Synchronization of concurrent computation engines
摘要
說(shuō)明書(shū)
In various examples, the activation engine 516 can include multiple separate execution channels. In these examples, the execution channels can correspond to the columns of the processing engine array 510, and can perform an operation on the outputs of a column, the result of which can be stored in the memory subsystem 504. In these examples, the activation engine 516 may be able to perform between 1 and n parallel computations, where n is equal to the number of columns in the processing engine array 510. In some cases, one or more of the computations can be performed simultaneously. Examples of computations that each execution channel can perform include exponentials, squares, square roots, identities, binary steps, bipolar steps, sigmoidals, and ramps, among other examples.
In some implementations, the accelerator engine 502 can include a pooling engine 518. Pooling is the combining of outputs of the columns of the processing engine array 510. Combining can include for example, computing a maximum value, a minimum value, an average value, a median value, a summation, a multiplication, or another logical or mathematical combination. In various examples, the pooling engine 518 can include multiple execution channels that can operating on values from corresponding columns of the processing engine array 510. In these examples, the pooling engine 518 may be able to perform between 1 and n parallel computations, where n is equal to the number of columns in the processing engine array 510. In various examples, execution channels of the pooling engine 518 can operate in parallel and/or simultaneously. In some examples, the pooling engine 518 can be bypassed.
該功能需要專(zhuān)業(yè)版企業(yè)版VIP權(quán)限，您可以：
掃描加專(zhuān)利匯官方微信交流群
