-
Octa-core ARM Cortex-A55 in a DynamIQ big.LITTLE configuration, up to 1.8 GHz
-
32KB L1 I-cache and 32KB L1 D-cache per A55 core
-
Optional 64KB L2 cache per “LITTLE” core
-
Optional 128KB L2 cache per “big” core
-
ARM G57 MC1 GPU -
Supports OpenGL ES 3.2/2.0/1.1, Vulkan 1.1/1.2/1.3, and OpenCL 2.2 -
Anti-aliasing algorithm -
High memory bandwidth and low power consumption in 3D graphics processing

-
2 TOPS NPU -
Embedded 512KB internal buffer -
Supports deep learning frameworks: TensorFlow, Pytorch, Caffe, Onnx NN, TFLite…….

-
HiFi4 Audio DSP -
32KB I-cache + 32KB D-cache

-
RISC-V CPU, up to 200 MHz -
16 KB I-cache and 16 KB D-cache·RV32IMAFC instructions





