Sambanova Systems, Inc.

Bigstream Solutions, Inc.

Weiwei Chen

David Alan Koeplinger

Kevin James Brown

Raghu Prabhakar

Xiaoming Gu

Sumti Jairath

Ruddhi Arun Chaphekar

Sitanshu Gupta

Ajit Punj

Yaqi Zhang

Matthew Feldman

Behnam Robatmili

Fei Wang

Maysam Lavasani

Weihang Fan

Adam Bordelon

Mark Wagner

John David Davis

The technology disclosed provides a system that comprises a processor with computing units on an integrated circuit substrate. The processor is configured to map a program across multiple hardware stages with each hardware stage executing a corresponding operation of the program at a different stage latency dependent on an operation type and an operand format. The system further comprises a runtime logic that configures the compute units with configuration data. The configuration data causes first and second producer hardware stages in a given compute unit to execute first and second data processing operations and produce first and second outputs at first and second stage latencies, and synchronizes consumption of the first and second outputs by a consumer hardware stage in the given compute unit for execution of a third data processing operation by introducing a register storage delay that compensates for a difference between the first and second stage latencies.

Compiler-based input synchronization for processor with variant stage latencies

A method in a reconfigurable computing system includes connecting a plurality of tensor consumers to their corresponding tensor producers via skip-buffers, which generates a plurality of skip-buffers. The method includes determining that at least one skip-buffer of the plurality of skip-buffers corresponding to a first set of tensor consumers and at least one skip-buffer of the plurality of skip-buffers corresponding to a second set of tensor consumers, are compatible to wholly or partially merge. The method also includes merging, wholly or partially, the compatible skip-buffers to produce a merged skip-buffer having a minimal buffer depth. The described method may reduce memory unit consumption and latency.

Merging skip-buffers in a reconfigurable dataflow processor

A method for merging buffers and associated operations includes receiving a compute graph for a reconfigurable dataflow computing system and conducting a buffer allocation and merging process responsive to determining that a first operation specified by a first operation node is a memory indexing operation and that the first operation node is a producer for exactly one consuming node that specifies a second operation. The buffer allocation and merging process may include replacing the first operation node and the consuming node with a merged buffer node within the graph responsive to determining that the first operation and the second operation can be merged into a merged indexing operation and that the resource cost of the merged node is less than the sum of the resource costs of separate buffer nodes. A corresponding system and computer readable medium are also disclosed herein.

Merging buffer access operations in a coarse-grained reconfigurable computing system

The technology disclosed relates to storing a dataflow graph with a plurality of compute nodes that transmit data along data connections, and controlling data transmission between compute nodes in the plurality of compute nodes along the data connections by using control connections to control writing of data.

Flow control for reconfigurable processors

A method for partitioning executable operations for a reconfigurable computing system includes receiving a set of expressions comprising a plurality of operations and dependencies for those operations, partitioning the plurality of operations into selected executable partitions wherein each selected executable partition conforms to resource constraints for a reconfigurable unit of the reconfigurable computing system. Partitioning the plurality of operations into selected executable partitions may include seeding a candidate partition with an operation, recursively generating an additional candidate partition for each operation adjacent to the candidate partition whose dependent operations are already within the candidate partition or a previously selected partition, and selecting a best candidate partition based on resource cost. A corresponding system and computer-readable medium are also disclosed herein. The system includes a partitioning module that that partitions the plurality of operations into selected executable partitions according to the method describe above.

Partitioning dataflow operations for a reconfigurable computing system through recursive candidate generation

A dataflow graph for an application has operation units that are configured to be producers and consumers of tensors. A write access pattern of a particular producer specifies an order in which the particular producer generates elements of a tensor, and a read access pattern of a corresponding consumer specifies an order in which the corresponding consumer processes the elements of the tensor. The technology disclosed detects conflicts between the producers and the corresponding consumers that have mismatches between the write access patterns and the read access patterns. A conflict occurs when the order in which the particular producer generates the elements of the tensor is different from the order in which the corresponding consumer processes the elements of the tensor. The technology disclosed resolves the conflicts by inserting buffers between the producers and the corresponding consumers.

Compile time logic for inserting a buffer between a producer operation unit and a consumer operation unit in a dataflow graph

A compiler configured to configure memory nodes with a ready-to-read credit counter and a write credit counter. The ready-to-read credit counter of a particular upstream memory node initialized with as many read credits as a buffer depth of a corresponding downstream memory node. The ready-to-read credit counter configured to decrement when a buffer data unit is written by the particular upstream memory node into the corresponding downstream memory node, and to increment when the particular upstream memory node receives from the corresponding downstream memory node a read ready token. The write credit counter of the particular upstream memory node initialized with one or more write credits and configured to decrement when the particular upstream memory node begins writing the buffer data unit into the corresponding downstream memory node, and to increment when the particular upstream memory node receives from the corresponding downstream memory node a write done token.

Anti-congestion flow control for reconfigurable processors

A dataflow graph has operation units that are configured to be producer operation units to produce tensors for execution of the application, and to be consumer operation units to consume the tensors for execution of the application. Compile time logic is configured to process the dataflow graph to determine, for the tensors, expected producer memory layouts, expected consumer memory layouts, and current memory layouts. The expected producer memory layouts specify memory layouts required by the producer operation units that produce the tensors. The expected consumer memory layouts specify the memory layouts required by the consumer operation units that consume the tensors. The current memory layouts specify the memory layouts of the tensors. Each of the memory layouts includes a vector dimension and at least one of a vector ordering and a data alignment.

Systems and methods for memory layout determination and conflict resolution

Compile time logic for detecting streaming compatible and broadcast compatible data access patterns

For one embodiment of the present invention, methods and systems for accelerating data operations with efficient memory management in native code and native dynamic class loading mechanisms are disclosed. In one embodiment, a data processing system comprises memory and a processing unit coupled to the memory. The processing unit is configured to receive input data, to execute a domain specific language (DSL) for a DSL operation with a native implementation, to translate a user defined function (UDF) into the native implementation by translating user defined managed software code into native software code, to execute the native software code in the native implementation, and to utilize a native memory management mechanism for the memory to manage object instances in the native implementation.

Growing community of inventors

Mountain View, CA, United States of America

Weiwei Chen