IP Library Granted Patent US 11,709,667
Granted Patent B2
US 11,709,667 · App. 17/347,141 · Granted Jul 25, 2023

Non-disruptive firmware upgrade of symmetric hardware accelerator systems

Inventors: Tao Chen (Beijing, CN); Yong Zou (San Jose, CA); Ran Liu (Beijing, CN)
Assignee: EMC IP Holding Company LLC
G06F8/65G06F9/5083G06F2209/509
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 11,709,667
App. No.
17/347,141
Granted
Jul 25, 2023
Kind
B2
Abstract

In a symmetric hardware accelerator system, an initial hardware accelerator is selected for an upgrade of firmware. The initial and other hardware accelerators handle workloads that have been balanced across the hardware accelerators. Workloads are rebalanced by directing workloads having low CPU utilization to the initial hardware accelerator. A CPU fallback is conducted of the workloads of the initial hardware accelerator to the CPU. While the CPU is handling the workloads, firmware of the initial hardware accelerator is upgraded.

Claims (65)

1. A method for upgrading firmware of a symmetric hardware accelerator system coupled to a central processing unit (CPU) comprising:

selecting an initial hardware accelerator to upgrade, the initial and other hardware accelerators handling workloads that have been balanced across the hardware accelerators;

rebalancing the workloads by directing workloads having low CPU utilization to the initial hardware accelerator;

conducting a CPU fallback of the workloads of the initial hardware accelerator to the CPU; and

while the CPU is handling the workloads, upgrading firmware of the initial hardware accelerator, the initial hardware accelerator now being an upgraded hardware accelerator, wherein the rebalancing the workloads comprises:

receiving a request to process a first workload of a first type;

directing the first workload to the initial hardware accelerator;

receiving a request to process a second workload of a second type; and

directing the second workload to a hardware accelerator other than the initial hardware accelerator, wherein an amount of CPU utilization of the first type of workload is less than an amount of CPU utilization of the second type of workload.

2. The method of claim 1 further comprising:

redirecting workloads of a next hardware accelerator to upgrade to the upgraded hardware accelerator; and

upgrading firmware of the next hardware accelerator.

3. The method of claim 2 further comprising:

repeating the redirecting workloads and upgrading firmware until all hardware accelerators have been upgraded; and

upon all hardware accelerators having been upgraded, offloading the workloads of the CPU back onto one or more hardware accelerators that have been upgraded.

4. The method of claim 1 wherein the hardware accelerators comprise QuickAssist Technology (QAT) compatible accelerators.

5. The method of claim 1 wherein at an initial state of the symmetric hardware accelerator system before the upgrading of the firmware, each hardware accelerator of the symmetric hardware accelerator system is configured to handle workloads of different types.

6. The method of claim 1 further comprising:

during the upgrading the firmware, receiving a request to process a third workload of a third type;

directing the third workload to the CPU;

after the upgrading the firmware, receiving a request to process a second fourth workload of the third type; and

directing the fourth workload to a hardware accelerator.

7. A system comprising: a processor; and memory configured to store one or more sequences of instructions which, when executed by the processor, cause the processor to carry out the steps of:

selecting an initial hardware accelerator to upgrade, the initial and other hardware accelerators handling workloads that have been balanced across the hardware accelerators;

rebalancing the workloads by directing workloads having low CPU utilization to the initial hardware accelerator;

conducting a CPU fallback of the workloads of the initial hardware accelerator to a central processing unit (CPU);

while the CPU is handling the workloads, upgrading firmware of the initial hardware accelerator, the initial hardware accelerator now being an upgraded hardware accelerator;

during the upgrading the firmware, receiving a request to process a first workload of a first type;

directing the first workload to the CPU;

after the upgrading the firmware, receiving a request to process a second workload of the first type; and

directing the second workload to a hardware accelerator.

8. The system of claim 7 wherein the processor further carries out the steps of:

redirecting workloads of a next hardware accelerator to upgrade to the upgraded hardware accelerator; and

upgrading firmware of the next hardware accelerator.

9. The system of claim 8 wherein the processor further carries out the steps of:

repeating the redirecting workloads and upgrading firmware until all hardware accelerators have been upgraded; and

upon all hardware accelerators having been upgraded, offloading the workloads of the CPU back onto one or more hardware accelerators that have been upgraded.

10. The system of claim 7 wherein the hardware accelerators comprise QuickAssist Technology (QAT) compatible accelerators.

11. The system of claim 7 wherein at an initial state of the symmetric hardware accelerator system before the upgrading of the firmware, each hardware accelerator of the symmetric hardware accelerator system is configured to handle workloads of different types.

12. The system of claim 7 wherein the rebalancing the workloads comprises:

receiving a request to process a third workload of a second type;

directing the third workload to the initial hardware accelerator;

receiving a request to process a fourth workload of a third type; and

directing the fourth workload to a hardware accelerator other than the initial hardware accelerator, wherein an amount of CPU utilization of the second type of workload is less than an amount of CPU utilization of the third type of workload.

13. A computer program product, comprising a non-transitory computer-readable medium having a computer-readable program code embodied therein, the computer-readable program code adapted to be executed by one or more processors to implement a method for upgrading firmware of a symmetric hardware accelerator system coupled to a central processing unit (CPU), the method comprising:

selecting an initial hardware accelerator to upgrade, the initial and other hardware accelerators handling workloads that have been balanced across the hardware accelerators;

rebalancing the workloads by directing workloads having low CPU utilization to the initial hardware accelerator;

conducting a CPU fallback of the workloads of the initial hardware accelerator to the CPU;

while the CPU is handling the workloads, upgrading firmware of the initial hardware accelerator, the initial hardware accelerator now being an upgraded hardware accelerator;

during the upgrading the firmware, receiving a request to process a first workload of a first type;

directing the first workload to the CPU;

after the upgrading the firmware, receiving a request to process a second workload of the first type; and

directing the second workload to a hardware accelerator.

14. The computer program produce of claim 13 wherein the method further comprises:

redirecting workloads of a next hardware accelerator to upgrade to the upgraded hardware accelerator; and

upgrading firmware of the next hardware accelerator.

15. The computer program product of claim 14 wherein the method further comprises:

repeating the redirecting workloads and upgrading firmware until all hardware accelerators have been upgraded; and

upon all hardware accelerators having been upgraded, offloading the workloads of the CPU back onto one or more hardware accelerators that have been upgraded.

16. The computer program product of claim 13 wherein at an initial state of the symmetric hardware accelerator system before the upgrading of the firmware, each hardware accelerator of the symmetric hardware accelerator system is configured to handle workloads of different types.

17. The computer program product of claim 13 wherein the rebalancing the workloads comprises:

receiving a request to process a third workload of a second type;

directing the third workload to the initial hardware accelerator;

receiving a request to process a fourth workload of a third type; and

directing the fourth workload to a hardware accelerator other than the initial hardware accelerator, wherein an amount of CPU utilization of the second type of workload is less than an amount of CPU utilization of the third type of workload.

Assignments (8)
RELEASE OF SECURITY INTEREST IN PATENTS PREVIOUSLY RECORDED AT REEL/FRAME (058014/0560) Recorded Jun 10, 2022
From: THE BANK OF NEW YORK MELLON TRUST COMPANY, N.A., AS NOTES COLLATERAL AGENT
To: DELL PRODUCTS L.P.; EMC IP HOLDING COMPANY LLC
Reel/Frame 062022/0473 →
RELEASE OF SECURITY INTEREST IN PATENTS PREVIOUSLY RECORDED AT REEL/FRAME (057931/0392) Recorded Jun 10, 2022
From: THE BANK OF NEW YORK MELLON TRUST COMPANY, N.A., AS NOTES COLLATERAL AGENT
To: DELL PRODUCTS L.P.; EMC IP HOLDING COMPANY LLC
Reel/Frame 062022/0382 →
RELEASE OF SECURITY INTEREST IN PATENTS PREVIOUSLY RECORDED AT REEL/FRAME (057758/0286) Recorded Jun 10, 2022
From: THE BANK OF NEW YORK MELLON TRUST COMPANY, N.A., AS NOTES COLLATERAL AGENT
To: DELL PRODUCTS L.P.; EMC IP HOLDING COMPANY LLC
Reel/Frame 061654/0064 →
SECURITY INTEREST Recorded Oct 6, 2021
From: DELL PRODUCTS L.P.; EMC IP HOLDING COMPANY LLC
To: THE BANK OF NEW YORK MELLON TRUST COMPANY, N.A., AS NOTES COLLATERAL AGENT
Reel/Frame 058014/0560 →
SECURITY INTEREST Recorded Oct 6, 2021
From: DELL PRODUCTS L.P.; EMC IP HOLDING COMPANY LLC
To: THE BANK OF NEW YORK MELLON TRUST COMPANY, N.A., AS NOTES COLLATERAL AGENT
Reel/Frame 057758/0286 →
SECURITY INTEREST Recorded Oct 6, 2021
From: DELL PRODUCTS L.P.; EMC IP HOLDING COMPANY LLC
To: THE BANK OF NEW YORK MELLON TRUST COMPANY, N.A., AS NOTES COLLATERAL AGENT
Reel/Frame 057931/0392 →
SECURITY AGREEMENT Recorded Oct 1, 2021
From: DELL PRODUCTS, L.P.; EMC IP HOLDING COMPANY LLC
To: CREDIT SUISSE AG, CAYMAN ISLANDS BRANCH
Reel/Frame 057682/0830 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jun 14, 2021
From: CHEN, TAO; ZOU, YONG; LIU, RAN
To: EMC IP HOLDING COMPANY LLC
Reel/Frame 056536/0442 →
Continuity (1)
Related Publication 20220398084A1 · Dec 15, 2022