# US 10,120,704 B2 · Parallel hardware hypervisor for virtualizing application-specific supercomputers

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 10,120,704 B2
- Document ID: US10120704B2
- Status: Granted
- Kind: B2
- Application number: 15/137,268
- Filed: 2016-04-25
- Published: 2016-08-25
- Granted: 2018-11-06
- Earliest claimed: 2012-02-04
- Assignees: Global Supercomputing Corporation
- Applicants: Kemal Ebcioglu; Atakan Dogan; Reha Oguz Altug
- Inventors: Kemal Ebcioglu; Atakan Dogan; Reha Oguz Altug
- Web page: https://patentgrep.com/patent/US10120704B2

## CPC

- G06F 9/45533: {Hypervisors; Virtual machine monitors}
- G06F 9/45558: {Hypervisor-specific management and integration aspects}
- G06F 9/5077: {Logical partitioning of resources; Management or configuration of virtualized resources (specific details on emulation or internal functioning of virtual machines G06F9/455)}
- G06F 15/7871: {Reconfiguration support, e.g. configuration loading, configuration switching, or hardware OS}
- G06F 2009/4557: {Distribution of virtual machine instances; Migration and load balancing}
- Y02D 10/00: Energy efficient computing, e.g. low power processors, power management or thermal management
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- Y02D: CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE

## Abstract

A parallel hypervisor system for virtualizing application-specific supercomputers is disclosed. The hypervisor system comprises (a) at least one software-virtual hardware pair consisting of a software application, and an application-specific virtual supercomputer for accelerating the said software application, wherein (i) The virtual supercomputer contains one or more virtual tiles; and (ii) The software application and the virtual tiles communicate among themselves with messages; (b) One or more reconfigurable physical tiles, wherein each virtual tile of each supercomputer can be implemented on at least one physical tile, by configuring the physical tile to perform the virtual tile's function; and (c) A scheduler implemented substantially in hardware, for parallel pre-emptive scheduling of the virtual tiles on the physical tiles.

## Claims

### Claim 1

A flat hypervisor system for virtualizing application-specific supercomputers, the system comprising:
(a) at least one software-virtual hardware pair consisting of a software application, and an application-specific virtual supercomputer for accelerating the software application, wherein:
i. the application-specific virtual supercomputer comprises a plurality of virtual tiles; and
ii. the software application and the virtual tiles communicate among themselves with communication messages;
(b) a plurality of reconfigurable physical tiles, wherein each virtual tile of the plurality of virtual tiles of the application-specific virtual supercomputer can be implemented on at least a reconfigurable physical tile of the plurality of reconfigurable physical tiles, by configuring the reconfigurable physical tile to perform a function of a virtual tile; and
(c) a scheduler consisting of a plurality of hardware components, wherein the scheduler communicates and synchronizes to perform parallel pre-emptive scheduling of the plurality of virtual tiles on the plurality of reconfigurable physical tiles;
(d) wherein all hardware components of the flat hypervisor system, the plurality of reconfigurable physical tiles, and the plurality of hardware components of the scheduler, are partitioned into two or more modules interconnected by a scalable network, wherein the two or more modules each comprise an I/O controller of a plurality of I/O controllers; and
wherein the plurality of I/O controllers collectively enable a first hardware component inside a first module of the two or more modules, to efficiently and scalably communicate messages with a second hardware component inside a second module of the two or more modules, and
wherein the communication of the messages is performed without altering message communication protocols among all of the hardware components of the flat hypervisor system, and
wherein each module of the two or more modules is realized by its respective copy of a cloud building block, where the cloud building block is a union module capable of realizing any one of the two or more modules, depending on configuration parameters supplied to the cloud building block.

### Claim 2 (depends on 1)

The hypervisor system of claim 1, further comprising:
wherein the cloud building block is capable of realizing any one of the modules within the hypervisor system, depending on configuration parameters supplied to the cloud building block; and wherein each module within the hypervisor system is realized by its respective copy of the cloud building block; and wherein an implementation of the cloud building block comprises ASIC technology.