28/12/2012 December 2012 OpenGL status
RSS Feed ~ Comment: by email - on Twitter ~ Share: Twitter - Facebook - Linked In - Google+ - Permanent link
This month I had my first experience with Intel OpenGL drivers since four or five years. Back then working on Intel was a nightmare, the implementation would report VBO supports but no entry point would be available... Intel has been more and more active on the GPU side so that having an Intel GPU on a laptop sounds like a reasonnable option to me... but is the OpenGL implementation descent?
The lastest drivers I found is the version 15.31.64.2885 which doesn't appears to be an official driver but it resolves a lot of issues regarding rendering to framebuffer objects. The implementation provides a full support of OpenGL 4.0 specification adn the work on OpenGL 4.1 and OpenGL 4.2 has began as we can see that ARB_ES2_compatibility (4.1), ARB_base_instance (4.2), ARB_conservative_depth (4.2), ARB_internalformat_query (4.2), ARB_map_buffer_alignment (4.2) and ARB_texture_storage (4.2) extensions are exposed.
More digging also shows that the work on more extensions has began as we can find entry points for ARB_get_program_binary (4.1), ARB_invalidate_subdata (4.3), ARB_shader_image_load_store (4.2), ARB_shader_atomic_counter (4.2) and ARB_separate_shader_objects (4.1).
At this point, the biggest issue is that ARB_debug_output is still not supported which makes debugging to painful for this modern age. ARB_debug_output support is definetely my number one feature request for this implementation!
For this new OpenGL status, I refactored the table using Excel to make it easier for me to maintain the status especially now that it includes Intel results. The complete and details status can be download at the end of this post. The PDF version doesn't show the comments embedded in the Excel version.
For this month status I am using Catalyst 12.11 beta 11 as AMD hasn't release new drivers. On NVIDIA side, I am testing Forceware 310.64 and for Intel drivers 15.31.64.2885. Lot of fixes for Intel regarding rendering to framebuffer, indirect draws and point sprite rendering.
All in all, I am quite surprized by the good quality of Intel OpenGL driver. Considering the size of Intel market share, it sounds reasonable to consider supporting this implementation by our software and the future sounds pretty promessing. Hopefully, in 2013 Intel will provide an OpenGL 4.2 implementation which will lead to some pretty interesting programming across all vendors.
Once again, don't forget to contribute to the OpenGL community by reporting your bugs!