Test Methodology
Sapphire Radeon HD 7950 OC with Boost specification |
|---|
 |
GPU Comparisons |
|||||||
|---|---|---|---|---|---|---|---|
| Graphics card | GPU clock (MHz) |
Stream processors |
Shader clock (MHz) |
Memory clock (MHz) |
Memory bus (bits) |
Graphics driver | |
| Sapphire Radeon HD 7970 (3,072MB) | 925 | 2,048 | 925 | 5,500 | 384 | Catalyst 12.7 beta | |
| Sapphire Radeon HD 7950 OC Vapor-X (3,072MB) | 950 | 1,792 | 950 | 5,000 | 384 | Catalyst 12.8 | Sapphire Radeon HD 7950 (3,072MB) | 800 | 1,792 | 800 | 5,000 | 384 | Catalyst 12.7 beta | Sapphire Radeon HD 7870 (3,072MB) | 1,000 | 1,280 | 1,000 | 4,800 | 256 | Catalyst 12.7 beta |
| KFA² GeForce GTX 660 Ti EX OC (3,072MB) | 1,006+ | 1,344 | 1,006+ | 6,008 | 192 | GeForce 305.37 beta | |
| EVGA GeForce GTX 680 (2,048MB) | 1,006+ | 1,536 | 1,006+ | 6,008 | 256 | GeForce 305.37 beta | |
| ASUS GeForce GTX 670 (2,048MB) | 915+ | 1,344 | 915+ | 6,008 | 256 | GeForce 305.37 beta | |
| Gigabyte GeForce GTX 660 Ti OC (2,048MB) | 1,032+ | 1,344 | 1,032+ | 6,008 | 256 | GeForce 305.37 beta | |
| KFA² GeForce GTX 660 Ti EX OC (3,072MB) | 1,006+ | 1,344 | 1,006+ | 6,008 | 256 | GeForce 305.37 beta | |
| EVGA GeForce GTX 660 Ti SC (2,048MB) | 980+ | 1,344 | 980+ | 6,008 | 256 | GeForce 305.37 beta | |
HEXUS High-End Test Bench |
||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Processor | Intel Core i7-3770K (3.5GHz, 8MB cache, quad-core) | |||||||||||
| CPU Cooler | Intel reference E97378-001 | |||||||||||
| Motherboard | Gigabyte GA-Z77X-UD5H | |||||||||||
| Memory | 8GB G.Skill TridentX (2x4GB) DDR3 @ 2,400MHz | |||||||||||
| Power Supply | Corsair AX750W | |||||||||||
| Storage Device | Samsung 830 256GB SSD | |||||||||||
| Optical Drive | Generic 24x DVD-RW | |||||||||||
| Chassis | Corsair Graphite Series 600T | |||||||||||
| Monitor | Dell 30in 3007WFP | |||||||||||
| Operating system | Windows 7 Ultimate (64-bit, SP1) | |||||||||||
HEXUS High-End Benchmark Suite |
||
|---|---|---|
| Mode and Resolutions | Quality Settings | |
| 3DMark 11 | DX11 at 1,280x720 and 1,920x1,080 | Performance and Extreme Presets |
| Aliens vs. Predator | DX11 at 1,920x1,080 and 2,560x1,600 | 4xAA, 16xAF, Maximum Image Quality |
| Battlefield 3 | DX11 at 1,920x1,080 and 2,560x1,600 | 4xMSAA, 16xAF, Ultra Preset |
| Batman: Arkham City | DX11 at 1,920x1,080 and 2,560x1,600 | 8xMSAA, Extreme Preset |
| Crysis 2 | DX11 at 1,920x1,080 and 2,560x1,600 | 4xAA, Ultra Preset, DX11 and High-Res Textures |
| DiRT Showdown | DX11 at 1,920x1,080 and 2,560x1,600 | 4xMSAA, Ultra Preset |
| Just Cause 2 | DX10 at 1,920x1,080 and 2,560x1,600 | 8xAA, 16xAF, Very High Quality, PLS On |
| Total War: Shogun 2 | DX11 at 1,920x1,080 and 2,560x1,600 | 4xMSAA, 16xAF, Custom High Quality |
| LuxMark 2.0 | OpenCL GPUs-only | Sala (488,000+ triangles) and Room (2,000,000+ triangles) |
| General Benchmarks | ||
| Power Consumption | To emulate real-world usage scenarios, we record mains power draw both when idle and while playing Batman: Arkham City | |
| Temperature | To emulate real-world usage scenarios, we record GPU core temperature both when idle and while playing Batman: Arkham City | |
| Noise | A PCE-318 meter is used to record noise levels when idle and while playing Batman: Arkham City | |
Notes
We're using this review as an opportunity to introduce a new test platform and a refreshed benchmark suite. Our trusty-ol' Sandy Bridge rig has been retired in favour of a new Ivy Bridge system that consists of an Intel Core i7-3770K processor, a Gigabyte GA-Z77X-UD5H motherboard and 8GB of G.Skill memory. DiRT Showdown and LuxMark 2.0 have been added to the benchmark suite.
In the interests of fairness, any GPU vendor-specific technologies are disabled during benchmarking. For example, NVIDIA-exclusive features such as PhysX in Batman: Arkham City and the Bokeh Filter in Just Cause 2 are not enabled.
We've run the Sapphire card with the PTWB feature enabled - the second BIOS position, so the speak - and the card flicks between 850MHz and 950MHz (PTWB) core on an even, regular basis. We also tested the card on Catalyst 12.7 beta and Catalyst 12.8 WHQL and found no performance difference between the sets.
