RISC-V Edge Computing SoCs
Powered by a 64-bit RISC-V high-performance CPU
and self-developed effcient NPU, it supports full-stack
floating-point computing for generative AI acceleration.
64-bit High-Performance RISC-V Processors
64-bit High-Performance RISC-V Processors
Edge computing SoC equipped with 64-bit out-of-order RISC-V CPU (processors) and self-developed high-performance NPU
Multiple Computing  Acceleration Units
Multiple Computing Acceleration Units
Multiple CV and computing acceleration units including NPU, GPU and DSP, which can be widely applied to various scenarios in computing
Graphics Processing
Graphics Processing
High-performance 3D graphics processing capability
High Computation
High Computation
Processing power for deep learning is up to:
20TOPS (INT8)
10TOPS (INT16)
10FTOPS (FP16)
Rich Peripherals
Rich Peripherals
USB 3.0/2.0, Ethernet RGMII, PCIe 3.0, I2C, HDMI, etc.
Strong Capabilities of  Audio and Video Processing
Strong Capabilities of Audio and Video Processing
Supports video decoding up to 8K@56fps and video encoding up to 8K@25fps, Support audio codecs such as ACC-LC, G.711, G.722.1, etc.
Security and Reliability
Security and Reliability
Hardware encryption engine supports the algorithms of TEE, TRNG, ECC, RSA, ECDSA, AES, DES, HMAC, SM3/4, SHA256, CRC32, etc.
High-precision LLMs
High-precision LLMs
Supports software development frameworks such as Pytorch, Tensorflow, PaddlePaddle, ONNX, etc., and high-precision LLMs