nVidia once again made the decision to launch a new GPU. Yet another number to pad out their already crowded FX line up. But before we damn this decision, lets take a look at it's performance.

We used the same test bed as we did for our round up at Christmas. So we replicated the same benchmarks giving us a very easy comparison. We have bought in the benchmarks from the Sparkle 5600 and Terratec 5200 so we can get an idea of how it compares to it's closest competition for your money.

This review is pretty much an extension of our last round up, so for an explanation of the benchmarks used, excellent buying advice or results from other cards, we highly recommend giving the round up a glance through.

The Testing System

All the tests were run on a Pentium 4 (Northwood) system, with an 800 mhz FSB Intel Springdale Motherboard, accompanied by 1 Gigabyte of Kingmax Dual DDR400, kindly supplied by Overclock.co.uk.

Windows XP SP1 was used for testing, with Direct X 9.0b, Fraps v2.0 build 1787, 3dmark 2001 SE build 330, 3dmark 2003 PRO build 340, and Aquamark 3.0.

The ForceWare 52.16 drivers were used for the 5200 and 5600 cards and ForceWare 56.55 where used for the 5500. Unless otherwise noted, all tests were run with the CPU at 3.2 ghz, with an 800mhz FSB. Tests run at 1.6 ghz were also run with an 800mhz FSB, the only change being the multiplier.

The Benchmarks

• 3dMark 2001 SE
• 3dMark 2003
• AquaMark 3.0
• Unreal Tournament 2003
• Grand Theft Auto: Vice City
• Serious Sam the Second Encounter

We have tried to get a range of benchmarks to test the cards in different scenarios. Despite attempts to cheat, we still feel 3dMark is a useful program and tallys with our results from the other benchmarks. We are of course using the latest build of 3dmark and drivers that don't allow cheating.

We feel 3dMark is a very good general summary of a cards performance so we have used it as the basis for our "bang for buck" rating. This is quite simply, the 3dmark divided by the price to give the number of 3d marks per pound.

<< Back | 1 | | 2 | | 3 | | 4 | | 5 | | 6 | Next >>