DeltaCortex A Series
DeltaCortex A-series is a soft NPU family from Deltacortex and DeltaCortex companies, optimized for FPGA neurocomputing & deep learning.
The advantage of FPGA technology is that it combines high energy efficiency with parallel computing.
In contrast to GPGPU computing, soft neural processing units (NPUs) can support real-time signal processing and high-speed online learning. A combination high-energy efficiency and high-performance makes it possible to use the NPUs in robotics, military, medicine, high-frequency trading, and other areas.
DeltaCortex's neural network processing units are divided into two series: Micro and Stella.
Stella A is designed for Altera FPGAs for high-performance neurocomputing & deep learning tasks.
Micro A is designed for creation of low-cost, energy-efficient neural network controllers and does not require a high-capacity FPGA. With Altera Hardcopy®, you can design an ASIC implementation of Stella A to achieve lower power consumption, better performance and much lower unit cost in production.
The advantage of FPGA technology is that it combines high energy efficiency with parallel computing.
In contrast to GPGPU computing, soft neural processing units (NPUs) can support real-time signal processing and high-speed online learning. A combination high-energy efficiency and high-performance makes it possible to use the NPUs in robotics, military, medicine, high-frequency trading, and other areas.
DeltaCortex's neural network processing units are divided into two series: Micro and Stella.
Stella A is designed for Altera FPGAs for high-performance neurocomputing & deep learning tasks.
Micro A is designed for creation of low-cost, energy-efficient neural network controllers and does not require a high-capacity FPGA. With Altera Hardcopy®, you can design an ASIC implementation of Stella A to achieve lower power consumption, better performance and much lower unit cost in production.
NextMind Micro A1
Micro A1 is a soft NPU from NextMind, optimized for creation of low-cost
energy-efficient neural network controllers with Altera® FPGAs
NextMind Stella A1
Stella A1 is designed for Altera® FPGAs for high-performance deep learning tasks