Posted on June 30th, 2009 by Mike Jennings
Will Nvidia PhysX ever be worthwhile?
Nvidia has talked up its PhysX system incessantly since it bought Ageia Technologies, creator of the engine, in February 2008, but it’s struggled to make a significant impact on the PC gaming landscape. So, despite the impressive tech demos and endless optimism, is PhysX looking more like a white elephant with every passing GPU and game release?
Certainly, the list of games using PhysX looks healthy enough. Already out are Bionic Commando, Army of Two, City of Villains, Empire: Total War, Gears of War 2, Haze, Mass Effect and Mirror’s Edge. Other promising titles, such as APB and Borderlands, both of which garnered plenty of attention at E3, are in the pipeline.
At a recent briefing, Nvidia unveiled another recent game to make heavy use of PhysX: Terminator Salvation. Unfortunately, though, it’s a typical movie tie-in, averaging scores of just above 50% since its release. And, while the demo (which showed the game running with and without PhysX looked impressive, we didn’t see any effects that couldn’t be achieved with systems other than PhysX. We’ve also struggled to think of many games where the presence of physics is genuinely game-changing rather than just an aesthetic enhancement.
There’s also a formidable list of games that don’t use Nvidia’s technology: BioShock, Company of Heroes, MotorStorm, Spore, Fallout 3, Fable 2, Halo 3, Killzone 2 and Half-Life 2 all use Havok, which was purchased by Intel in 2007 but, at the moment at least, is available for use on any hardware including PC and the big three consoles.
Nevertheless, the fragmented nature of the gaming physics market means that the Irish firm’s technology faces many of the same problems as PhysX even without being tied down to one hardware manufacturer.
Other gaming firms, including Crytek, are often happier to use their own physics engines instead of third-party technology. “Crytek has its own in-house physics system”, says Mark Atkinson, Crytek’s director of technology, which gives them “a single solution which can be optimised for all target platforms” at the same time – and, since Crysis 2 has been confirmed as a PC, PS3 and Xbox 360 release, this is obviously the easiest solution.
When developing Far Cry 2, Ubisoft eschewed PhysX and instead used Havok, a competing physics engine that was bought by Intel in 2007. Vincent Greco, a technical co-ordinator with Ubisoft, explained that “different games have different needs” and that, while Havok was the system of choice for Far Cry 2, PhysX “is a great technology” that Ubisoft has used to good effect in other big titles, such as Rainbow Six Vegas and EndWar.
So, why are some big publishers reluctant to use PhysX in their big games? Perhaps it’s a question of hardware.
The activation of PhysX on all GeForce 8000-series and newer cards in August 2008 brought to an end the days of having to buy a discrete card to handle physics calculations, instead moving these procedures onto the GPU.
In theory, that meant that the majority of gamers using Nvidia cards were able to unlock the hidden power of physics and suddenly populate their games with realistic sparks, gently swelling water and other impressive effects, increasing realism ten-fold. It also meant that using PhysX had never been cheaper or easier.
In practise, though, Nvidia needs a near-monopoly on hardware to make PhysX worthwhile – and its position in the discrete GPU market is looking more precarious now that it has done over the past couple of years. Various reports from the end of 2008 cite the release of ATI’s Radeon HD 4000-series as the major reason why ATI’s market share had grown from 35% to 40% in three months, forcing Nvidia to cut the prices of some of its flagship products in the meantime. In the same period, overall sales of desktop GPUs fell, shrinking the potential market for PhysX (and any other competing physics system) further.
With ATI coming up with superb GPUs and gaining market share, it’s become even less tempting for developers to spend hundreds of hours incorporating PhysX into their games while knowing that a large proportion of players won’t be able to see those effects in action.
The picture is no rosier on consoles, either. Both systems use older hardware and, while Nvidia makes its PhysX middleware engine available to developers on all three consoles, we’ve seen that plenty of developers would rather use their own engines or other third party tools instead of Nvidia’s technology.
In this sort of market – with ATI closing the gap and Intel’s Larrabee apparently on the horizon – it’s going to become more difficult for Nvidia to cram PhysX into enough games to make its purchase of Ageia a worthwhile venture.
At the moment, there’s little chance of most of the big developers and publishers using PhysX but, without widespread support, there’s no chance of Nvidia being able to grow PhysX into the dominant physics system that it surely hopes for. It’s a double-edged sword and, without a drastic shift, it’s difficult to see Nvidia making any sort of genuine progress in the physics market.
Tags: AMD, ati, graphics, intel, larrabee, Nvidia, physx
Posted in: Hardware
Follow any responses to this entry through the RSS 2.0 feed.
3 Responses to “ Will Nvidia PhysX ever be worthwhile? ”
Leave a Reply
Authors
- Barry Collins
- Chris Brennan
- Christine Horton
- Darien Graham-Smith
- Dave Stevenson
- Davey Winder
- David Bayon
- David Fearon
- Ewen Rankin
- Ian Devlin
- Jon Honeyball
- Jonathan Bray
- Kevin Partner
- Mike Jennings
- Nicole Kobie
- Sasha Muller
- Steve Cassidy
- Stewart Mitchell
- Stuart Turton
- Tim Danton
- Tom Arah
Categories
- About the bloggers
- Android App of the Week
- cloud computing
- Green
- Hardware
- How To
- iPhone App of the Week
- Just in
- Microsoft Office 2010
- Newsdesk
- Online business
- Random
- Rant
- Real World Computing
- Software
- View from the Labs
- Windows 7
- Windows 8
Archives
- February 2012
- January 2012
- December 2011
- November 2011
- October 2011
- September 2011
- August 2011
- July 2011
- June 2011
- May 2011
- April 2011
- March 2011
- February 2011
- January 2011
- December 2010
- November 2010
- October 2010
- September 2010
- August 2010
- July 2010
- June 2010
- May 2010
- April 2010
- March 2010
- February 2010
- January 2010
- December 2009
- November 2009
- October 2009
- September 2009
- August 2009
- July 2009
- June 2009
- May 2009
- April 2009
- March 2009
- February 2009
- January 2009
- December 2008
- November 2008
- October 2008
- September 2008
- August 2008
- July 2008
- June 2008
- May 2008
- April 2008
advertisement





June 30th, 2009 at 7:19 pm
[...] [...]
June 30th, 2009 at 8:36 pm
Surely what is needed to increase take up is for Microsoft to add a physics element into Direct X?
July 1st, 2009 at 6:58 pm
I ASSUME that there’s a bunch of useful physics commands that developers could agree upon to be added to a future Direct X or open standard alternative. However if OpenCL or an OpenCL like DX11 alternative is flexible and easy enough to use for physics another set of physics only API’s may not have any benefit.
PhysX is useless I know if I was a developer I’d want my games and tools to be based around an open standard with a future of hardware compatibility way into the future.
I hope OpenCL becomes the standard for games as it could potentially help allow a shift in future towards linux being the standard game platform.