhz and GTX 275:

4x AA was used, the rest is maximum quality. Some options allow Ultra quality that cannot be selected (Shader Quality, Terrain Detail). It can probably be enabled only with dx10. Environmental reverb cannot be enabled.

Minimum Frame Rate is probably a benchmark glitch. The value is not reliable.

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain about 12 fps
Turion 64 at 2.4Ghz, 1x registered DDR 480 stick, CR1 ( Re: Battle of the platforms: socket 754! )

I reran the benchmark 2nd time to validate average and max frame rate. Minimum was about 18 fps 2nd time. It seems system with 1x registered DDR 480 stick was held back by latency. This benchmark is a cinematic sequence so not necessarily representative of the game.

AlexZ wrote on Today, 06:30:

That's very nice, although 256MB is low amount. You want 512MB for Windows 98. It's an option for someone not having Kingston HyperX, OCZ Platinum or Corsair which can run at CAS 2, DDR400 even for 1GB modules.

Ended up with the same performance at 512MB. Playing around with rank and bank it appears that dropping the stick to one rank is where the performance comes from. I'm only doing this on junk ram so no worries.

AlexZ wrote on Today, 13:34:

Splinter Cell Chaos Theory (2005) benchmark with Turion 64 at 2400Mhz and GTX 275:

Saw those benchmark results and just wanted to comment: this is a great example of a CPU bottleneck. The framerate remains the same at 1024x768 and 1600x1200. The GPU can clearly do more, but it is held back by the CPU. I'm aware that you guys already know this, I'm just pointing it out for new readers who might not be as familiar with the subject.

Would have been interesting to see the min and max FPS values too.

Unreal Tournament 3 Black Edition (2009, base game from 2007) benchmark with Turion 64 at 2400Mhz and GTX 275:

Game version is 2.1. Benchmarked using UT3 Benchmark Tool (version 0.2.0.45). HOC Unreal Tournament 3 Bench 1.3 crashes the game. Resolution is 1600x1200.

Comparison benchmarks:
Turion 64 at 2.4Ghz, 2x unbuffered DDR 400 sticks, CR1 ( Re: Battle of the platforms: socket 754! ), we gain about 1.5 fps

All benchmarked maps are bot matches. Due to randomness of bot matches 12 maps were benchmarked and sum of gains was calculated, then divided by the number of maps.

WAR-Downtown -1.6
WAR-Avalanche +6.63
vCTF-Containment +3.35
vCTF-Sandstorm -1.9
CTF-Coret -0.11
DM-Santuary +4.86
DM-Sentinel +5.53
DM-HeatRay -3.16
DM-Deimos -3.08
DM-Biohazard +5.96
CTF-OmicronDawn +0.38
CTF-Hydrosis +1.2
-----------------------
Sum +18.06 fps
Per map +1.505 fps

Joseph_Joestar wrote on Today, 13:53:

Saw those benchmark results and just wanted to comment: this is a great example of a CPU bottleneck. The framerate remains the same at 1024x768 and 1600x1200. The GPU can clearly do more, but it is held back by the CPU. I'm aware that you guys already know this, I'm just pointing it out for new readers who might not be as familiar with the subject.

Would have been interesting to see the min and max FPS values too.

The CPU bottleneck is unavoidable when chasing top performance out of s754. We have usable min/max values for FEAR, Crysis, Unreal tournament 2004 and Half-Life 2. I didn't test the last 2 because they aren't sufficiently challenging for PCIe and difficult to reproduce for me. Once nd22 tests X2 CPUs in s939 topic, bigger minimum should be one of main advantages.

The last couple of builds demonstrate how far s754 can be pushed without too crazy OC. We are not doing world records here, only pushing things on default voltage. Almost anyone can have Lancaster CPUs running at 2.2-2.4Ghz by buying a 1.8Ghz one and OCing it. DDR533 offers very small benefits and seems unnecessary. It was meant to address the CPU bottleneck and improve sound playback.

This topic also shows how vast differences there can be between the most powerful Athlon XP, entry level Athlon 64 (slightly better than Athlon XP) and Athlon 64 with PCIe.

nd22 did K7 vs K8 comparison here Re: Battle of the platforms: socket 754!
Far Cry, FEAR and Quake 4 can be compared in 1600x1200. Athlon 64 can deliver 5x higher framerate in FEAR and 2x higher framerate in Far Cry than the most powerful Athlon XP. We get 4x higher minimums as well, we aren't getting higher average fps just due to higher maximums.