PDA

View Full Version : GTX 680: PhysX Benchmarks roundup


News
03-22-12, 08:30 PM
Finally, NVIDIA has officially presented (http://www.geforce.com/whats-new/articles/introducing-the-geforce-gtx-680-gpu/) their new GPU ‚?? GTX 680, built on 28nm GK104 chip, which itself is based on next-gen Kepler architecture.

http://physxinfo.com/news/wp-content/uploads/2012/03/NV_GF_GTX3.jpg

Many rumors were floating around Kepler and its physics acceleration capabilities, some (http://news.techeye.net/chips/nvidias-kepler-suffers-wobbly-perturbations) were endowing new GPU with dedicated PhysX blocks, others (http://semiaccurate.com/2012/02/01/physics-hardware-makes-keplergk104-fast/) were claiming ability to run CPU PhysX games in hardware ‚?? none of this happens to be true. Nothing special (like NVENC video encoder) for physics calculations, just general improvements to chip architecture, SM clusters, etc.

Nonetheless, we will try to gather all available information regarding PhysX computation performance of the GTX 680, in comparison to previous generation GPUs.

NVIDIA Kepler GPU GeForce GTX 680 Video Card Review (http://www.hardocp.com/article/2012/03/22/nvidia_kepler_gpu_geforce_gtx_680_video_card_revie w/4) by HARDOCP

http://physxinfo.com/news/wp-content/uploads/2012/03/Batman_HardOCP.png

Interesting results ‚?? in this case, GeForce GTX 680 performance is higher than Radeon HD 7970 even with GPU PhysX effects enabled. However, we can not directly compare GTX 580 and GTX 680 due to different resolution settings.



NVIDIA GeForce GTX 680 (Kepler) (http://www.hardwareheaven.com/reviews/1452/pg10/nvidia-geforce-gtx-680-kepler-graphics-card-review-batman-arkham-city.html) by HardwareHeaven

http://physxinfo.com/news/wp-content/uploads/2012/03/Kepler_HHeaven.jpg

At lower resolutions situation is quite different ‚?? enabled PhysX acceleration introduces significant penalty to GTX 680 performance.

NVIDIA GeForce GTX 680 Review (http://www.eurogamer.net/articles/df-hardware-nvidia-gtx-680-review) by Eurogamer

http://physxinfo.com/news/wp-content/uploads/2012/03/Kepler_eurogamer.png

GTX 680 is showing 20% performance improvement over its predecessor, but again on medium PhysX settings only.

Test: Nvidia GeForce GTX 680 (http://www.computerbase.de/artikel/grafikkarten/2012/test-nvidia-geforce-gtx-680/24/) by ComputerBase

http://physxinfo.com/news/wp-content/uploads/2012/03/kepler_CB1.pnghttp://physxinfo.com/news/wp-content/uploads/2012/03/kepler_CB2.png

GTX 680 vs GTX 580: similar results in Dark Void Benchmark (only ‚??PhysX Low‚?? settings are used ‚?? simple particle effects) and significant performance difference in Mars Benchmark (http://physxinfo.com/news/5837/new-gpu-physx-benchmark-released-for-online-shooter-mars/) (which contains more diverse PhysX content like fluids, particles and tearable cloth, but also produces very disputable (http://www.geeks3d.com/20110623/tested-mars-physx-benchmark/) results).

Nvidia‚??s GTX 680 arrives! ‚??Faster, Smoother, Richer‚?? (http://alienbabeltech.com/abt/viewtopic.php?f=4&t=24517) by Alienbabeltech.com

http://physxinfo.com/news/wp-content/uploads/2012/03/Kepler_alien.jpg

Most comprehensive set of PhysX benchmarks comes from Alienbabeltech ! According to their numbers, average PhysX performance difference between GTX 680 and GTX 580 is around 11 %. Not as impressive as one might expect, huh?

Note: surpisingly, from several dozens of GTX 680 reviews we were able to find, only a few are containing actual PhysX benchmarks, majority of the reviewers just leave hardware acceleration disabled.



More... (http://physxinfo.com/news/7862/gtx-680-physx-benchmarks-roundup/)