US 12,277,444 B2
Software-defined tensor streaming multiprocessor for large-scale machine learning
Dennis Charles Abts, Eau Claire, WI (US); Jonathan Ross, Palo Alto, CA (US); Garrin Kimmell, Mountain View, CA (US); Michael Bye, Chippewa Falls, WI (US); Matthew Boyd, Gresham, OR (US); and Andrew Ling, Toronto (CA)
Assigned to Groq, Inc., Mountain View, CA (US)
Filed by Groq, Inc., Mountain View, CA (US)
Filed on Nov. 23, 2022, as Appl. No. 17/993,564.
Claims priority of provisional application 63/283,094, filed on Nov. 24, 2021.
Prior Publication US 2023/0161621 A1, May 25, 2023
Int. Cl. G06F 9/48 (2006.01); G06F 9/50 (2006.01); G06F 15/163 (2006.01); G06F 15/78 (2006.01)
CPC G06F 9/4881 (2013.01) [G06F 9/5072 (2013.01); G06F 15/163 (2013.01); G06F 15/7867 (2013.01)] 21 Claims
OG exemplary drawing
 
1. A system comprising:
a network of processors, comprising:
a plurality of processors arranged in a plurality of nodes, comprising at least a first node having a first plurality of deterministic processors connected by local links;
wherein the plurality of nodes are connected by global links; and wherein the plurality of processors communicate with each other to establish a global counter for the network, enabling deterministic communication between the plurality of processors of the network; and
a compiler to explicitly schedule communication traffic across the global and local links with explicit send or receive instructions executed at specific times to establish a specific ordering of operations performed by the network of processors.