Back to results
US 12,596,579 B2Granted
Method and system for accelerating application performance in solid state drive
AssigneeSamsung Electronics Co., Ltd.
InventorsJing Yang, Shuyi Pei, Jingpei Yang, Rekha Pitchumani
Life of this patent
Estimating the term…
Abstract
A method and a memory device are provided. One or more host central processing units (HCPUs) of the memory device may receive a workload from a host application. The workload includes an identifier (ID). The workload may be distributed to a central processing unit (CPU) of the memory device based on the ID. The workload may be distributed to channels of the memory device based on the CPU.
Claims18
A method of a memory device, comprising:
- receiving a workload from a host application at one or more host central processing units (HCPUs) of the memory device, wherein the workload comprises an identifier (ID);
- distributing the workload to one of first and second central processing units (CPUs) of the memory device based on the ID, wherein the first and second CPUs comprise respective hardware processing units and are configured to load firmware corresponding to the host application based on the ID in the workload;
- and
- distributing the workload to channels of the memory device based on the CPU that received the workload.
1 more element
A memory device, comprising:
- one or more host central processing units (HCPUs) configured to receive a workload from a host application, wherein the workload comprises an identifier (ID), and to distribute the workload to one of first and second central processing units (CPUs) of the memory device based on the ID, wherein the first and second CPUs comprise respective hardware processing units and are configured to load firmware corresponding to the host application based on the ID in the workload;
- and
- the CPU that received the workload configured to distribute the workload to channels of the memory device based on the CPU.
A memory device, comprising:
- one or more host central processing units (HCPUs) configured to receive workloads from host applications, wherein the workloads comprise respective identifiers (IDs), and to distribute the workloads to first and second central processing units (CPUs) of the memory device based on the respective IDs, and wherein the first and second CPUs comprise respective hardware processing units and are configured to load firmware corresponding to the host application based on the ID in the workload;
- the CPUs configured to distribute the workloads to respective channels of the memory device based on the CPUs;
- and
- channels configured to receive the workloads from the CPUs for data storage and retrieval.
1 more element
Dependencies are inferred from the claim text.