Hewlett-packard Development Company, L.p.

Advanced Micro Devices Corporation

Michael Christopher Sedmak

Ken Gary Pomaranski

Emil Talpes

Guy Lowell Kuntz

Stephen Patrick Hack

Howard Calkin

Jeff Barlow

Lacey Joyal

Jeff Brauch

Raymond Gratias

Jay E Fleischman

Michael Estlick

Erik D Swanson

Sneha V Desai

Ray Gratias

Robert E Weidner

Richard McGowen, Ii

A coprocessor such as a floating-point unit includes a pipeline that is partitioned into a first portion and a second portion. A controller is configured to provide control signals to the first portion and the second portion of the pipeline. A first physical distance traversed by control signals propagating from the controller to the first portion of the pipeline is shorter than a second physical distance traversed by control signals propagating from the controller to the second portion of the pipeline. A scheduler is configured to cause a physical register file to provide a first subset of bits of an instruction to the first portion at a first time. The physical register file provides a second subset of the bits of the instruction to the second portion at a second time subsequent to the first time.

Differential pipeline delays in a coprocessor

A pipeline includes a first portion configured to process a first subset of bits of an instruction and a second portion configured to process a second subset of the bits of the instruction. A first clock mesh is configured to provide a first clock signal to the first portion of the pipeline. A second clock mesh is configured to provide a second clock signal to the second portion of the pipeline. The first and second clock meshes selectively provide the first and second clock signals based on characteristics of in-flight instructions that have been dispatched to the pipeline but not yet retired. In some cases, a physical register file is configured to store values of bits representative of instructions. Only the first subset is stored in the physical register file in response to the value of the zero high bit indicating that the second subset is equal to zero.

Clock mesh-based power conservation in a coprocessor based on in-flight instruction characteristics

In one embodiment, a cache element allocation method is provided. Each cache element on a CPU is assigned a quality rank based on the error rate of the cache element. If an allocated cache element is deemed to be faulty, the quality rank of the faulty allocated cache element is compared with the quality rank of the non-allocated cache elements. If a non-allocated cache element has a lower quality rank than the allocated cache element, the non-allocated cache element is swapped in for the allocated cache element.

Systems and methods for CPU repair

In one embodiment, a CPU cache management system is provided. The CPU management system includes, for example, a CPU chip and cache management logic. The CPU chip include cache elements that are initially in use and spare cache elements that not initially in use. The cache management logic determines whether currently-used cache elements are faulty. If a cache element is determined to be faulty, the cache management logic schedules a reboot of the computer and swaps in a spare cache element for the faulty currently-used cache element during the reboot.

Systems and methods for repairing a processor are provided. In one embodiment, a method for repairing a processor is provided that includes, for example, the steps of initializing and executing an operating system, determining that a cache element is faulty, and swapping in a spare cache element for said faulty cache element while the operating system is executing.

Systems and methods for conducting processor health-checks are provided. In one embodiment, a method for evaluating the status of a processor is provided. The method includes, for example, initializing and executing an operating system, de-allocating the processor from the available pool or system resources and performing a health-check on the processor while the operating system is executing.

Methods and systems for conducting processor health-checks

In one embodiment, a method for repairing a faulty cache element is provided. Once a monitored cache element is determined to be faulty, the system stores the repair information, and cache configuration in an EEPROM or non-volatile memory on the CPU module. Then the computer is rebooted. During the reboot, the faulty cache element is repaired by being swapped out for a spare cache element based on the information stored in the EEPROM or the non-volatile memory.

Method and apparatus for implementing a plurality of RDR chains, wherein each of the RDR chains comprises at least one RDR is described. In one embodiment, the method comprises, responsive to execution of a first instruction identifying one of the RDR chains and one of the RDRs of the identified RDR chain, reading data from the identified RDR of the identified RDR chain and, responsive to issuance of a second instruction identifying one of the RDR chains and one of the RDRs of the identified RDR chain, writing data to the identified RDR of the identified RDR chain.

Growing community of inventors

Fort Collins, CO, United States of America

Michael Christopher Sedmak