Some embodiments provide a three-dimensional (3D) circuit structure that has two or more vertically stacked bonded layers with a machine-trained network on at least one bonded layer. As described above, each bonded layer can be an IC die or an IC wafer in some embodiments with different embodiments encompassing different combinations of wafers and dies for the different bonded layers. The machine-trained network in some embodiments includes several stages of machine-trained processing nodes with routing fabric that supplies the outputs of earlier stage nodes to drive the inputs of later stage nodes. In some embodiments, the machine-trained network is a neural network and the processing nodes are neurons of the neural network. In some embodiments, one or more parameters associated with each processing node (e.g., each neuron) is defined through machine-trained processes that define the values of these parameters in order to allow the machine-trained network (e.g., neural network) to perform particular operations (e.g., face recognition, voice recognition, etc.). For example, in some embodiments, the machine-trained parameters are weight values that are used to aggregate (e.g., to sum) several output values of several earlier stage processing nodes to produce an input value for a later stage processing node.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for three-dimensionally stacked neural network accelerators. In one aspect, a method includes obtaining data specifying that a tile from a plurality of tiles in a three-dimensionally stacked neural network accelerator is a faulty tile. The three-dimensionally stacked neural network accelerator includes a plurality of neural network dies, each neural network die including a respective plurality of tiles, each tile has input and output connections. The three-dimensionally stacked neural network accelerator is configured to process inputs by routing the input through each of the plurality of tiles according to a dataflow configuration and modifying the dataflow configuration to route an output of a tile before the faulty tile in the dataflow configuration to an input connection of a tile that is positioned above or below the faulty tile on a different neural network die than the faulty tile.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for three-dimensionally stacked neural network accelerators. In one aspect, a method includes obtaining data specifying that a tile from a plurality of tiles in a three-dimensionally stacked neural network accelerator is a faulty tile. The three-dimensionally stacked neural network accelerator includes a plurality of neural network dies, each neural network die including a respective plurality of tiles, each tile has input and output connections. The three-dimensionally stacked neural network accelerator is configured to process inputs by routing the input through each of the plurality of tiles according to a dataflow configuration and modifying the dataflow configuration to route an output of a tile before the faulty tile in the dataflow configuration to an input connection of a tile that is positioned above or below the faulty tile on a different neural network die than the faulty tile.

Some embodiments of the invention provide an integrated circuit (IC) with a defect-tolerant neural network. The neural network has one or more redundant neurons in some embodiments. After the IC is manufactured, a defective neuron in the neural network can be detected through a test procedure and then replaced by a redundant neuron (i.e., the redundant neuron can be assigned the operation of the defective neuron). The routing fabric of the neural network can be reconfigured so that it re-routes signals around the discarded, defective neuron. In some embodiments, the reconfigured routing fabric does not provide any signal to or forward any signal from the discarded, defective neuron, and instead provides signals to and forwards signals from the redundant neuron that takes the defective neuron's position in the neural network. In some embodiments that implement a neural network by re-purposing (i.e., reconfiguring) one or more individual neurons to implement neurons of multiple stages of the neural network, the IC discards a defective neuron by removing it from the pool of neurons that it configures to perform the operation(s) of neurons in one or more stages of neurons, and assigning this defective neuron's configuration(s) (i.e., its machine-trained parameter set(s)) to a redundant neuron. In some of these embodiments, the IC would re-route around the defective neuron and route to the redundant neuron, by (1) supplying machine-trained parameters and input signals (e.g., previous stage neuron outputs) to the redundant neuron instead of supplying these parameters and signals to the defective neuron, and (2) storing the output(s) of the redundant neuron instead of storing the output(s) of the defective neuron.

Some embodiments of the invention provide a three-dimensional (3D) circuit structure that uses latches to transfer signals between two bonded circuit layers. In some embodiments, this structure includes a first circuit partition on a first bonded layer and a second circuit partition on a second bonded layer. It also includes at least one latch to transfer signals between the first circuit partition on the first bonded layer and the second circuit partition on the second bonded layer. In some embodiments, the latch operates in (1) an open first mode that allows a signal to pass from the first circuit partition to the second circuit partition and (2) a closed second mode that maintains the signal passed through during the prior open first mode. By allowing the signal to pass through the first circuit partition to the second circuit partition during its open mode, the latch allows the signal to borrow time from a first portion of a clock cycle of the second circuit partition for a second portion of the clock cycle of the second circuit partition.

Some embodiments provide a three-dimensional (3D) circuit structure that has two or more vertically stacked bonded layers with a machine-trained network on at least one bonded layer. As described above, each bonded layer can be an IC die or an IC wafer in some embodiments with different embodiments encompassing different combinations of wafers and dies for the different bonded layers. The machine-trained network in some embodiments includes several stages of machine-trained processing nodes with routing fabric that supplies the outputs of earlier stage nodes to drive the inputs of later stage nodes. In some embodiments, the machine-trained network is a neural network and the processing nodes are neurons of the neural network. In some embodiments, one or more parameters associated with each processing node (e.g., each neuron) is defined through machine-trained processes that define the values of these parameters in order to allow the machine-trained network (e.g., neural network) to perform particular operations (e.g., face recognition, voice recognition, etc.). For example, in some embodiments, the machine-trained parameters are weight values that are used to aggregate (e.g., to sum) several output values of several earlier stage processing nodes to produce an input value for a later stage processing node.

Some embodiments provide a three-dimensional (3D) circuit structure that has two or more vertically stacked bonded layers with a machine-trained network on at least one bonded layer. As described above, each bonded layer can be an IC die or an IC wafer in some embodiments with different embodiments encompassing different combinations of wafers and dies for the different bonded layers. The machine-trained network in some embodiments includes several stages of machine-trained processing nodes with routing fabric that supplies the outputs of earlier stage nodes to drive the inputs of later stage nodes. In some embodiments, the machine-trained network is a neural network and the processing nodes are neurons of the neural network. In some embodiments, one or more parameters associated with each processing node (e.g., each neuron) is defined through machine-trained processes that define the values of these parameters in order to allow the machine-trained network (e.g., neural network) to perform particular operations (e.g., face recognition, voice recognition, etc.). For example, in some embodiments, the machine-trained parameters are weight values that are used to aggregate (e.g., to sum) several output values of several earlier stage processing nodes to produce an input value for a later stage processing node.

A system can receive an untrusted onboard announcement message from a remote computer, wherein the untrusted onboard announcement message comprises first data that identifies the remote computer and second data that indicates a current configuration of the remote computer. The system can identify a stored indication of an authorized configuration of the remote computer based on the data that identifies the remote computer. The system can determine that there is a mismatch between the authorized configuration of the remote computer and the current configuration of the remote computer. The system can determine a trust metrics evaluation score for the remote computer based on a type of hardware component change between the authorized configuration of the remote computer and the current configuration of the remote computer. The system can, in response to determining that the trust metrics evaluation score is greater than a threshold value, onboard the remote computer.

Systems and techniques are described for protecting virtual machines from network failures. A described technique includes running a virtual machine on a first source host; replicating, over a first network, data related to the virtual machine to a destination host; determining that the destination host has become unreachable, over the first network, from the first source host; determining whether a second source host can reach the destination host over the first network or a second network; determining whether the virtual machine can run on the second source host; and running the virtual machine on the second source host.

Systems and methods take into account the criticality of workloads, the warranty needs of workloads, the warranty available time, and the lifetime of a workload to provide an optimal solution that ensures servers are used to highest extent. The warranty health of servers is computed and categorized as critical, warning, or healthy based on the number of days remaining in warranty. Workloads are tagged as short-term or long-term workloads. Workloads are also classified based on criticality. The quarantine mode for proactive high availability of servers is divided into multiple modes, including a long-time, critical-workload quarantine mode, a critical-workload quarantine mode, and a standard quarantine mode. Servers that are in quarantine mode are assigned new workloads based upon the warranty health, workload term, and workload criticality.