The head background
The head background
  • Quad-Engine CAISA Chip Quad-Engine CAISA Chip
  • 10.9 TOPS High Cost Efficiency
  • 3 ms 3 ms

Nebula Accelerator X3

High Performance, Low Latency and High Cost Efficiency AI Acceleration Solution

  • High Performance High Performance
  • Low Latency Low Latency
  • High Cost Efficiency High Cost Efficiency
  • Easy-to-use User-friendly
  • Easy-to-use High Versatility


Architecture: CAISA 3.0

Chip Utilization Ratio: Up to 95.4%

Peak performance: 10.9 TOPS

PCIe: PCI Express3.0 x8

Memory: 8GB DDR4

Power Supply: from PCIe slot without AUX power

Cooling: Active Cooling Fan

Temperature: -20℃ ~ 70℃

Dimension: 169.5 mm x 69.6mm

(Standard half-length, half-height, single-slot)


3.06 ms

1306.93 FPS


8.68 ms

460.27 FPS


31.06 ms

125.75 FPS


21.96 ms

182.16 FPS

U-Net Industrial

74.07 ms

54.01 FPS

Note: Batch=4, INT8 The above CNN models are created by TensorFlow framework

Empowering Industries’AI Upgrading

  • Image recognition

    Smart City

  • Semantic segmentation

    Smart Manufacturing

  • Time series analysis

    Smart Grid

  • Speech semantics

    Intelligent Safety