Recent content by renderstate

Not on discrete GPUs. PowerVR does it IIRC, it's quite a natural fit for TBDRs, although I don't like much the idea of having programs that are not orthogonal to some render states (i.e. a render state change might force a partial or full shader recompile).

Same story.

Incorrect actually. The API doesn't specify tasks from different queues must be scheduled concurrently. Enforcing it it would make no sense since if a particular workload already nicely fills the GPU co-scheduling other work would cause oversubscription of HW resources, with the risk of...

It invalidates it because you are possibly slowing down some HW making official scores comparison unreliable. On the other hand if an IHV disables concurrent execution (or the HW doesn't support it in the first place) the score is still valid, of course.

It depends on how state is handled. If you keep state stored in fixed size on chip memory/context and you run out of it you will stall.

Or more likely the pipeline state hasn't changed at all.

IIRC It's the lowest common denominator that includes GCN, that still doesn't support conservative raster and raster order views. It would have been nice to see those new two features being used in this benchmark :(

So I am making a technical argument and you reply with that? Please put me on your ignored list so you can save time and effort.

It would be nice if people would stop conflating async compute with instruction scheduling. How many times has to be explained/repeated that they are not related?!? If you think they are you still don't get what async compute is. Also stop this FUD about Pascal not having a HW task scheduler...

Quite the contrary. It's very likely that host and shader code heavily optimized for GCN might heavily penalize NVIDIA HW and viceversa.

Apple is already abandoning OpenCL for compute on Metal. OpenCL is a slowly dying API.

Fireaxis has always been in very good terms with AMD. Several of their employees are former AMD engineers and researchers. Dan Baker of AotS "fame" is a big AMD supporter (often partecipates to AMD events) and used to work there. This partnership really makes sense given how both companies go...

Mantle was definitely a big success for AMD since it has heavily influenced DX12 and Vulkan. I'd argue it's in fact a good thing for AMD that Mantle has never seen the light of the day as it eliminates for them having to support yet another API and therefore save money. DX12 and Vulkan carry the...

Polaris runs half precision ops at the same rate of single precision ops, there is no 2X rate. If you use half precision math in DX/OpenGL NVIDIA will run it at full precision on gaming parts, so there is no perf penalty. Are you saying that statements made by NVIDIA in the past should reflect...

Will it invalidate reviews/affect performance?