Technologies for dividing memory across socket partitions
AssigneeIntel Corporation
InventorsMurugasamy K. Nachimuthu, Mohan J. Kumar
Life of this patent
Estimating the term…
Abstract
Technologies for dividing resources across partitions include a compute sled. The compute sled is to determine partitions among sockets of the compute sled. Each socket is associated with a corresponding processor. The compute sled is also to establish a separate memory space for each determined partition, obtain, from an application executed in one of the sockets, a request to access a logical memory address, identify the partition associated with the memory access request, determine a corresponding physical memory address as a function of the identified partition and the logical memory address, and access a memory of the compute sled at the determined physical memory address. Other embodiments are also described and claimed.
Claims16
A compute sled comprising:
- a compute engine having (i) multiple processors and sockets, wherein each processor is associated with a respective socket of the compute sled and (ii) a memory, wherein the compute engine is to:
- determine partitions among the sockets;
- establish a separate memory space for each determined partition;
- obtain, from an application executed in one of the sockets, a request to access a logical memory address;
- identify the partition associated with the memory access request;
- add a partition identifier to the to the memory access request;
- determine a corresponding physical memory address as a function of the identified partition and the logical memory address, including multiplying the partition identifier by a predefined memory space size to determine a base physical memory address, and adding the logical address to the base physical memory address;
- and
- access the memory at the determined physical memory address.
One or more machine-readable storage media comprising a plurality of executable computer program instructions stored thereon that, in response to being executed, cause a compute sled to:
- determine partitions among sockets of the compute sled, wherein each socket is associated with a corresponding processor of the compute sled;
- establish a separate memory space for each determined partition;
- obtain from an application executed in one of the sockets, a request to access a logical memory address;
- identify the partition associated with the memory access request;
- add a partition identifier to the to the memory access request;
- determine a corresponding physical memory address as a function of the identified partition and the logical memory address, including multiplying the partition identifier by a predefined memory space size to determine a base physical memory address, and adding the logical address to the base physical memory address;
- and
- access a memory of the compute sled at the determined physical memory address.
A method comprising:
- determining, by a compute sled, partitions among sockets of the compute sled, wherein each socket is associated with a corresponding processor of the compute sled;
- establishing, by the compute sled, a separate memory space for each determined partition;
- obtaining, by the compute sled, from an application executed in one of the sockets, a request to access a logical memory address;
- identifying, by the compute sled, the partition associated with the memory access request;
- adding a partition identifier to the to the memory access request;
- determining, by the compute sled, a corresponding physical memory address as a function of the identified partition and the logical memory address, including multiplying the partition identifier by a predefined memory space size to determine a base physical memory address, and adding the logical address to the base physical memory address;
- and
- accessing, by the compute sled, a memory of the compute sled at the determined physical memory address.
Dependencies are inferred from the claim text.