Parallel Processor Startups

Provider of AI-powered microprocessors for parallel computing applications. Its multi-core technology uses heterogeneous cores in highly concurrent and parallel manners. It accelerates the performance using its TruStream computing as a fabric on its ultra-high core density, ultra-high memory bandwidth data centre chip and reduces latency and power slashing operational costs. It finds applications in fields like big data, machine learning, and finance applications.

Upmem is a processing-in-memory solution to boost the data-intensive applications in the data center. The DPUs (Data processing units) are integrated into the DRAM chips for faster access and larger bandwidth to form a network of co-processors in a scalable parallel environment. It plans to have 2048 cores with 128GByte of RAM per socket. It is packaged on DIMM modules to allow integration with standard servers and comes with an end-to-end SDK for development. The solution is scalable and OS agnostic. The company claims that the solutions could find applications in real-time analytics, pattern matching, database serving and implementation of artificial neural networks. It has partnerships with CEA-Leti, Dolphin Integration, TU Kaiserslautern and Inria.

Provider of AI-based chips for applications. It offers a data processing solution by implementing a serial data processing method known as a parallel data processing method or CNN. It provides a module manufacturing solution for various RF standards such as BLE, Wifi, and 5G, for realizing data processing.

Adapteva is a Massachusetts based fabless semiconductor company developing multicore processor chips for parallel computing. It has adopted an open-source approach for heterogeneous development of parallel programming. It offers Epiphany which is an energy efficient solution for developing chips. Company's Parallela supercomputer is aimed at making parallel computing more accessible.