Typed unordered access view overloading on pixel pipeline
Methods, systems and apparatuses provide for graphics processor technology that routes untyped unordered access view (UAV) messages to a next level memory cache, routes typed UAV messages and render target messages to a pixel pipeline, and processes, via the pixel pipeline, the typed UAV messages. The technology can also provide for the pixel pipeline to perform a format conversion of one or more pixels associated with a typed UAV message based on a surface format of a UAV resource, calculate a memory address for each pixel associated with the typed UAV message, and collect a plurality of fragments from processed typed UAV messages.
1. A computing system comprising:
a network controller; and
a graphics processor coupled to the network controller, wherein the graphics processor includes logic coupled to one or more substrates, the logic comprising:
a pixel pipeline to process typed unordered access view (UAV) messages; and
a message router to:
route untyped UAV messages to a next level memory cache; and
route typed UAV messages and render target messages to the pixel pipeline.
2. The computing system of claim 1 , wherein to process the typed UAV messages, the pixel pipeline is to:
perform a format conversion of one or more pixels associated with a typed UAV message based on a surface format of a UAV resource;
calculate a memory address for each pixel associated with the typed UAV message; and
collect a plurality of fragments from processed typed UAV messages.
3. The computing system of claim 2 , wherein to perform a format conversion of one or more pixels associated with a typed UAV message comprises to convert a 32 bit per channel format to one of an 8 bit per channel format or a 16 bit per channel format.
4. The computing system of claim 2 , wherein to calculate a memory address for each pixel associated with the typed UAV message comprises using location coordinates and surface properties associated with each respective pixel.
5. The computing system of claim 2 , wherein the plurality of fragments from processed typed UAV messages are collected based on an address boundary associated with each of the fragments.
6. The computing system of claim 5 , wherein the plurality of fragments are collected so long as the address of each fragment falls within a threshold address range, and wherein the pixel pipeline is further to route the collected plurality of fragments to the next level memory cache once a next fragment falls outside of the threshold address range.
7. A semiconductor apparatus comprising:
one or more substrates; and
logic coupled to the one or more substrates, the logic implemented at least partly in one or more of configurable logic or fixed-functionality hardware logic, the logic comprising:
a pixel pipeline to process typed unordered access view (UAV) messages; and
a message router to:
route untyped UAV messages to a next level memory cache; and
route typed UAV messages and render target messages to the pixel pipeline.
8. The apparatus of claim 7 , wherein to process the typed UAV messages, the pixel pipeline is to:
perform a format conversion of one or more pixels associated with a typed UAV message based on a surface format of a UAV resource;
calculate a memory address for each pixel associated with the typed UAV message; and
collect a plurality of fragments from processed typed UAV messages.
9. The apparatus of claim 8 , wherein to perform a format conversion of one or more pixels associated with a typed UAV message comprises to convert a 32 bit per channel format to one of an 8 bit per channel format or a 16 bit per channel format.
10. The apparatus of claim 8 , wherein to calculate a memory address for each pixel associated with the typed UAV message comprises using location coordinates and surface properties associated with each respective pixel.
11. The apparatus of claim 8 , wherein the plurality of fragments from processed typed UAV messages are collected based on an address boundary associated with each of the fragments.
12. The apparatus of claim 11 , wherein the plurality of fragments are collected so long as the address of each fragment falls within a threshold address range, and wherein the pixel pipeline is further to route the collected plurality of fragments to the next level memory cache once a next fragment falls outside of the threshold address range.
13. At least one non-transitory computer readable storage medium comprising a set of instructions which, when executed by a graphics apparatus, cause the graphics apparatus to:
route untyped unordered access view (UAV) messages to a next level memory cache;
route typed UAV messages and render target messages to a pixel pipeline; and
process, by the pixel pipeline, the typed UAV messages.
14. The at least one non-transitory computer readable storage medium of claim 13 , wherein to process the typed UAV messages, the pixel pipeline is to:
perform a format conversion of one or more pixels associated with a typed UAV message based on a surface format of a UAV resource;
calculate a memory address for each pixel associated with the typed UAV message; and
collect a plurality of fragments from processed typed UAV messages.
15. The at least one non-transitory computer readable storage medium of claim 14 , wherein to perform a format conversion of one or more pixels associated with a typed UAV message comprises to convert a 32 bit per channel format to one of an 8 bit per channel format or a 16 bit per channel format.
16. The at least one non-transitory computer readable storage medium of claim 14 , wherein to calculate a memory address for each pixel associated with the typed UAV message comprises using location coordinates and surface properties associated with each respective pixel.
17. The at least one non-transitory computer readable storage medium of claim 14 , wherein the plurality of fragments from processed typed UAV messages are collected based on an address boundary associated with each of the fragments.
18. The at least one non-transitory computer readable storage medium of claim 17 , wherein the plurality of fragments are collected so long as the address of each fragment falls within a threshold address range, and wherein the instructions, when executed, further cause the graphics apparatus to route the collected plurality of fragments to the next level memory cache once a next fragment falls outside of the threshold address range.
19. A method comprising:
routing untyped unordered access view (UAV) messages to a next level memory cache;
routing typed UAV messages and render target messages to a pixel pipeline; and
processing, by the pixel pipeline, the typed UAV messages.
20. The method of claim 19 , wherein processing the typed UAV messages comprises:
performing a format conversion of one or more pixels associated with a typed UAV message based on a surface format of a UAV resource;
calculating a memory address for each pixel associated with the typed UAV message; and
collecting a plurality of fragments from processed typed UAV messages.
21. The method of claim 20 , wherein performing a format conversion of one or more pixels associated with a typed UAV message comprises converting a 32 bit per channel format to one of an 8 bit per channel format or a 16 bit per channel format.
22. The method of claim 20 , wherein calculating a memory address for each pixel associated with the typed UAV message comprises using location coordinates and surface properties associated with each respective pixel.
23. The method of claim 20 , wherein the plurality of fragments from processed typed UAV messages are collected based on an address boundary associated with each of the fragments.
24. The method of claim 23 , wherein the plurality of fragments are collected so long as the address of each fragment falls within a threshold address range, and further comprising routing the collected plurality of fragments to the next level memory cache once a next fragment falls outside of the threshold address range.