IP Library › Granted Patent US 12,450,008
Granted Patent B2
US 12,450,008 · App. 17/558,268 · Granted Oct 21, 2025

Remote storage for hardware microservices hosted on XPUs and SOC-XPU platforms

Inventors: Andrzej Kuriata (Gdansk, PL); Susanne M. Balle (Hudson, NH); Duane E. Galbi (Wayland, MA); Sundar Nadathur (Cupertino, CA); Nagabhushan Chitlur (Portland, OR); Francesc Guim Bernat (Barcelona, ES); Alexander Bachmutsky (Sunnyvale, CA)
Assignee: Intel Corporation
G06F3/0659G06F3/0604G06F3/0613G06F3/067G06F15/7807
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,450,008
App. No.
17/558,268
Granted
Oct 21, 2025
Kind
B2
Abstract

Methods, apparatus, and software for remote storage of hardware microservices hosted on other processing units (XPUs) and SOC-XPU Platforms. The apparatus may be a platform including a System on Chip (SOC) and an XPU, such as a Field Programmable Gate Array (FPGA). Software, via execution on the SOC, enables the platform to pre-provision storage space on a remote storage node and assign the storage space to the platform, wherein the pre-provisioned storage space includes one or more container images to be implemented as one or more hardware (HW) microservice front-ends. The XPU/FPGA is configured to implement one or more accelerator functions used to accelerate HW microservice backend operations that are offloaded from the one or more HW microservice front-ends. The platform is also configured to pre-provision a remote storage volume containing worker node components and access and persistently store worker node components.

Claims (54)

1. A platform configured to be deployed in a network environment including a remote storage node coupled to a fabric, comprising:

a circuit board to which multiple components are operatively coupled, including,

a System On Chip (SOC), having a plurality of cores;

at least one fabric interface configured to be coupled to the fabric;

memory, in which software to be executed on one or more of the plurality of cores in the SOC is stored, the software including software code configured to pre-provision storage space on the remote storage node and assign the storage space to the platform, wherein the pre-provisioned storage space includes one or more container images to be implemented as one or more hardware (HW) microservice front-ends; and

an Other Processing Unit (XPU), communicatively coupled to the SOC via interconnect circuitry in the circuit board and configured to implement one or more accelerator functions used to accelerate HW microservice backend operations that are offloaded from the one or more HW microservice front-ends.

2. The platform of claim 1 , wherein the remote storage node includes a plurality of NVMe (Non-Volatile Memory Express) drives, and the platform is configured to employ an NVMe over fabric (NVMe-oF) protocol to pre-provision the storage space on the remote storage node.

3. The platform of claim 2 , wherein the storage space is pre-provisioned as one or more NVMe-oF volumes including a pre-provisioned NVMe-oF volume containing a plurality of HW microservice container images.

4. The platform of claim 2 , wherein the storage space is pre-provisioned as one or more NVMe-oF volumes including a pre-provisioned NVMe-oF volume containing worker node components.

5. The platform of claim 4 , wherein a portion of the memory comprises non-volatile memory implemented as persistent storage, and wherein execution of the software code further enables the platform to:

retrieve worker node components from the pre-provisioned NVMe-oF volume containing worker node components; and

persistently store the worker node components that are retrieved in the non-volatile memory.

6. The platform of claim 1 , wherein the XPU comprises a Field Programmable Gate Array (FPGA), and the one or more accelerated functions comprise FPGA kernels.

7. The platform of claim 6 , wherein execution of the software further enables to platform to:

retrieve an FPGA kernel bitstream from the remote storage node;

provision an accelerator function in the FPGA using the FPGA kernel bitstream to program the FPGA to accelerate HW microservice backend operations for at least one HW microservice.

8. The platform of claim 1 , wherein the platform comprises an infrastructure processing unit (IPU).

9. The platform of claim 8 , wherein the XPU comprises a Field Programmable Gate Array (FPGA), wherein the IPU comprises a Peripheral Component Interconnect Express (PCIe) card, and the FPGA is configured to implement one or more PCIe interfaces.

10. The platform of claim 1 , wherein the platform is configured to be installed in a server including a central processing unit (CPU) that is communicatively coupled to at least one of the SOC and XPU when the platform is installed in the server.

11. A method implemented on a platform configured to be deployed in a data center including a remote storage node coupled to a fabric, the platform including a System on Chip (SOC) having a plurality of cores and operatively coupled to a circuit board and communicatively coupled to an other processing unit (XPU) operatively coupled to the circuit board, the method comprising:

communicatively coupling the platform to the fabric; and

pre-provisioning storage space on the remote storage node and assigning the storage space to the platform, wherein the pre-provisioned storage space includes one or more container images to be implemented as one or more hardware (HW) microservice front-ends on the platform.

12. The method of claim 11 , wherein the remote storage node includes a plurality of NVMe (Non-Volatile Memory Express) drives, further comprising:

employing an NVMe over fabric (NVMe-oF) protocol to pre-provision the storage space on the remote storage node,

wherein the storage space is pre-provisioned as one or more NVMe-oF volumes including a pre-provisioned NVMe-oF volume containing a plurality of HW microservice front-end container images.

13. The method of claim 12 , further comprising:

pre-provisioning an NVMe-oF volume containing worker node components;

retrieving worker node components from the pre-provisioned NVMe-oF volume containing worker node components; and

persistently storing the worker node components that are retrieved in non-volatile storage on the platform.

14. The method of claim 12 , further comprising:

searching for a HW microservice front-end container image among the plurality of HW microservice front-end container images stored on the remote storage node;

receiving, via the remote storage nodes, a HW microservice front-end container image the is provided by the remote storage node;

testing the HW microservice front-end container image that is provided; and

determining whether the provided HW microservice front-end container image successfully works or fails.

15. The method of claim 11 , wherein the XPU comprises a Field Programmable Gate Array (FPGA), and the one or more accelerated functions comprise FPGA kernels, further comprising:

retrieving an FPGA kernel bitstream from the remote storage node;

provision an accelerator function in the FPGA using the FPGA kernel bitstream to program the FPGA to accelerate HW microservice backend operations for at least one HW microservice.

16. A system comprising:

a fabric or network;

a remote storage node, coupled to the fabric and including a plurality of storage devices;

a server, including,

one or more central processing units (CPUs); and

an infrastructure processing unit (IPU) comprising,

a circuit board to which multiple components are operatively coupled, including,

a System On Chip (SOC), having a plurality of cores;

at least one fabric interface coupled to the fabric or network;

memory, in which software to be executed on one or more of the plurality of cores in the SOC is stored, the software including software code configured to pre-provision storage space on the remote storage node and assign the storage space to the platform, wherein the pre-provisioned storage space includes one or more container images to be implemented as one or more hardware (HW) microservice front-ends; and

a Field Programmable Gate Array (FPGA), communicatively coupled to the SOC via interconnect circuitry in the circuit board and configured to implement one or more accelerator functions used to accelerate HW microservice backend operations that are offloaded from the one or more HW microservice front-ends.

17. The system of claim 16 , wherein the remote storage node includes a plurality of NVMe (Non-Volatile Memory Express) drives, and the platform is configured to employ an NVMe over fabric (NVMe-oF) protocol to pre-provision the storage space on the remote storage node.

18. The system of claim 17 , wherein the storage space is pre-provisioned as one or more NVMe-oF volumes including a pre-provisioned NVMe-oF volume containing a plurality of HW microservice container images and including a pre-provisioned NVMe-oF volume containing worker node components.

19. The system of claim 18 , wherein a portion of the memory in the IPU comprises non-volatile memory implemented as persistent storage, and wherein execution of the software code further enables the IPU to:

retrieve worker node components from the pre-provisioned NVMe-oF volume containing worker node components; and

persistently store the worker node components that are retrieved in the non-volatile memory.

20. The system of claim 16 , wherein the remote storage node is installed in a first rack in a datacenter, the server is installed in a second rack in the data center, and the first rack is connected to the second rack via the fabric or network.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 18, 2022
From: KURIATA, ANDRZEJ; BALLE, SUSANNE M.; GALBI, DUANE E.; NADATHUR, SUNDAR; CHITLUR, NAGABHUSHAN; GUIM BERNAT, FRANCESC; BACHMUTSKY, ALEXANDER
To: INTEL CORPORATION
Reel/Frame 058680/0416 →
Continuity (1)
Related Publication 20220113911A1 · Apr 14, 2022
References Cited (9)
US 20180337991A1 · Kumar · 2018 [cited by examiner]
US 20190073255A1 · Sato · 2019 [cited by examiner]
US 20190095229A1 · Guo · 2019 [cited by examiner]
US 20200142735A1 · Maciocco · 2020 [cited by examiner]
US 20200409591A1 · Kulli · 2020 [cited by examiner]
US 20210286745A1 · Smith · 2021 [cited by examiner]
Extended European Search Report for Patent Application No. 22203130.4, Mailed May 9, 2023, 9 pages. [cited by applicant]
Williams Chris: “Xilinx pops a 16-core 64-bit Arm system-on-chip from NXP into its latest FPGA-based IOOGbps smart NIC—The Register”, Mar. 23, 2021 (Mar. 23, 2021), pp. 1-7, XP093042995, Retrieved from the Internet: URL… [cited by applicant]
Xilinx Sn1000: “Product Brief—Alveo(TM) SN1000 SmartNICs”, Nov. 24, 2021 (Nov. 24, 2021), pp. 1-2, XP093043010, Retrieved from the Internet:URL:https://web.archive.org/web/2021112421 4547if_/https://www.xilinx.com/conte… [cited by applicant]