Tianshu Zhixin Tiangai BI-V100 32GB
Deep learning acceleration card with 128 TFLOPS peak power, supporting major frameworks like PyTorch and TensorFlow. Supports training transformer-based models (BERT/GPT), computer vision, AI-driven medical diagnostics, and scientific research.
VRAM
32
FP32
32
TDP
250
Interface
PCIe 4.0x16
Pre-order
Product status
Memory Specifications
VRAM
32
VRAM Type
HBM2
Memory Bandwidth
250
Interface
PCIe 4.0x16
Interconnect Type
Inter-chip Interconnect
Interconnect Speed
200-400
Computing Performance
FP64 Vector-
FP32 Vector32
FP16 Vector128
TF32 Tensor-
FP/BF16 Tensor128
FP8 Tensor-
INT8 Tensor256
Graphics & Thermal
Pixel Rate (GPixels/s)
-
Texture Rate (GTexels/s)
-
TDP
250
Cooling Type
Пассивное
Form Factor
PCIe
Video Encoding
-
Video Decoding
-
Physical Dimensions
Slots
2
Length
312 mm
Height
111 mm
Width
39 mm
Architecture
Architecture
CoreX
Cores
-
Price
On request