RAPTM’s architecture is based on “agent” as a basic compute element. Each agent is a “group of Tensors” thus packing a whooping performance and enabling a high dimensional compute. These agents operate in multi-agent environment for asynchronous processing of AI algorithms.
Agents are defined as autonomous, problem- solving computational entities capable of operating interactively in dynamic heterogeneous environments. Multiple agents interact, operate individually or in cooperation with other agents, with some degree of autonomy to achieve overarching objectives.
Agents represent the most important new paradigm for combining perception and decision making harmoniously in real life applications.
SIMATM enables Multiple Agents working in groups bringing huge level of parallelism at Agent level thereby significantly increasing rate of computations.
RAPTM’s Instruction Set Architecture enables high data flow as well as high level of program control. High data flow is required for Deep Learning algorithms and high level of program control is required in Reinforcement Learning and Unsupervised Learning.
All fields marked with * are required.
691 S Milpitas Blvd, Suite 217
Milpitas CA 95035-5478
AlphaICs India Pvt Ltd
6th Floor, H2 Block, Mountain Ash Building,
Manyata Embassy Business Park,
Outer Ring Road, Bengaluru 560 045