English | 中文 | Русский | Français
2,856 Posts served
8,606 Conversations started
Hello media developers. This is a follow up to my original post asking you about what your pain points are for media development. I've recently been thinking about some of the challenges of hardware accelerated video decode with DXVA. A whitepaper was written to help developers access H.264 decode in our current G45/GM45 chipsets, and talking with some developers it seems that DXVA is not as simple as one might think. The benefits of HW acceleration for decode are interesting to me from a potential increased battery life perspective as well as in more professional editing applications where someone might be working with a dozen or more files at once and fast decode for preview, etc would be helpful in streamlining the workflow. I'm wondering what some of you think.
First of all, how important is HW acceleration for video decode to you, and secondly, have you had challenges in using DXVA that could be simplified?
| July 23, 2009 4:34 PM PDT
Craig Hurst (Intel)
| Yes. We have looked at several media frameworks where this would be interesting, but we are focussing on DirectShow and Media Foundation first. Open source remains an interesting area for exploration. |
| October 14, 2009 9:20 PM PDT
Dave Abrams |
I'm working on an IP video wall to display 50+ H.264 streams from mega-pixel surveillance video cameras. Can the Media SDK do multiple decoding in multi-core, or is it just designed for one video (e.g. laptops) at a time? Can you suggest a graphics card combination with chipset for multiple camera decode on more than one monitor display? |

Che Kristo