Apple Inc.

Advanced Micro Devices Corporation

Ati Technologies Ulc

Ati Technologies Inc.

Jeffrey T Brady

Michael Mantor

Andrew M Havlir

Angel E Socarras

Terence M Potter

Ralph Clay Taylor

Frank W Liljeros

Christopher L Spencer

Benjiman L Goodman

Dimitri Tan

Andrew Evan Gruber

Mark Fowler

Rex Eldon McCrary

Yoong Chert Foo

Michael Andrew Mang

Daniel Wong

Jason David Carroll

Karl D Mann

Benjamin Bowman

Brian A Buchner

Brian K Reynolds

Anjana Rajendran

Jingfei Kong

Marcos Zini

Jeffrey A Lohman

Daniel Clifton

Adil M Sadik

Mark D Earl

Jeffrey M Broton

Sindhuja Sethuraman

Michael J Mantor

Techniques are disclosed relating to thread preemption in the context of memory-backed registers. In some embodiments, a memory hierarchy includes one or more cache levels and one or more memory circuits. Execution circuitry may operate on operands in architectural registers to execute instructions of threads, where data for the architectural registers is stored and backed by the memory hierarchy. Control circuitry may, in response to a context switch indication for a given thread: flush and invalidate a set of architectural register data from a first cache level and store memory page information (e.g., a page catalog base address) associated with the set of architectural register data.

Preemption techniques for memory-backed registers

Techniques are disclosed relating to using an initial version of an object shader to determine a child count and distribute geometry work based on the child count. In some embodiments, graphics shader circuitry is configured to execute shader programs including object shaders and mesh shaders. Vertex control circuitry is configured to, for a given object shader: launch an initial version of the given object shader to determine a number of meshlets to be generated by the given object shader (e.g., where the initial version of the given object shader does not commit side effects to architectural state of the apparatus) and select shader circuitry to execute a complete version of the given object shader based on the determined number of meshlets.

Initial object shader run for graphics workload distribution

Methods and systems are described. A system includes a redundant shader pipe array that performs rendering calculations on data provided thereto and a shader pipe array that includes a plurality of shader pipes, each of which performs rendering calculations on data provided thereto. The system also includes a circuit that identifies a defective shader pipe of the plurality of shader pipes in the shader pipe array. In response to identifying the defective shader pipe, the circuit generates a signal. The system also includes a redundant shader switch. The redundant shader switch receives the generated signal, and, in response to receiving the generated signal, transfers the data for the defective shader pipe to the redundant shader pipe array.

Redundancy method and apparatus for shader column repair

Techniques are disclosed relating to context switching in a SIMD processor. In some embodiments, an apparatus includes pipeline circuitry configured to execute graphics instructions included in threads of a group of single-instruction multiple-data (SIMD) threads in a thread group. In some embodiments, context switch circuitry is configured to atomically: save, for the SIMD group, a program counter and information that indicates whether threads in the SIMD group are active using one or more context switch registers, set all threads to an active state for the SIMD group, and branch to handler code for the SIMD group. In some embodiments, the pipeline circuitry is configured to execute the handler code to save context information for the SIMD group and subsequently execute threads of another thread group. Disclosed techniques may allow instruction-level context switching even when some SIMD threads are non-active.

Instruction-level context switch in SIMD processor

Techniques are disclosed relating to fetching items from a compute command stream that includes compute kernels. In some embodiments, stream fetch circuitry sequentially pre-fetches items from the stream and stores them in a buffer. In some embodiments, fetch parse circuitry iterate through items in the buffer using a fetch parse pointer to detect indirect-data-access items and/or redirect items in the buffer. The fetch parse circuitry may send detected indirect data accesses to indirect-fetch circuitry, which may buffer requests. In some embodiments, execute parse circuitry iterates through items in the buffer using an execute parse pointer (e.g., which may trail the fetch parse pointer) and outputs both item data from the buffer and indirect-fetch results from indirect-fetch circuitry for execution. In various embodiments, the disclosed techniques may reduce fetch latency for compute kernels.

Low latency fetch circuitry for compute kernels

Techniques are disclosed relating to caches that support transient storage fields for cache entries. In some embodiments, cache circuitry includes a set of multiple cache entries that each include a tag field and a data field. In some embodiments, transient storage circuitry includes a transient storage field for each of the multiple cache entries. In some embodiments, cache control circuitry stores received first data in the data field of a cache entry and stores received transient data in a corresponding transient storage field. In response to an eviction determination for the cache entry, however, the cache control circuitry may write the first data but not the transient data to a backing memory for the cache circuitry. In various embodiments, disclosed techniques may allow caching additional data that is transient without increasing bandwidth to the backing memory.

Cache memory with transient storage for cache lines

In various embodiments, a resource allocation management circuit may allocate a plurality of different types of hardware resources (e.g., different types of registers) to a plurality of threads. The different types of hardware resources may correspond to a plurality of hardware resource allocation circuits. The resource allocation management circuit may track allocation of the hardware resources to the threads using state identification values of the threads. In response to determining that fewer than a respective requested number of one or more types of the hardware resources are available, the resource allocation management circuit may identify one or more threads for deallocation. As a result, the hardware resource allocation system may allocate hardware resources to threads more efficiently (e.g., may deallocate hardware resources allocated to fewer threads), as compared to a hardware resource allocation system that does not track allocation of hardware resources to threads using state identification values.

Hardware resource allocation system for allocating resources to threads

Techniques are disclosed relating to context switching using distributed compute workload parsers. In some embodiments, an apparatus includes a plurality of shader units configured to perform operations for compute workgroups included in compute kernels, a plurality of distributed workload parser circuits each configured to dispatch workgroups to a respective set of the shader units, a communications fabric, and a master workload parser circuit configured to communicate with the distributed workload parser circuits via the communications fabric. In some embodiments, the master workload parser circuit maintains a first set of master state information that does not change for a compute kernel based on operations by the shader units and a second set of master state information that may be changed by operations specified by the kernel. In some embodiments, the master workload parser circuit performs a multi-phase state storage process in communications with the distributed workload parser circuits.

Techniques for context switching using distributed compute workload parsers

Methods, systems and non-transitory computer readable media are described. A system includes a shader pipe array, a redundant shader pipe array, a sequencer and a redundant shader switch. The shader pipe array includes multiple shader pipes, each of which perform rendering calculations on data provided thereto. The redundant shader pipe array also performs rendering calculations on data provided thereto. The sequencer identifies at least one defective shader pipe in the shader pipe array, and, in response, generates a signal. The redundant shader switch receives the generated signal, and, in response, transfers the data destined for each shader pipe identified as being defective independently to the redundant shader pipe array.

Techniques are disclosed relating to distributing work from compute kernels using distributed parser circuitry. In some embodiments, a master parser is configured to communicate with distributed parsers over a communications fabric. In some embodiments, the master parser generates batches of compute workgroups from a compute kernel and assigns batches to ones of the distributed workload parser circuits. In some embodiments, based on an indicated number of sequential workgroups to send to the same distributed workload parser, the master parser selects a distributed parser to receive a set of batches and avoids selecting a distributed workload parser whose queue would be filled by the batches. In some embodiments, this may avoid stalling for a kernel terminate command. In some embodiments, the master parser may adjust a batch size to avoid filling a distributed parser's queue. In some embodiments, the distributed parsers may use similar techniques when sending work to shader units.

Distributing compute work using distributed parser circuitry

An example system includes a plurality of execution units, a shared resource, and an allocation control circuit. Each execution unit may generate a resource allocation request that includes a resource allocation size. The allocation control circuit may select a particular resource allocation request from the plurality of resource allocation requests, and determine an availability, based on an allocation register, of contiguous resource blocks within the shared resource. In response to determining that a number of the contiguous resource blocks satisfies a requested allocation size, the allocation control circuit may select an address corresponding to a particular resource block of the one or more contiguous resource blocks, and allocate the resource blocks to a corresponding execution unit. In response to a beginning of a second system clock cycle, the allocation control circuit may also update the allocation register based on the selected address and the requested allocation size.

Pipelined resource allocation using size aligned allocation

Techniques are disclosed relating to distributing work from compute kernels using a distributed hierarchical parser architecture. In some embodiments, an apparatus includes a plurality of shader units configured to perform operations for compute workgroups included in compute kernels processed by the apparatus, a plurality of distributed workload parser circuits, and a communications fabric connected to the plurality of distributed workload parser circuits and a master workload parser circuit. In some embodiments, the master workload parser circuit is configured to iteratively determine a next position in multiple dimensions for a next batch of workgroups from the kernel and send batch information to the distributed workload parser circuits via the communications fabric to assign the batch of workgroups. In some embodiments, the distributed parsers maintain coordinate information for the kernel and update the coordinate information in response to the batch information, even when the distributed parsers are not assigned to execute the batch.

Distributed compute work parser circuitry using communications fabric

Techniques are disclosed relating to managing dependencies in a compute control stream that specifies operations to be performed on a programmable shader (e.g., of a graphics unit). In some embodiments, the compute control stream includes commands and kernels. In some embodiments, dependency circuitry is configured to maintain dependencies such that younger kernels are allowed to execute ahead of a type of cache-related command (e.g., a command that signals a cache flush and/or invalidate). Disclosed circuitry may include separate buffers for commands and kernels, command dependency circuitry, and kernel dependency circuitry. In various embodiments, the disclosed architecture may improve performance in a highly scalable manner.

Growing community of inventors

Orlando, FL, United States of America

Jeffrey T Brady