International Business Machines Corporation

Apple Inc.

Globalfoundries Inc.

Benjiman L Goodman

Jeffrey A Stuecheli

William John Starke

Guy L Guthrie

Stephen J Powell

Praveen S Reddy

Jody B Joyner

John Steven Dodson

Eric Eugene Retter

Ravi Kumar Arimilli

James Stephen Fields, Jr

Terence M Potter

Derek E Williams

David Michael Daly

Hillery C Hunter

Kenneth Lee Wright

Aaron Christoph Sawdey

Robert D Kenney

Yoong Chert Foo

Samuel I Ward

Harrison M McCreary

Anjana Rajendran

Paul Frank Lecocq

Linton B Ward, Jr

Luis Alfonso Lastras-Montano

Marc Avery Gollub

Steven Leonard Roberts

Miles Robert Dooley

Sujatha Kashyap

Mark Andrew Brittain

Jeffrey T Brady

Vicente Enrique Chung

Steven J Hnatko

Winnie W Yeung

Joshua P Hernandez

Yuen C Tschang

Jeffrey A Lohman

Michael Stephen Floyd

Naresh Nayar

James Han Wang

James S Blomgren

Leo James Clark

Phillip G Williams

Liang-Kai Wang

Tatsuya Iwamoto

Nathaniel Douglas Hieter

Jonathan M Redshaw

Kutty Banerjee

William V Miller

Steven Powell

Milton Devon Miller, Ii

Frank W Liljeros

Karl D Mann

Brian K Reynolds

Jeremy T Hopkins

Dzung Q Vu

Zelin Zhang

Cheng Li

Aroun Demeure

Liang Xia

Justin Friesenhahn

Adam T Moerschell

Donald R Desota

Gregory D Roberts

Sertac Cakici

Sheenam Jayaswal

David K Li

Mark I Luffel

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 multi-stage thread scheduling. In some embodiments, processor circuitry includes multiple channel pipelines for multiple channels and multiple execution pipelines shared by the channel pipelines and configured to perform different types of operations provided by the channel pipelines. First scheduler circuitry may arbitrate among threads to assign threads to channels. Second scheduler circuitry may arbitrate among channels to assign an operation from a given channel to a given execution pipeline. The execution pipelines may provide backpressure information to the first scheduler circuitry based on execution status and the first scheduler circuitry may adjust priority of a thread for assignment to a channel based on the backpressure information. Disclosed techniques may reduce channel conflicts and starvation for execution resources.

Multi-stage thread scheduling

Techniques are disclosed relating to eviction control for cache lines that store register data. In some embodiments, memory hierarchy circuitry is configured to provide memory backing for register operand data in one or more cache circuits. Lock circuitry may control a first set of lock indicators for a set of registers for a first thread, including to assert one or more lock indicators for registers that are indicated, by decode circuitry, as being utilized by decoded instructions of the first thread. The lock circuitry may preserve register operand data in the one or more cache circuits, including to prevent eviction of a given cache line from a cache circuit based on an asserted lock indicator. The lock circuitry may clear the first set of lock indicators in response to a reset event. Disclosed techniques may advantageously retain relevant register information in the cache with limited control circuit area.

Cache control to preserve register data

Techniques are disclosed relating to instruction scheduling in the context of instruction cache misses. In some embodiments, first-stage scheduler circuitry is configured to assign threads to channels and second-stage scheduler circuitry is configured to assign an operation from a given channel to a given execution pipeline based on decode of an operation for that channel. In some embodiments, thread replacement circuitry is configured to, in response to an instruction cache miss for an operation of a first thread assigned to a first channel, deactivate the first thread from the first channel.

Thread channel deactivation based on instruction cache misses

Techniques are disclosed relating to channel stalls or deactivations based on the latency of prior operations. In some embodiments, a processor includes a plurality of channel pipelines for a plurality of channels and a plurality of execution pipelines shared by the channel pipelines and configured to perform different types of operations provided by the channel pipelines. First scheduler circuitry may assign threads to channels and second scheduler circuitry may assign an operation from a given channel to a given execution pipeline based on decode of an operation for that channel. Dependency circuitry may, for a first operation that depends on a prior operation that uses one of the execution pipelines, determine, based on status information for the prior operation from the one of the execution pipelines, whether to stall the first operation or to deactivate a thread that includes the first operation from its assigned channel.

Fence enforcement techniques based on stall characteristics

Techniques are disclosed relating to cache footprint management. In some embodiments, execution circuitry is configured to perform operations for instructions from multiple threads in parallel. Cache circuitry may store information operated on by threads executed by the execution circuitry. Scheduling circuitry may arbitrate among threads to schedule threads for execution by the execution circuitry. Tracking circuitry may determine one or more performance metrics for the cache circuitry. Control circuitry may, based on the one or more performance metrics meeting a threshold, reduce a limit on a number of threads considered for arbitration by the scheduling circuitry, to control a footprint of information stored by the cache circuitry. Disclosed techniques may advantageously reduce or avoid cache thrashing for certain processor workloads.

Cache footprint management

Techniques are disclosed relating to private memory management using a mapping thread, which may be persistent. In some embodiments, a graphics processor is configured to generate a pool of private memory pages for a set of graphics work that includes multiple threads. The processor may maintain a translation table configured to map private memory addresses to virtual addresses based on identifiers of the threads. The processor may execute a mapping thread to receive a request to allocate a private memory page for a requesting thread, select a private memory page from the pool in response to the request, and map the selected page in the translation table for the requesting. The processor may then execute one or more instructions of the requesting thread to access a private memory space, wherein the execution includes translation of a private memory address to a virtual address based on the mapped page in the translation table. The mapping thread may be a persistent thread for which resources are allocated for an entirety of a time interval over which the set of graphics work is executed.

Private memory management using utility thread

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 a thread-group-scoped gate instruction. In some embodiments, graphics processor circuitry is configured to execute, for multiple SIMD groups of a thread group, a graphics program that includes a gate instruction. During execution of the gate instruction for a first SIMD group, the processor accesses state information to determine that a threshold number of other SIMD groups in the thread group have not yet executed the gate instruction. Based on the determination, the processor executes a particular set of instructions of the graphics program for the first SIMD group (that is not executed by one or more other SIMD groups that reach the gate instruction after the first SIMD group). For example, the particular set of instructions may be a utility program that performs one or more operations for the entire thread group but is only executed by a subset of the SIMD groups.

Thread-group-scoped gate instruction

Techniques are disclosed relating to scheduling tasks for graphics processing. In one embodiment, a graphics unit is configured to render a frame of graphics data using a plurality of pass groups and the frame of graphics data includes a plurality of frame portions. In this embodiment, the graphics unit includes scheduling circuitry configured to receive a plurality of tasks, maintain pass group information for each of the plurality of tasks, and maintain relative age information for the plurality of frame portions. In this embodiment, the scheduling circuitry is configured to select a task for execution based on the pass group information and the age information. In some embodiments, the scheduling circuitry is configured to select tasks from an oldest frame portion and current pass group before selecting other tasks. This scheduling approach may result in efficient execution of various different types of graphics workloads.

GPU task scheduling

A data queuing and format apparatus is disclosed. A first selection circuit may be configured to selectively couple a first subset of data to a first plurality of data lines dependent upon control information, and a second selection circuit may be configured to selectively couple a second subset of data to a second plurality of data lines dependent upon the control information. A storage array may include multiple storage units, and each storage unit may be configured to receive data from one or more data lines of either the first or second plurality of data lines dependent upon the control information.

Data alignment and formatting for graphics processing unit

In general, techniques are disclosed for tracking and allocating graphics processor hardware over specified periods of time. More particularly, hardware sensors may be used to determine the utilization of graphics processor hardware after each of a number of specified intervals (referred to as 'sample intervals'). The utilization values so captured may be combined after a first number of sample intervals (the combined interval referred to as an 'epoch interval') and used to determine a normalized utilization of the graphic processor's hardware resources. Normalized epoch utilization values have been adjusted to account for resources used by concurrently executing processes. In some embodiments, a lower priority process that obtains and fails to release resources that should be allocated to one or more higher priority processes may be detected, paused, and its hardware resources given to the higher priority processes.

GPU resource priorities based on hardware utilization

A method and apparatus for saving power in integrated circuits is disclosed. An IC includes functional circuit blocks which are not placed into a sleep mode when idle. A power management circuit may monitor the activity levels of the functional circuit blocks not placed into a sleep mode. When the power management circuit detects that an activity level of one of the non-sleep functional circuit blocks is less than a predefined threshold, it reduce the frequency of a clock signal provided thereto by scheduling only one pulse of a clock signal for every N pulses of the full frequency clock signal. The remaining N&#x2212;1 pulses of the clock signal may be inhibited. If a high priority transaction inbound for the functional circuit block is detected, an inserted pulse of the clock signal may be provided to the functional unit irrespective of when a most recent regular pulse was provided.

Power saving with dynamic pulse insertion

Techniques are disclosed relating to a cache configured to store state information for texture mapping. In one embodiment, a texture state cache includes a plurality of entries configured to store state information relating to one or more stored textures. In this embodiment, the texture state cache also includes texture processing circuitry configured to retrieve state information for one of the stored textures from one of the entries in the texture state cache and determine pixel attributes based on the texture and the retrieved state information. The state information may include texture state information and sampler state information, in some embodiments. The texture state cache may allow for reduced rending times and power consumption, in some embodiments.

Texture state cache

Techniques for handling queuing of memory accesses prevent passing excessive requests that implicate a region of memory subject to a high latency memory operation, such as a memory refresh operation, memory scrubbing or an internal bus calibration event, to a re-order queue of a memory controller. The memory controller includes a queue for storing pending memory access requests, a re-order queue for receiving the requests, and a control logic implementing a queue controller that determines if there is a collision between a received request and an ongoing high-latency memory operation. If there is a collision, then transfer of the request to the re-order queue may be rejected outright, or a count of existing queued operations that collide with the high latency operation may be used to determine if queuing the new request will exceed a threshold number of such operations.

Memory queue handling techniques for reducing impact of high-latency memory operations

Growing community of inventors

Cedar Park, TX, United States of America

Benjiman L Goodman