T-Head Zhenwu M890 vs T-Head Zhenwu 810E
Side-by-side specifications, performance, and rental pricing for datacenter AI workloads.
Zhenwu M890: Zhenwu, 2026 Zhenwu 810E: Zhenwu, 2026
Memory
144 vs 96 GB
48 GB more for the Zhenwu M890
T-Head Zhenwu M890
Full specs →Architecture
Zhenwu
Year
2026
Memory
144 GB
T-Head Zhenwu 810E
Full specs →Architecture
Zhenwu
Year
2026
Memory
96 GB
Specifications
Architecture
ZhenwuZhenwu
Launch Year
20262026
Form Factor
——
Memory
144 GB96 GB
Memory Bandwidth
—2.7 TB/s
TDP
——
Process Node
——
Spec Zhenwu M890 Zhenwu 810E
Architecture ZhenwuZhenwu
Launch Year 20262026
Form Factor ——
Memory 144 GB96 GB
Memory Bandwidth —2.7 TB/s
TDP ——
Process Node ——
Performance (TFLOPS)
FP64
No verified data available No verified data available
FP32
No verified data available No verified data available
TF32
No verified data available No verified data available
BF16
No verified data available No verified data available
FP16
No verified data available No verified data available
FP8
No verified data available No verified data available
FP6
No verified data available No verified data available
FP4
No verified data available No verified data available
INT8
No verified data available No verified data available
Precision Zhenwu M890 Zhenwu 810E
FP64 No verified data available No verified data available
FP32 No verified data available No verified data available
TF32 No verified data available No verified data available
BF16 No verified data available No verified data available
FP16 No verified data available No verified data available
FP8 No verified data available No verified data available
FP6 No verified data available No verified data available
FP4 No verified data available No verified data available
INT8 No verified data available No verified data available
FLOPS by Precision
What actually differs
Both GPUs launched in 2026: the Zhenwu M890 on Zhenwu and the Zhenwu 810E on Zhenwu.
Memory is 144 GB against 96 GB. More memory per GPU means larger models fit before you have to shard across cards, which often matters more than raw TFLOPS for inference.
For LLM training and inference, weight the FP8 and FP16 rows and memory capacity most heavily. For scientific and HPC workloads, the FP64 row is the one to read.
Frequently asked questions
Which has more memory, the Zhenwu M890 or the Zhenwu 810E?
The Zhenwu M890 carries 144 GB of memory versus 96 GB for the Zhenwu 810E.
Get Comparison Updates
New GPUs added weekly. Be the first to see how they compare.