IP Library › Granted Patent US 12,422,990
Granted Patent B2
US 12,422,990 · App. 18/146,538 · Granted Sep 23, 2025

Extended training for memory

Inventors: Alicia Wen Ju Yurie Leong (Austin, TX); Jayesh Hari Joshi (Austin, TX); William Robert Alverson (Del Valle, TX); Joshua Taylor Knight (Georgetown, TX); Jerry Anton Ahrens (Sister Bay, WI); Amitabh Mehra (Fort Collins, CO); Grant Evan Ley (Eden, UT)
Assignee: Advanced Micro Devices, Inc.
G06F3/0613G06F3/0634G06F3/0673
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 12,422,990
App. No.
18/146,538
Granted
Sep 23, 2025
Kind
B2
Abstract

Extended training for memory is described. In accordance with the described techniques, a training request to train a memory with extended training is received. The extended training corresponds to a longer amount of time than a default training. The extended training of the memory is performed using a set of target memory settings. In one or more implementations, the extended training is performed during a boot up phase of the computing device.

Claims (39)

1. A computing device comprising:

a memory; and

a processor to execute an application configured to:

display a user interface having:

an option that is selectable to train the memory with an extended training instead of a default training; and

a control that is selectable to define an additional amount of time for performing the extended training relative to the default training;

responsive to selection of the option via the user interface, request a memory controller to perform the extended training of the memory over a duration defined by the additional amount of time specified via the user interface control using a set of target memory settings; and

display, after the extended training, training results via the user interface that indicate a response of the memory to the extended training.

2. The computing device of claim 1 , wherein the memory controller is configured to perform the extended training during a boot up phase of the computing device.

3. The computing device of claim 1 , wherein the memory controller is configured to cause the memory to operate using the set of target memory settings.

4. The computing device of claim 1 , wherein the memory controller is configured to communicate one or more change signals to cause the memory to operate using the set of target memory settings.

5. The computing device of claim 1 , wherein the set of target memory settings includes at least one overclocked memory setting for operating the memory in an overclocked mode.

6. A system comprising:

a memory; and

a processor to execute an application to:

display a user interface having:

an option that is selectable to toggle between one of at least two training modes for training the memory, the at least two training modes including a default training mode and an extended training mode; and

a control that is selectable to define an additional amount of time for performing the extended training mode relative to an amount of time for performing the default training mode;

based on a selection of the option and input to the control via the user interface, request a memory controller to train the memory using the extended training mode, during a boot up process, over a duration defined by the additional amount of time specified via the input to the control; and

output, after the boot up process, training results based on the training via the user interface, the training results including at least one overclocked memory setting with which the memory is operable to operate in an overclocked mode.

7. The system of claim 6 , wherein the memory controller performs a first set of operations while training the memory using the default training mode, and wherein the memory controller performs a second set of operations while training the memory using the extended training mode, wherein the second set of operations includes more operations than the first set of operations.

8. The system of claim 6 , wherein the memory controller is further configured to select the extended training mode based on a training request received from a requesting application.

9. The system of claim 8 , wherein the training request is received from the requesting application responsive to the application being launched, installed, or updated.

10. The system of claim 6 , wherein the memory controller trains the memory using the extended training mode during the boot up process by directly accessing the memory to perform at least one training operation that is destructive to content of the memory.

11. The system of claim 6 , wherein the memory controller is further configured to cause the memory to be operated using the at least one overclocked memory setting automatically responsive to the system advancing from a boot up phase to a working state.

12. A method comprising:

displaying a user interface that includes:

an option that is selectable to train a memory with an extended training instead of a default training; and

a control that is selectable to define an additional amount of time for performing the extended training relative to the default training;

responsive to receiving a selection of the option via the user interface, requesting a memory controller to perform the extended training of the memory over a duration defined by the additional amount of time specified via the user interface control using a set of target memory settings; and

after the extended training, displaying training results via the user interface that indicate a response of the memory to the extended training.

13. The method of claim 12 , wherein the user interface control includes descriptions of multiple different levels of extended training, wherein each of the multiple different levels of extended training are associated with different amounts of additional time.

14. The method of claim 12 , wherein the user interface control enables a user to enter an amount of time available for the extended training, the method further comprising defining the additional amount of time for the extended training based on the amount of time entered via the user interface control.

15. The method of claim 12 , wherein the training results indicate a signal integrity improvement from the extended training of the memory for operating the memory with the set of target memory settings.

16. The computing device of claim 1 , wherein the memory controller is configured to perform the default training of the memory unless user input is received via the user interface selecting the option.

17. The computing device of claim 1 , wherein the memory controller is configured to perform at least one training operation for the set of target memory settings for a longer time during the extended training than during the default training to improve a signal integrity for reading data to and writing data from the memory when operating using the set of target memory settings.

18. The computing device of claim 1 , wherein the user interface control includes a plurality of extended training options that each correspond to different selectable amounts of time, wherein the additional amount of time corresponds to a selected one of the plurality of extended training options.

19. The computing device of claim 1 , wherein the memory controller is configured to perform at least one training operation for a longer time during the extended training than during the default training to use the longer time for additional prediction of one or more stability margins when operating the memory with the set of target memory settings.

20. The system of claim 6 , wherein the user interface control includes descriptions of multiple different levels of extended training, wherein each of the multiple different levels of extended training are associated with different amounts of additional time presented in the user interface.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 12, 2023
From: LEONG, ALICIA WEN JU YURIE; JOSHI, JAYESH HARI; ALVERSON, WILLIAM ROBERT; KNIGHT, JOSHUA TAYLOR; AHRENS, JERRY ANTON; MEHRA, AMITABH; LEY, GRANT EVAN
To: ADVANCED MICRO DEVICES, INC
Reel/Frame 062362/0514 →
Continuity (1)
Related Publication 20240211142A1 · Jun 27, 2024
References Cited (31)
US 9633570B2 · Peterson · 2017 [cited by examiner]
US 10559335B2 · Kim · 2020 [cited by examiner]
US 11320885B2 · Wang · 2022 [cited by examiner]
US 20040221239A1 · Hachigian · 2004 [cited by examiner]
US 20060271773A1 · Marquiz · 2006 [cited by examiner]
US 20080104386A1 · Van Rooyen · 2008 [cited by examiner]
US 20100082967A1 · Lo · 2010 [cited by examiner]
US 20110016214A1 · Jackson · 2011 [cited by applicant]
US 20130198502A1 · Rothman · 2013 [cited by examiner]
US 20130275739A1 · Kim · 2013 [cited by examiner]
US 20140047227A1 · Breternitz · 2014 [cited by examiner]
US 20170087410A1 · Cone · 2017 [cited by examiner]
US 20180095806A1 · Aneja · 2018 [cited by examiner]
US 20190004825A1 · Rangarajan · 2019 [cited by examiner]
US 20190122745A1 · Kim · 2019 [cited by examiner]
US 20200043362A1 · Vlaskamp · 2020 [cited by examiner]
US 20200097201A1 · Irshad et al. · 2020 [cited by applicant]
US 20210064390A1 · Park · 2021 [cited by examiner]
US 20210081127A1 · Xiao et al. · 2021 [cited by applicant]
US 20210373642A1 · Wang · 2021 [cited by examiner]
US 20220039197A1 · Dao · 2022 [cited by examiner]
US 20220113881A1 · Roberts · 2022 [cited by applicant]
US 20220197832A1 · Ruttenberg et al. · 2022 [cited by applicant]
US 20230061291A1 · Nealis · 2023 [cited by examiner]
US 20230214141A1 · Chien · 2023 [cited by examiner]
US 20230215475A1 · Chien · 2023 [cited by examiner]
US 20230289192A1 · Peterson · 2023 [cited by examiner]
US 20230350715A1 · Knight et al. · 2023 [cited by applicant]
Knight, Joshua Taylor, et al., “US Application as Filed”, U.S. Appl. No. 17/732,987, filed Apr. 29, 2022, 41 pages. [cited by applicant]
“Non-Final Office Action”, U.S. Appl. No. 17/732,987, filed Nov. 22, 2024, 27 pages. [cited by applicant]
“Final Office Action”, U.S. Appl. No. 17/732,987, filed May 14, 2025, 17 pages. [cited by applicant]