Explore >> Select a destination
|
You are here 
|
logins.github.io | ||
| | | | |
www.gfxstrand.net
|
|
| | | | | [AI summary] The post discusses the complexities and challenges of descriptor sets in graphics APIs like Vulkan and D3D12, focusing on hardware differences and the trade-offs between various descriptor binding methods. | |
| | | | |
therealmjp.github.io
|
|
| | | | | Basics of GPU Memory Integrated/UMA GPUs Dedicated/NUMA GPUs How It Works In D3D12 Common Patterns in D3D12 Textures And The Two-Step Upload Should We Upload Buffers? Working With The COPY Queue Two COPY Queues Are Better Than One? Allocating Staging Memory What About DirectStorage? Results From My Testing App CPU Write Performance CPU Read Performance GPU Read Performance, Normal Access GPU Read Performance, Non-Coalesced Access GPU Read Performance, Various Buffer Sizes Conclusion When the monkey's paw granted our wish for lower-level/explicit graphics APIs, one of the consequences was that we were much more directly exposed to the fact that GPUs can have their own separate set of physical memory. | |
| | | | |
themaister.net
|
|
| | | | | [AI summary] The author discusses the complexities and challenges of implementing D3D12's binding model in vkd3d-proton, highlighting the intricacies of bindless resources, local root signatures, and the transition to SM 6.6. They emphasize the need for careful handling of descriptors, validation, and the impact of hardware differences. The author also mentions the benefits of AMD's GCN architecture for bindless operations and the importance of thorough testing and code generation for different GPU architectures. | |
| | | | |
blog.selfshadow.com
|
|
| | | [AI summary] This text discusses advanced techniques for occlusion culling and visibility determination in computer graphics, particularly focusing on GPU and SPU implementations. It outlines methods such as hierarchical z-buffering, HZB (Hierarchical Z-Buffer) sampling, and frustum subdivision for efficient rendering of large environments. The text also touches on challenges like latency, hardware limitations, and future directions for visibility processing, including potential integration with next-generation hardware. | ||
