US 11,144,207 B2 · Accelerating memory compression of a physically scattered buffer

Abstract

Embodiments herein describe using compression engines in a processor subsystem to compress only the data fragments stored locally. That is, an application may be allocated a buffer where the physical memory of that buffer is spread across multiple processor subsystems. Rather than asking a single actor (e.g., a single host processor or compression engine) to compress all the fragments of the buffer, a compression library can instead instruct the individual compression engines in each of the processor subsystems to compress only the fragments stored in local memory in the same processor subsystem. Doing so leverages the memory affinity between the compression engines in the local memory which can reduce the overall time required to perform compression.

Claims20

Dependent claims

A method comprising:

receiving a request to compress a buffer comprising a plurality of fragments;
identifying a plurality of processor subsystems in a computing system that each stores at least one of the plurality of fragments in a local memory, wherein the plurality of fragments are assigned contiguous virtual memory addresses associated with the buffer, wherein each of the plurality of processor subsystems comprises a compression engine and share a virtual address space with each other;
instructing each compression engine in the plurality of processor subsystems to compress only fragments of the plurality of fragments stored in the same processor subsystem as the respective compression engine;
and
storing the compressed fragments in an output buffer.

2 more elements

A computer readable storage medium having computer readable program code embodied therewith, the computer readable program code executable by one or more computer processors to perform an operation, the operation comprising:

receiving a request to compress a buffer comprising a plurality of fragments;
identifying a plurality of processor subsystems in a computing system that each stores at least one of the plurality of fragments in a local memory, wherein the plurality of fragments are assigned contiguous virtual memory addresses associated with the buffer, wherein each of the plurality of processor subsystems comprises a compression engine and share a virtual address space with each other;
instructing each compression engine in the plurality of processor subsystems to compress only fragments of the plurality of fragments stored in the same processor subsystem as the compression engine;
and
storing the compressed fragments in an output buffer.

2 more elements

A system, comprising:

a plurality of processor subsystems, each of the plurality of processor subsystems comprising a local memory and a compression engine, and the plurality of processor subsystems sharing a virtual address space;
and
a compression library configured to:
receive a request to compress a buffer comprising a plurality of fragments,
identify a subset of the plurality of processor subsystems each stores at least one of the plurality of fragments, wherein the plurality of fragments are assigned contiguous virtual memory addresses associated with the buffer, and
instruct each compression engine in the plurality of processor subsystems to compress only fragments of the plurality of fragments stored in the same processor subsystem as the compression engine and store the compressed fragments in an output buffer.

3 more elements

Dependencies are inferred from the claim text.

Accelerating memory compression of a physically scattered buffer

Life of this patent

Accelerating memory compression of a physically scattered buffer

Life of this patent

Abstract

Claims20