/Patent Tracker
GooglePatent applicationChips & hardwareInterest 4/10

Google patents system to fix broken connections in large computer clusters

Patent drawing
Patent drawing from the official filing.

This filing describes a method for managing groups of high-speed computer chips connected by fast data links. If one link breaks, the system identifies the specific group of chips affected, marks that group as temporarily unavailable to prevent errors, and sends a repair instruction to a technician.

Why it mattersIt helps data centers keep running smoothly by quickly isolating broken parts of a computer cluster so they can be fixed without shutting down the entire system.

In the filing’s own words

A system for repair of an accelerator architecture including a plurality of accelerator host having a plurality of accelerator elements in each host and high-speed interconnects between accelerator elements and including a link health monitor service in the host determining a status of each interconnect internal and external to the host. The link health monitor service detects link status of the interconnects and communicates this to a pod manager responsible for managing a number of interconnected hosts. The pod manager communicates with a scheduler for scheduling processing tasks to each accelerator element or group of elements. Upon detection of an interconnect problem, the pod manager flags a least common ancestor in a binary tree of elements containing elements adjacent to the failed interconnect and flags the ancestor and its parent nodes as unavailable. Pod manager further generates a corrective action repairing the problem interconnect communicating the action to a technician.

What exactly they claim (claim 1)

1. A pod manager apparatus for managing distributed execution of workflows across a pod of accelerator hosts, the apparatus comprising: at least one processor; and a memory storing instructions that, when executed by the at least one processor, cause the pod manager apparatus to: receive a link status of interconnects from a link health monitor service of each of a plurality of accelerator hosts; responsive to receiving an indication of a failed interconnect connecting a first accelerator element and a second accelerator element, determine a least common ancestor slice in a hierarchical tree of slices, wherein the least common ancestor slice is the lowest slice in the hierarchical tree that contains both the first accelerator element and the second accelerator element; flag the least common ancestor slice and all parent slices of the least common ancestor slice in the hierarchical tree as unavailable; and communicate an updated slice availability status indicating the unavailable slices to a scheduler to disable communication with the unavailable slices.

Facts

Company
Google (Google LLC)
Official title
System And Method Of Drainless Link Repair For Increased Accelerator Utilization During Failure
Publication number
US20260310560A1
Application number
19711390
Filed
Jun 17, 2026
Published
Oct 8, 2026
Inventors
9
Claims
10

Classified under

G06F 11/1616 Physics & computingG06F 2201/805 Physics & computing
Read the full filing (official PDF) ↗ Open in the Patent Tracker

More from Google in Chips & hardware

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.