NEOX™ is a parallel multicore and multithreaded GPU architecture based on the RISC-V RV64IMFC instruction set with adaptive NoC. The number of cores varies from 4 to 64 organized in 1-16 cluster elements, each configured for cache sizes and thread counts. Depending on cluster/core configuration, NEOX™ compute power ranges from 12.8 to 409.6 GFLOPS at 800MHz with support for FP16, and FP32, and SIMD instructions.
Ultra-low-power RISC-V based GPU Processor
Overview
Key Features
- ARCHITECTURE
- RISC V64GC ISA
- Multicore Array
- Multithreaded
- Adaptive NoC
- Configurable 4-64 Cores
- SOFTWARE
- C/C++ LLVM Compiler
- C/C++ GCC Compiler
- Posix Threads
- Open Graphics Frameworks
- EVALUATION
- Xilinx Zynq FPGA
- Cycle Accurate Simulator
Benefits
- Scalable to match multiple applications & performance levels
- NEOX™ is highly configurable in the number of cores, cache sizes and thread count of the array elements. Multithreading hides long latency delays from external memory controller maintaining high computation throughput for the entire array.
Block Diagram
Applications
- GPGPU Compute
- Artificial Intelligence
- Vision Processing
- Video Processing
- Computer Graphics
Deliverables
- System Verilog RTL
- Configuration Tool
- Verification Suite
- Synthesis Scripts
- Software Emulator
- FPGA Prototype
- SDK
Technical Specifications
Related IPs
- Small, ultra-low-power, and very processing-efficient 32-bit processor
- RISC-V processor - 32 bit, 5-stage pipeline
- RISC-V Processor - RV12 - 32/64 bit, Single Core CPU
- Low-power 32-bit RISC-V processor
- 64-bit RISC-V core with in-order dual issue pipeline based complex for Linux-based systems
- 64-bit RISC-V core with out-of-order pipeline based complex for Linux-based embedded systems