In-storage compute pipeline controller for enterprise SSDs
This technology describes a storage system that can receive tasks from applications, each task specifying multiple compute functions and the order in which those functions should be run. The system's processing circuit reads the ordering information embedded in the task itself and executes the functions accordingly. This is essentially a smarter task scheduler living inside a storage or compute device, allowing applications to directly orchestrate complex, multi-step processing workflows without relying on a separate external scheduler. It targets environments where data processing needs to happen close to where data lives, such as computational storage or near-data processing architectures.
What you could build
A computational storage drive or controller firmware that allows enterprise applications to push ordered function pipelines directly into storage hardware, reducing latency and data movement; target buyers would be cloud infrastructure vendors, data center operators, and enterprise SSD manufacturers.
Who in Virginia should care
Northern Virginia hyperscale data center operators and federal IT infrastructure contractors would have direct interest in near-data compute technology that reduces data movement overhead.
Readiness: Concept
Concept — described but not yet demonstrated. Lab validated — supported by experimental results in the patent. Prototype likely — the text describes a built, working embodiment.
Readiness is inferred from the patent text, not from a lab visit.
The record
- Inventors
- Madhava Krishnan Ramanathan, Shashwat Jain, Changwoo Min, Vishwanath Maram, Naga Sanjana Bikonda
- Filed
- Patent pending — filed August 25, 2023
- Status
- Application
- Publication number
- US20240069965A1
Ready to talk?
Virginia Tech Intellectual Properties handles licensing for this technology.
Prosim summaries are generated from public patent text and are not legal advice.