Use big cores for compute intensive tasks.Avoid disk I/O or blocking operations on main/render threads.Avoid single-thread 100% CPU Utilization bottleneck.How to ensure Vulkan Subpasses merged properly?.What is the recommended usage SSBO vs.Is there a performance benefit or cost to using push constants?.Is there a performance impact of using Vulkan Secondary Command Buffers?.If a primitive spans multiple tiles, will the GPU insert synthetic vertices at tile boundaries?.Is binning affected by fragment Z occlusion?.Is the full Vertex Shader used when performing binning?.What conditions trigger direct rendering with FlexRender?.Do you have any details of how the tiling and binning process works?.What is the performance of 1010102 vs.Is float or half float more performant for 8-bit texture lookup?.What cache levels do textures remain compressed/decompressed?.What depth buffer format provides the best performance?.What is the effect on LRZ if discard/clip is used?.Can LRZ buffers and visibility streams be stored and reused?.What is the behavior on dynamic branching in fragment shaders?.Best single pass stereo? Texture array? Need GS?.Which has better performance, alpha test vs alpha blend?.What is the performance of user clip planes?.How are timer queries calculated in Adreno GPUs?.

What is the Occlusion Query performance?.How many Occlusion Queries should I use?.Which is better: vertex stream or attribute fetching in VS?.Are there “no-copy’ paths available for other Snapdragon hardware blocks?.What is the optimal way to sort objects? Is front-to-back object submission needed or, given the tiling architecture, is that not necessary?.Vertex texture access or vertex texture fetch.Percentage Closer Filtering for depth textures.FlexRender™ technology (Hybrid Deferred and Direct Rendering mode).

Opengl es 2.0 apps