IP Library › Granted Patent US 12,033,714
Granted Patent B2
US 12,033,714 · App. 17/224,603 · Granted Jul 9, 2024

Interconnect architecture for three-dimensional processing systems

Inventors: Nuwan S. Jayasena (Santa Clara, CA); Yasuko Eckert (Bellevue, WA)
Assignee: Advanced Micro Devices, Inc.
G11C5/025G06F12/0888G06F12/0811G06F12/0815G06F2212/1016G06F2212/251G06F2212/254G06F2212/502
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 12,033,714
App. No.
17/224,603
Granted
Jul 9, 2024
Kind
B2
Abstract

A processing system includes a plurality of processor cores formed in a first layer of an integrated circuit device and a plurality of partitions of memory formed in one or more second layers of the integrated circuit device. The one or more second layers are deployed in a stacked configuration with the first layer. Each of the partitions is associated with a subset of the processor cores that have overlapping footprints with the partitions. The processing system also includes first memory paths between the processor cores and their corresponding subsets of partitions. The processing system further includes second memory paths between the processor cores and the partitions.

Claims (28)

1. An apparatus comprising:

a first processor core formed in a base logic layer of a three-dimensional processing system; and

a first vertical memory column formed in the three-dimensional processing system above the base logic layer, the first vertical memory column having a footprint that overlaps with an area occupied by the first processor core in the base logic layer;

wherein the first processor core is to select, in response to a first memory access request, one of the first vertical memory column and a second vertical memory column in the three-dimensional processing system, the second vertical memory column having a footprint that overlaps with an area occupied by a second processor core in the base logic layer.

2. The apparatus of claim 1 , wherein the first processor core is to select the one of the first vertical memory column and the second vertical memory column based on a memory address associated with the first memory access request.

3. The apparatus of claim 1 , wherein the first vertical memory column and the first processor core are interconnected by one or more through-silicon vias (TSVs).

4. The apparatus of claim 1 , wherein the first processor core is to select the one of the first vertical memory column and the second vertical memory column based on an instruction that invokes the first memory access request.

5. The apparatus of claim 1 , wherein the first processor core is to select the one of the first vertical memory column and the second vertical memory column based on a type of operations associated with the first memory access request.

6. The apparatus of claim 1 , wherein the first processor core is to select the one of the first vertical memory column and the second vertical memory column based on content of information to be accessed by the first memory access request.

7. The apparatus of claim 1 , further comprising:

a cache associated with the first processor core, wherein at least one of the first vertical memory column and the second vertical memory column bypasses the cache.

8. The apparatus of claim 1 , wherein the second processor core to select, in response to a second memory access request, one of the first vertical memory column and the second vertical memory column.

9. A method comprising:

receiving a memory access request at a first processor core that is one of a plurality of processor cores formed in a base logic layer of a three-dimensional processing system; and

selectively accessing by the first processor core, based on the memory access request, one of a first vertical memory column and a second vertical memory column formed in the three-dimensional processing system above the base logic layer, wherein the first vertical memory column has a footprint that overlaps with an area occupied by the first processor core in the base logic layer and the second vertical memory column has a footprint that overlaps with an area occupied by a second processor core of the plurality of processor cores in the base logic layer.

10. The method of claim 9 , wherein selectively accessing comprises selectively accessing the one of the first vertical memory column and the second vertical memory column based on a memory address associated with the memory access request.

11. The method of claim 9 , wherein the first vertical memory column and the first processor core are interconnected by one or more through-silicon vias (TSVs).

12. The method of claim 9 , wherein selectively accessing comprises selectively accessing based on an instruction that invokes the memory access request.

13. The method of claim 9 , wherein selectively accessing comprises selectively accessing the one of the first vertical memory column and the second vertical memory column based on a type of operations associated with the memory access request.

14. The method of claim 9 , wherein selectively accessing comprises selectively accessing the one of the first vertical memory column and the second vertical memory column based on content of information to be accessed by the memory access request.

15. An apparatus comprising:

a first processor core formed in a base logic layer of a three-dimensional processing system; and

a first vertical memory column comprising random access memory (RAM) and formed in the three-dimensional processing system above the base logic layer, the first vertical memory column having a footprint that overlaps with an area occupied by the first processor core in the base logic layer,

wherein the first processor core is to select, in response to a first memory access request, one of the first vertical memory column and a second vertical memory column in the three-dimensional processing system, the second vertical memory column having a footprint that overlaps with an area occupied by a second processor core in the base logic layer.

16. The apparatus of claim 15 , wherein the first processor core is to select the one of the first vertical memory column and the second vertical memory column based on a memory address associated with the first memory access request.

17. The apparatus of claim 15 , further comprising a first memory access path between the first processor core and the first vertical memory column and a second memory access path between the first processor core and the second vertical memory column.

18. The apparatus of claim 17 , wherein the second memory access path comprises a memory switch between the first processor core and the second vertical memory column.

19. The apparatus of claim 18 , wherein the first memory access path does not include the memory switch.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 30, 2021
From: JAYASENA, NUWAN S; ECKERT, YASUKO
To: ADVANCED MICRO DEVICES INC.
Reel/Frame 057657/0026 →
Continuity (2)
Continuation 14944099 · Nov 17, 2015
Related Publication 20210312952A1 · Oct 7, 2021