Distributed phase redundancy
Embodiments herein describe a switchable redundant voltage regulator (referred to herein as a redundant phase) that can be attached to multiple different power domains. Instead of providing a redundant phase for every load or voltage domain in a computing system, the embodiments herein describe sharing redundant phases between voltage domains. Thus, if a primary phase serving any one of the phase fails, the switch can be controlled to connect the redundant phase to the corresponding voltage domain. In this manner, the redundant phase is a shared phase for a plurality of voltage domains.
1 . A system comprising:
a plurality of primary voltage regulators coupled to respective loads;
a plurality of redundant voltage regulators coupled to respective switches, wherein each of the switches is coupled to a plurality of the respective loads; and
a controller configured to:
detect a failure of one of the plurality of primary voltage regulators;
identify a load powered by the failed voltage regulator; and
control one of the respective switches to couple a corresponding one of the plurality of redundant voltage regulators to the identified load.
2 . The system of claim 1 , wherein the plurality of primary voltage regulators, the plurality of redundant voltage regulators, and the respective switches are disposed on a same card.
3 . The system of claim 2 , wherein the respective loads are disposed on one or more cards that are separate from the same card.
4 . The system of claim 3 , wherein the same card and the one or more cards are disposed in an input/output (IO) drawer.
5 . The system of claim 2 , wherein the respective loads are disposed on the same card as the plurality of primary voltage regulators, the plurality of redundant voltage regulators, and the respective switches.
6 . The system of claim 3 , wherein the respective loads are respective processors.
7 . The system of claim 2 , wherein the controller is configured to:
determine when a number of available redundant voltage regulators satisfies a threshold, and in response, transmit an alert to replace the same card.
8 . The system of claim 1 , wherein each of the plurality of primary voltage regulators is coupled to only one of the respective loads, wherein each of the plurality of redundant voltage regulators can be coupled to any one of the respective loads using the respective switches.
9 . A method, comprising:
detecting a failure of one of a plurality of primary voltage regulators, wherein the plurality of primary voltage regulators are coupled to respective loads;
identifying a load powered by the failed voltage regulator; and
controlling, using a controller, a switch to couple a redundant voltage regulator to the identified load, wherein the switch is coupled to a plurality of the respective loads and is separate from the controller.
10 . The method of claim 9 , wherein the plurality of primary voltage regulators, the redundant voltage regulator, and the switch are disposed on a same card.
11 . The method of claim 10 , wherein load is disposed on a card that is separate from the same card.
12 . The method of claim 9 , further comprising:
detecting a failure of a second one of the plurality of primary voltage regulators;
identifying a second load powered by the second voltage regulator; and
controlling a second switch to couple a second redundant voltage regulator to the second load.
13 . The method of claim 9 , further comprising:
determining when a number of available redundant voltage regulators satisfies a threshold, and in response, transmitting an alert to replace a card containing the plurality of primary voltage regulators and a plurality of redundant voltage regulators.
14 . A printed circuit board (PCB), comprising:
a plurality of primary voltage regulators configured to provide power to a plurality of loads;
a plurality of redundant voltage regulators coupled to respective switches, wherein each of the respective switches are configured to couple to the plurality of the loads; and
a controller configured to:
detect a failure of one of the plurality of primary voltage regulators;
identify a load powered by the failed voltage regulator; and
control one of the respective switches to couple a corresponding one of the plurality of redundant voltage to the identified load.
15 . The PCB of claim 14 , wherein the plurality of loads are not disposed on the PCB.
16 . The PCB of claim 14 , wherein the PCB is designed to be inserted into an IO drawer.
17 . The PCB of claim 14 , wherein the controller is configured to:
determine when a number of available redundant voltage regulators satisfies a threshold, and in response, transmit an alert to replace the PCB.
18 . The PCB of claim 14 , wherein each of the plurality of primary voltage regulators is configured to couple to only one of the plurality of loads, wherein each of the plurality of redundant voltage regulators can be coupled to any one of the plurality of loads using the respective switches.
19 . The PCB of claim 18 , wherein each of the plurality of redundant voltage regulators is coupled to a different one of the respective switches.
20 . The PCB of claim 14 , wherein the plurality of loads are disposed on the PCB, wherein the plurality of loads comprises processors.