公开(公告)号：US20190294439A1

公开(公告)日：2019-09-26

申请号：US15934179

申请日：2018-03-23

Applicant: Arm Limited

Inventor： Kenneth Edvard Ostby

IPC: G06F9/30 ,  G06F9/38 ,  G06T1/20

Abstract: A data processor in which execution threads may be grouped together into thread groups in which the plural threads of a thread group can each execute a set of instructions in lockstep, one instruction at a time. The data processor comprises a plurality of execution lanes for executing respective execution threads of a thread group. For each thread group in a pool 51 of thread groups available to be issued to the execution lanes, an indication 54 of the active threads of the thread group is stored, and sets of at least one thread group from the pool 51 of available thread groups to issue 73 to the execution lanes for execution are selected 72 based on the indications of the active threads for the thread groups in the thread group pool.

公开(公告)号：US09286714B2

公开(公告)日：2016-03-15

申请号：US13922471

申请日：2013-06-20

Applicant: ARM LIMITED

Inventor： Frode Heggelund ,  Kenneth Edvard Ostby

IPC: G06T15/30 ,  G06T11/40 ,  G06T15/40

CPC classification number: G06T15/40 ,  G06T15/30 ,  G06T2200/28

Abstract: A method and apparatus includes primitive setup circuitry for determining a plurality of functions for an input graphics primitive, including an edge function associated with each edge of the input graphics primitive and a depth function associated with the input graphics primitive. Rasterization circuitry performs a rasterization operation in order to calculate position data for a plurality of graphics fragments to be used to represent the input graphics primitive. In a default mode of operation, depth bound clipping circuitry performs a depth bound clipping operation by determining, for each graphics fragment in said plurality of graphics fragments, a depth value for said graphics fragment using the depth function, and determining whether said depth value resides within a valid depth range of a view frustum, the graphics fragment being discarded from further processing if its depth value does not reside within said valid depth range.

Abstract translation: 方法和装置包括用于确定输入图形原语的多个功能的原始设置电路，包括与输入图形基元的每个边缘相关联的边缘函数和与输入图形基元相关联的深度函数。 栅格化电路执行光栅化操作以便计算要用于表示输入图形原语的多个图形片段的位置数据。 在默认操作模式下，深度限制剪辑电路通过使用深度函数确定所述多个图形片段中的每个图形片段的所述图形片段的深度值，并且确定所述深度值是否位于 在视锥体的有效深度范围内，如果其深度值不在所述有效深度范围内，则图形片段将从进一步处理中被丢弃。

公开(公告)号：US10606595B2

公开(公告)日：2020-03-31

申请号：US15934179

申请日：2018-03-23

Applicant: Arm Limited

Inventor： Kenneth Edvard Ostby

IPC: G06F9/30 ,  G06T1/20 ,  G06F9/38

Abstract: A data processor in which execution threads may be grouped together into thread groups in which the plural threads of a thread group can each execute a set of instructions in lockstep, one instruction at a time. The data processor comprises a plurality of execution lanes for executing respective execution threads of a thread group. For each thread group in a pool 51 of thread groups available to be issued to the execution lanes, an indication 54 of the active threads of the thread group is stored, and sets of at least one thread group from the pool 51 of available thread groups to issue 73 to the execution lanes for execution are selected 72 based on the indications of the active threads for the thread groups in the thread group pool.

公开(公告)号：US20200097061A1

公开(公告)日：2020-03-26

申请号：US16137537

申请日：2018-09-20

Applicant: Arm Limited

Inventor： Kenneth Edvard Ostby ,  Andrew Burdass

IPC: G06F1/32 ,  G06T1/20 ,  G06F9/38

Abstract: A graphics processing unit is operable to execute graphics processing programs comprising sequences of instructions to perform graphics processing operations. The graphics processing unit includes execution processing circuitry operable to execute instructions to perform graphics processing operations and instruction issuing circuitry operable to issue instructions to be executed to the execution processing circuitry. The graphics processing unit also includes energy management circuitry operable to monitor the energy usage by the execution processing circuitry when executing instructions, determine, based on the monitoring of the energy usage, a permitted energy usage range for the execution processing circuitry when executing instructions for a future time period, and control the issuing of instructions to the execution processing circuitry by the instruction issuing circuitry during the future time period based on the permitted energy usage range determined for the future time period.

公开(公告)号：US10127626B1

公开(公告)日：2018-11-13

申请号：US15657014

申请日：2017-07-21

Applicant: ARM Limited

Inventor： Kenneth Edvard Ostby

IPC: G06T1/00 ,  G06T15/00 ,  G06T1/20 ,  G06T1/60 ,  G06F9/30

Abstract: In a data processing system, a program to be executed by a programmable processing unit of the data processing system is analyzed to identify a sequence of instructions that would produce the same result for plural execution threads were those plural execution threads each to execute the sequence of instructions using the same input data. Then, when the program is being executed, when an execution thread is to execute the identified sequence of instructions, it is determined whether a result produced by an earlier execution thread executing the sequence of instructions, and that used the same input data, is stored in memory or not. The current thread then either executes the sequence of instructions, or retrieves the stored result produced by the earlier execution of the sequence of instructions and skips execution of the sequence of instructions for which the result is stored, accordingly.