Performance tests

Testbed:

• 2 x Intel Core i7-3960X (o/c to 4 GHz) CPU
• Hydro Series H100i Extreme Performance CPU cooler
• Intel Thermal Solution RTS2011LC cooler
• ASUS Sabertooth X79 motherboard on the Intel X79 chipset
• MSI X79A-GD45(8B) motherboard on the Intel X79 chipset
• 16GB of 1600MHz Corsair Vengeance CMZ16GX3M4A1600C9 DDR3 SDRAM
• Seagate Barracuda 7200.14 3TB SATA2 HDD
• WD Caviar Blue WD10EZEX 1TB SATA2 HDD
• 2 x Corsair Neutron CSSD-N120GB3-BK SSD
• 2 x 1200W Corsair CMPSU-1200AXEU PSU
• Corsair Obsidian 800D Full-Tower case
• Windows 7 Ultimate 64-bit, DirectX 11
• 30" Dell UltraSharp U3011 display
• VSync disabled
• AMD Catalyst 12.11beta8; NVIDIA Drivers 310.70

Benchmarks:

• 3DMark 11 (FutureMark) — DirectX 11.0, Performance settings.
• Aliens vs. Predator (Rebellion/SEGA) — DirectX 11.0, Very High settings, run from in game.
• Nexuiz (2012) (IllFonic/THQ) — DirectX 11.0, built-in benchmark, maximum quality settings.
• Crysis 2 Maximum Edition (Crytek/EA) — DirectX 11.0, Very High settings, Central Park level, launched with the Adrenaline Crysis 2 Benchmark Tool.
• DiRT: Showdown (Codemasters) — DirectX 11.0, Ultra High settings, launched as follows: dirt showdown.exe -benchmark example_benchmark.xml.
• F1 2010 (Codemasters) — DirectX 11.0, Ultra High settings, run as follows: "formulaone.exe -benchmark example_benchmark.xml".
• Hard Reset (Flying Wild Hog) — DirectX 11.0, built-in benchmark, maximum quality settings.
• Heaven Benchmark 2.0 (Unigine) — DirectX 11.0, High settings.
• Metro 2033 (4A Games/THQ) — DirectX 11.0, Super High settings, PhysX disabled, run from in game.
• Total War: Shogun 2 (Creative Assembly/SEGA) — DirectX 11.0, maximum quality settings.
• Sleeping Dogs (United Front Games/Square Enix) — DirectX 11.0, built-in benchmark, maximum quality settings.

Final thoughts

The capability rating compares each graphics card with GeForce GT 630, which is considered a baseline, or 100%. The rating is based on both synthetic and gaming results and shows, as the name implies, what a product is capable of.

The usability rating is obtained by dividing each card's capability rating by its price. It basically shows whether a given product is over or underpriced, considering what it can do, and thus how reasonable it is to buy it.

The complete ratings can be found in our i3Dspeed. Below are just the parts relevant to today's review.

Sorted by capability (high to low)

Inno3D GTX 670 iChill takes the lead in its segment, outperforming even Radeon HD 7970, prices being almost the same. Inno3D GTX 660 Ti performs on a par with the reference GTX 660 Ti, and they both beat the AMD Radeon HD 7950.

Sorted by usability (high to low)

On average, Inno3D GTX 660 Ti is cheaper than other GTX 660 Ti adapters. As for Inno3D GTX 670 iChill, it yields to Radeon HD 7950 because of the higher price. In turn, Radeon HD 7970 falls behind.

Inno3D GeForce GTX 670 iChill is an excellent graphics adapter, working at very low temperatures and with very little noise. The only disadvantage is its big size: 3 slots thick and longer than 30 cm.

Inno3D GeForce GTX 660 Ti is also a good choice, with a smaller, dual-slot cooling system that does its job very nicely. Compared to the iChill, the only drawback is more noise.

Write a comment below. No registration needed!