Moore Threads MTT S80 vs NVIDIA B100 SXM 192GB

The B100 delivers 4.1x the FP32 throughput of the MTT S80 (60 vs 15 TFLOPS dense). The B100 also carries 176 GB more memory (192 GB vs 16 GB).

MTT S80: MUSA ChunXiao, 2024 B100: Blackwell, 2024

FP32 dense lead

4.1x

B100: 60 vs 15 TFLOPS

Memory

16 vs 192 GB

176 GB more for the B100

Bandwidth

448 GB/s vs 8.0 TB/s

B100 moves data faster

Download comparison sheet Printable version →

Moore Threads MTT S80

Full specs →

What is the Moore Threads MTT S80?

The MTT S80 is a GPU from Moore Threads on the MUSA ChunXiao architecture, an accelerator for AI training, inference and high-performance computing. It is a PCIe add-in card, so it fits standard servers and workstations. MTT S80 full specifications →

Can I rent the MTT S80 in the cloud?

Flopper is not tracking any live rental listings for the MTT S80 yet. The provider directory lists the clouds that rent comparable accelerators on demand. GPU cloud providers →

How does the MTT S80 compare with other accelerators?

This page sets the MTT S80 against the B100. Both parts launched in the same year. This is a cross-vendor matchup, Moore Threads against NVIDIA; the MTT S80 is a PCIe card while the B100 is an SXM module. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the MTT S80 figures come from?

Every MTT S80 figure on this page traces to Moore Threads's own datasheet or specification table. How Flopper verifies specifications →

NVIDIA B100 SXM 192GB

Leads on FP32 Full specs →

What is the NVIDIA B100?

The B100 is a GPU from NVIDIA on the Blackwell architecture, an accelerator for AI training, inference and high-performance computing. It is an SXM module, the mezzanine format used in dense multi-GPU servers. B100 full specifications →

Can I rent the B100 in the cloud?

Flopper is not tracking any live rental listings for the B100 yet. The provider directory lists the clouds that rent comparable accelerators on demand. GPU cloud providers →

How does the B100 compare with other accelerators?

This page sets the B100 against the MTT S80. Both parts launched in the same year. This is a cross-vendor matchup, NVIDIA against Moore Threads; the B100 is an SXM module while the MTT S80 is a PCIe card. Flopper also compares it head to head with the B200 SXM 180GB, across every published precision, memory and power. B100 vs B200 SXM 180GB →

Where do the B100 figures come from?

NVIDIA published the B100 figures in marketing material rather than a datasheet, and Flopper has no independent confirmation of them. How Flopper verifies specifications →

Specifications

MTT S80

B100

Architecture

MUSA ChunXiaoBlackwell

Launch Year

20242024

Form Factor

PCIeSXM

Memory

16 GB192 GB

Memory Bandwidth

448 GB/s8.0 TB/s

TDP

—700 W

Process Node

—4nm

Spec

MTT S80

B100

Architecture MUSA ChunXiaoBlackwell

Launch Year 20242024

Form Factor PCIeSXM

Memory 16 GB192 GB

Memory Bandwidth 448 GB/s8.0 TB/s

TDP —700 W

Process Node —4nm

Performance (TFLOPS)

MTT S80

B100

30 TFLOPS

15 TFLOPS 60 TFLOPS

875 TFLOPS

sparse not published

1,750 TFLOPS

sparse not published

1,750 TFLOPS

sparse not published

3,500 TFLOPS

sparse not published

FP6

7,000 TFLOPS

sparse not published

3,500 TOPS

sparse not published

Precision

MTT S80

B100

30 TFLOPS

15 TFLOPS 60 TFLOPS

875 TFLOPS

sparse not published

1,750 TFLOPS

sparse not published

1,750 TFLOPS

sparse not published

3,500 TFLOPS

sparse not published

FP6

7,000 TFLOPS

sparse not published

3,500 TOPS

sparse not published

Figures are dense. Where a vendor publishes a figure it sits under the dense one and never decides a winner. Hover any precision for its definition, or read the GPU glossary.

FLOPS by Precision

What actually differs

Both launched in 2024, the MTT S80 on MUSA ChunXiao and the B100 on Blackwell.

The B100 carries 176 GB more memory (192 GB against 16 GB), so larger models fit on one GPU before you shard. Bandwidth is 448 GB/s against 8.0 TB/s, which sets the ceiling on decode speed.

Weight the FP8 and FP16 rows for LLM work and the FP64 row for HPC.

New to the terms? The glossary explains FP64, FP32, TF32, BF16, FP16, FP8, FP4, INT8, Sparsity and TFLOPS. For how the generations relate, see NVIDIA architectures explained.

Frequently asked questions

Is the MTT S80 faster than the B100?

At FP32 precision the B100 reaches 60 TFLOPS dense against 15 TFLOPS for the MTT S80. The performance table on this page lists every published precision for both GPUs.

Which should you choose, the MTT S80 or the B100?

Choose the B100 for training and high-throughput inference: it has 4.1x the FP32 throughput of the MTT S80. It also carries 176 GB more memory, so larger models fit on one GPU before you shard. Packaging differs too: the MTT S80 is a PCIe card while the B100 is an SXM module, so check what your servers accept.

Which has more memory, the MTT S80 or the B100?

The B100 carries 192 GB of memory versus 16 GB for the MTT S80. Memory bandwidth is 448 GB/s for the MTT S80 and 8.0 TB/s for the B100.

Get Comparison Updates

New GPUs added weekly. Be the first to see how they compare.

Open in Full Compare Tool → MTT S80 full specs B100 full specs Estimate cluster cost in the calculator →

Moore Threads MTT S80 vs NVIDIA B100 SXM 192GB

Moore Threads MTT S80

What is the Moore Threads MTT S80?

Can I rent the MTT S80 in the cloud?

How does the MTT S80 compare with other accelerators?

Where do the MTT S80 figures come from?

NVIDIA B100 SXM 192GB

What is the NVIDIA B100?

Can I rent the B100 in the cloud?

How does the B100 compare with other accelerators?

Where do the B100 figures come from?

Specifications

Performance (TFLOPS)

FLOPS by Precision

What actually differs

Frequently asked questions

Is the MTT S80 faster than the B100?

Which should you choose, the MTT S80 or the B100?

Which has more memory, the MTT S80 or the B100?

Related comparisons

Get Comparison Updates

Stay Updated on GPU Specs