专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明申请

US20220058025A1 THROTTLING WHILE MANAGING UPSTREAM RESOURCES 有权
公开(公告)号：US20220058025A1
公开(公告)日：2022-02-24
申请号：US17519902
申请日：2021-11-05
申请人： Advanced Micro Devices, Inc.
发明人： Paul James Moyer , Douglas Benson Hunt , Kai Troester
IPC分类号： G06F9/38
摘要： Systems, apparatuses, and methods for arbitrating threads in a computing system are disclosed. A computing system includes a processor with multiple cores, each capable of simultaneously processing instructions of multiple threads. When a thread throttling unit receives an indication that a shared cache has resource contention, the throttling unit sets a threshold number of cache misses for the cache. If the number of cache misses exceeds this threshold, then the throttling unit notifies a particular upstream computation unit to throttle the processing of instructions for the thread. After a time period elapses, if the cache continues to exceed the threshold, then the throttling unit notifies the upstream computation unit to more restrictively throttle the thread by performing one or more of reducing the selection rate and increasing the time period. Otherwise, the unit notifies the upstream computation unit to less restrictively throttle the thread.

2. 发明申请

US20210096920A1 SHARED RESOURCE ALLOCATION IN A MULTI-THREADED MICROPROCESSOR 有权
公开(公告)号：US20210096920A1
公开(公告)日：2021-04-01
申请号：US16585424
申请日：2019-09-27
申请人： Advanced Micro Devices, Inc.
发明人： Kai Troester , Neil Marketkar , Matthew T. Sobel , Srinivas Keshav
IPC分类号： G06F9/50 , G06F9/48
摘要： An approach is provided for allocating a shared resource to threads in a multi-threaded microprocessor based upon the usefulness of the shared resource to each of the threads. The usefulness of a shared resource to a thread is determined based upon the number of entries in the shared resource that are allocated to the thread and the number of active entries that the thread has in the shared resource. Threads that are allocated a large number of entries in the shared resource and have a small number of active entries in the shared resource, indicative of a low level of parallelism, can operate efficiently with fewer entries in the shared resource, and have their allocation limit in the shared resource reduced.

3. 发明授权

US11144353B2 Soft watermarking in thread shared resources implemented through thread mediation 有权
公开(公告)号：US11144353B2
公开(公告)日：2021-10-12
申请号：US16585586
申请日：2019-09-27
申请人： Advanced Micro Devices, Inc.
发明人： Kai Troester
IPC分类号： G06F9/50
摘要： Techniques for use in a microprocessor core for soft watermarking in thread shared resources implemented through thread mediation. A thread is removed from a thread mediation decision involving multiple threads competing or requesting to use a shared resource at a current clock cycle based on a number of entries in the shared resource that the thread is estimated to have allocated to it at the current clock cycle. By removing the thread from the thread mediation decision, the thread is stalled from allocating additional entries in the shared resource.

4. 发明授权

US11048506B2 Tracking stores and loads by bypassing load store units 有权
公开(公告)号：US11048506B2
公开(公告)日：2021-06-29
申请号：US16450897
申请日：2019-06-24
申请人： Advanced Micro Devices, Inc.
发明人： Krishnan V. Ramani , Kai Troester , Frank C. Galloway , David N. Suggs , Michael D. Achenbach , Betty Ann McDaniel , Marius Evers
IPC分类号： G06F9/28 , G06F9/30 , G06F9/38
摘要： A system and method for tracking stores and loads to reduce load latency when forming the same memory address by bypassing a load store unit within an execution unit is disclosed. Store-load pairs which have a strong history of store-to-load forwarding are identified. Once identified, the load is memory renamed to the register stored by the store. The memory dependency predictor may also be used to detect loads that are dependent on a store but cannot be renamed. In such a configuration, the dependence is signaled to the load store unit and the load store unit uses the information to issue the load after the identified store has its physical address.

5. 发明申请

US20210096874A1 RETIRE QUEUE COMPRESSION 有权
公开(公告)号：US20210096874A1
公开(公告)日：2021-04-01
申请号：US16586642
申请日：2019-09-27
申请人： Advanced Micro Devices, Inc.
发明人： Matthew T. Sobel , Joshua James Lindner , Neil N. Marketkar , Kai Troester , Emil Talpes , Ashok Tirupathy Venkatachar
IPC分类号： G06F9/38
摘要： Systems, apparatuses, and methods for compressing multiple instruction operations together into a single retire queue entry are disclosed. A processor includes at least a scheduler, a retire queue, one or more execution units, and control logic. When the control logic detects a given instruction operation being dispatched by the scheduler to an execution unit, the control logic determines if the given instruction operation meets one or more conditions for being compressed with one or more other instruction operations into a single retire queue entry. If the one or more conditions are met, two or more instruction operations are stored together in a single retire queue entry. By compressing multiple instruction operations together into an individual retire queue entry, the retire queue is able to be used more efficiently, and the processor can speculatively execute more instructions without the retire queue exhausting its supply of available entries.

6. 发明授权

US09292292B2 Stack access tracking 有权
标题翻译：堆栈访问跟踪
公开(公告)号：US09292292B2
公开(公告)日：2016-03-22
申请号：US13922296
申请日：2013-06-20
申请人： Advanced Micro Devices, Inc.
发明人： Kai Troester , Luke Yen
IPC分类号： G06F12/06 , G06F9/38
CPC分类号： G06F9/38 , G06F9/30043 , G06F9/3826 , G06F9/3838
摘要： A processor employs a prediction table at a front end of its instruction pipeline, whereby the prediction table stores address register and offset information for store instructions; and stack offset information for stack access instructions. The stack offset information for a corresponding instruction indicates the entry of the stack accessed by the instruction stack relative to a base entry. The processor uses pattern matching to identify predicted dependencies between load/store instructions and predicted dependencies between stack access instructions. A scheduler unit of the instruction pipeline uses the predicted dependencies to perform store-to-load forwarding or other operations that increase efficiency and reduce power consumption at the processing system.
摘要翻译：处理器在其指令流水线的前端采用预测表，由此预测表存储用于存储指令的地址寄存器和偏移信息; 和堆栈访问指令的堆栈偏移信息。相应指令的堆栈偏移信息指示由指令栈相对于基本条目访问的堆栈的条目。处理器使用模式匹配来识别加载/存储指令之间的预测依赖性以及堆栈访问指令之间的预测依赖性。指令流水线的调度器单元使用预测的依赖性来执行存储到负载转发或提高处理系统的效率并降低功耗的其他操作。

7. 发明申请

US20220027162A1 RETIRE QUEUE COMPRESSION 有权
公开(公告)号：US20220027162A1
公开(公告)日：2022-01-27
申请号：US17497572
申请日：2021-10-08
申请人： Advanced Micro Devices, Inc.
发明人： Matthew T. Sobel , Joshua James Lindner , Neil N. Marketkar , Kai Troester , Emil Talpes , Ashok Tirupathy Venkatachar
IPC分类号： G06F9/38
摘要： Systems, apparatuses, and methods for compressing multiple instruction operations together into a single retire queue entry are disclosed. A processor includes at least a scheduler, a retire queue, one or more execution units, and control logic. When the control logic detects a given instruction operation being dispatched by the scheduler to an execution unit, the control logic determines if the given instruction operation meets one or more conditions for being compressed with one or more other instruction operations into a single retire queue entry. If the one or more conditions are met, two or more instruction operations are stored together in a single retire queue entry. By compressing multiple instruction operations together into an individual retire queue entry, the retire queue is able to be used more efficiently, and the processor can speculatively execute more instructions without the retire queue exhausting its supply of available entries.

8. 发明授权

US11169812B2 Throttling while managing upstream resources 有权
公开(公告)号：US11169812B2
公开(公告)日：2021-11-09
申请号：US16584701
申请日：2019-09-26
申请人： Advanced Micro Devices, Inc.
发明人： Paul James Moyer , Douglas Benson Hunt , Kai Troester
IPC分类号： G06F9/38
摘要： Systems, apparatuses, and methods for arbitrating threads in a computing system are disclosed. A computing system includes a processor with multiple cores, each capable of simultaneously processing instructions of multiple threads. When a thread throttling unit receives an indication that a shared cache has resource contention, the throttling unit sets a threshold number of cache misses for the cache. If the number of cache misses exceeds this threshold, then the throttling unit notifies a particular upstream computation unit to throttle the processing of instructions for the thread. After a time period elapses, if the cache continues to exceed the threshold, then the throttling unit notifies the upstream computation unit to more restrictively throttle the thread by performing one or more of reducing the selection rate and increasing the time period. Otherwise, the unit notifies the upstream computation unit to less restrictively throttle the thread.

9. 发明申请

US20210096914A1 SOFT WATERMARKING IN THREAD SHARED RESOURCES IMPLEMENTED THROUGH THREAD MEDIATION 有权
公开(公告)号：US20210096914A1
公开(公告)日：2021-04-01
申请号：US16585586
申请日：2019-09-27
申请人： Advanced Micro Devices, Inc.
发明人： Kai Troester
IPC分类号： G06F9/50
摘要： Techniques for use in a microprocessor core for soft watermarking in thread shared resources implemented through thread mediation. A thread is removed from a thread mediation decision involving multiple threads competing or requesting to use a shared resource at a current clock cycle based on a number of entries in the shared resource that the thread is estimated to have allocated to it at the current clock cycle. By removing the thread from the thread mediation decision, the thread is stalled from allocating additional entries in the shared resource.

10. 发明申请

US20190310845A1 TRACKING STORES AND LOADS BY BYPASSING LOAD STORE UNITS 审中-公开
公开(公告)号：US20190310845A1
公开(公告)日：2019-10-10
申请号：US16450897
申请日：2019-06-24
申请人： Advanced Micro Devices, Inc.
发明人： Krishnan V. Ramani , Kai Troester , Frank C. Galloway , David N. Suggs , Michael D. Achenbach , Betty Ann McDaniel , Marius Evers
IPC分类号： G06F9/28 , G06F9/30
摘要： A system and method for tracking stores and loads to reduce load latency when forming the same memory address by bypassing a load store unit within an execution unit is disclosed. Store-load pairs which have a strong history of store-to-load forwarding are identified. Once identified, the load is memory renamed to the register stored by the store. The memory dependency predictor may also be used to detect loads that are dependent on a store but cannot be renamed. In such a configuration, the dependence is signaled to the load store unit and the load store unit uses the information to issue the load after the identified store has its physical address.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式