How we test
We've changed the system setup to better reflect the kind of setup that would house these mid-range cards. Out goes Intel's Core i7 965 and X58, replaced by Core i5 750 and P55.We've also turned the settings down on the games and benchmarked them at 1,280x1,024, 1,680x1,050, and 1,920x1,080. Providing the comparison multi-GPU numbers from an AMD perspective are two Radeon HD 5750 1,024MB cards
GPU comparisons |
||||||||
|---|---|---|---|---|---|---|---|---|
| Graphics cards | Current pricing | Shader model | Stream processors | GPU clock speed (MHz) | Shader clock speed (MHz) | Memory clock speed (MHz) | Memory bus width (bits) | Graphics driver |
| NVIDIA GeForce GTS 450 1,024MB SLI | £198 | 5.0 | 192 x 2 | 783 | 1,566 | 3,608 | 128 x 2 | ForceWare 260.52 |
| NVIDIA GeForce GTS 450 1,024MB | £99 | 5.0 | 192 | 783 | 1,566 | 3,608 | 128 | ForceWare 260.52 |
| ASUS GeForce GTS 450 TOP 1,024MB | £115 | 5.0 | 192 | 925 | 1,850 | 4,000 | 128 | ForceWare 260.52 |
| ZOTAC GeForce GTX 460 768MB | £130 | 5.0 | 336 | 675 | 1,350 | 3,600 | 192 | ForceWare 260.52 |
| Sapphire Radeon HD 5750 1,024MB XF | £210 | 5.0 | 720 x 2 |
700 | 700 | 4,600 | 128 x 2 | Catalyst 10.8 |
| Sapphire Radeon HD 5750 1,024MB | £105 | 5.0 | 720 |
700 | 700 | 4,600 | 128 | Catalyst 10.8 |
| Sapphire Radeon HD 5670 1,024MB | £85 | 5.0 | 400 | 775 | 775 | 4,000 | 128 | Catalyst 10.8 |
| HIS Radeon HD 5770 1,024MB | £115 | 5.0 | 800 | 850 | 850 | 4,800 | 128 | Catalyst 10.8 |
Test bench |
||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CPU | Intel Core i5 750 (2.67GHz, 8MB L3) | |||||||||||
| Motherboard | ASUS P7P55D Premium | |||||||||||
| Motherboard BIOS | 1807 | |||||||||||
| Memory | 4GB Kingston DDR3 | |||||||||||
| Memory timings and speed | 9-9-9-24-1T @ DDR3-1,333 | |||||||||||
| PSU | Corsair HX650W | |||||||||||
| Monitor | Dell 24in 2405FPW - 1,920x1,200px | |||||||||||
| Disk drive(s) | Corsair Nova V128 SSD | |||||||||||
| Operating system | Windows 7 Ultimate, 64-bit | |||||||||||
Benchmarks |
||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Aliens vs. Predator | DX11, 1,280x1,024, 1,680x1,050 and 1,920x1,080 resolutions, 2xAA, 8xAF, low quality. | |||||||||||
| Mafia II | DX9, 1,280x1,024, 1,680x1,050 and 1,920x1,080 resolutions, 0xAA, 8xAF, medium quality | |||||||||||
| Call of Duty: Modern Warfare 2 | DX9, 1,280x1,024, 1,680x1,050 and 1,920x1,080 resolutions, 2xAA, normal quality, FRAPS-recorded benchmark. | |||||||||||
| Crysis Warhead | DX10, 1,280x1,024, 1,680x1,050 and 1,920x1,080 resolutions, 0xAA, mainstream quality, Frost map, FRAPS-recorded benchmark. | |||||||||||
| DiRT 2 | DX11, 1,280x1,024, 1,680x1,050 and 1,920x1,080 resolutions, 2xAA, ultra quality, London map. | |||||||||||
| Temperature | To emulate real-world usage scenarios, we record GPU core temperature both when idle and whilst playing Crysis Warhead. And, for a worst-case scenario, we throw in numbers from the FurMark stress test. | |||||||||||
| Power consumption | To emulate real-world usage scenarios, we record mains power draw both when idle and whilst playing Crysis Warhead. And, for a worst-case scenario, we throw in numbers from the FurMark stress test. | |||||||||||
| Noise | Albeit not scientific, we use the tried-and-trusted human ear to share an opinion on GPU noise. | |||||||||||