Huawei Ascend 950DT vs Huawei Atlas 350

The Ascend 950DT delivers 29% more FP8 throughput than the Atlas 350 (1,034 vs 804 TFLOPS dense). The Ascend 950DT also carries 32 GB more memory (144 GB vs 112 GB).

Ascend 950DT: Da Vinci v3, 2026 Atlas 350: Da Vinci v3, 2026

FP8 dense lead

+29%

Ascend 950DT: 1,034 vs 804 TFLOPS

Memory

144 vs 112 GB

32 GB more for the Ascend 950DT

Bandwidth

4.0 TB/s vs 1.4 TB/s

Ascend 950DT moves data faster

Download comparison sheet Printable version →

Huawei Ascend 950DT

Leads on FP8 Full specs →

What is the Huawei Ascend 950DT?

The Ascend 950DT is an NPU from Huawei on the Da Vinci v3 architecture, an accelerator built for AI inference. It is an OAM module, the open accelerator format used in multi-GPU servers. Ascend 950DT full specifications →

Can I rent the Ascend 950DT in the cloud?

Flopper is not tracking any live rental listings for the Ascend 950DT yet. The provider directory lists the clouds that rent comparable accelerators on demand. GPU cloud providers →

How does the Ascend 950DT compare with other accelerators?

This page sets the Ascend 950DT against the Atlas 350. Both parts launched in the same year. Both are Huawei parts. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the Ascend 950DT figures come from?

Every Ascend 950DT figure on this page traces to Huawei's own datasheet or specification table. How Flopper verifies specifications →

Huawei Atlas 350

Full specs →

What is the Huawei Atlas 350?

The Atlas 350 is an NPU from Huawei on the Da Vinci v3 architecture, an accelerator built for AI inference. Atlas 350 full specifications →

Can I rent the Atlas 350 in the cloud?

Flopper is not tracking any live rental listings for the Atlas 350 yet. The provider directory lists the clouds that rent comparable accelerators on demand. GPU cloud providers →

How does the Atlas 350 compare with other accelerators?

This page sets the Atlas 350 against the Ascend 950DT. Both parts launched in the same year. Both are Huawei parts. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the Atlas 350 figures come from?

Every Atlas 350 figure on this page traces to Huawei's own datasheet or specification table. How Flopper verifies specifications →

Specifications

Ascend 950DT

Atlas 350

Architecture

Da Vinci v3Da Vinci v3

Launch Year

20262026

Form Factor

OAMCustom

Memory

144 GB112 GB

Memory Bandwidth

4.0 TB/s1.4 TB/s

TDP

—600 W

Process Node

——

Spec

Ascend 950DT

Atlas 350

Architecture Da Vinci v3Da Vinci v3

Launch Year 20262026

Form Factor OAMCustom

Memory 144 GB112 GB

Memory Bandwidth 4.0 TB/s1.4 TB/s

TDP —600 W

Process Node ——

Performance (TFLOPS)

Ascend 950DT

Atlas 350

30 TFLOPS

273 TFLOPS

547 TFLOPS 425 TFLOPS

1,034 TFLOPS 804 TFLOPS

FP6

804 TFLOPS

2,007 TFLOPS 1,561 TFLOPS

1,034 TOPS 804 TOPS

Precision

Ascend 950DT

Atlas 350

30 TFLOPS

273 TFLOPS

547 TFLOPS 425 TFLOPS

1,034 TFLOPS 804 TFLOPS

FP6 804 TFLOPS

2,007 TFLOPS 1,561 TFLOPS

1,034 TOPS 804 TOPS

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 2026, the Ascend 950DT on Da Vinci v3 and the Atlas 350 on Da Vinci v3.

The Ascend 950DT carries 32 GB more memory (144 GB against 112 GB), so larger models fit on one GPU before you shard. Bandwidth is 4.0 TB/s against 1.4 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 FP32, TF32, BF16, FP16, FP8, FP4, INT8, Sparsity and TFLOPS.

Frequently asked questions

Is the Ascend 950DT faster than the Atlas 350?

At FP8 precision the Ascend 950DT reaches 1,034 TFLOPS dense against 804 TFLOPS for the Atlas 350. The performance table on this page lists every published precision for both GPUs.

Which should you choose, the Ascend 950DT or the Atlas 350?

Choose the Ascend 950DT for training and high-throughput inference: it has 29% more FP8 throughput than the Atlas 350. It also carries 32 GB more memory, so larger models fit on one GPU before you shard.

Which has more memory, the Ascend 950DT or the Atlas 350?

The Ascend 950DT carries 144 GB of memory versus 112 GB for the Atlas 350. Memory bandwidth is 4.0 TB/s for the Ascend 950DT and 1.4 TB/s for the Atlas 350.

Get Comparison Updates

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

Open in Full Compare Tool → Ascend 950DT full specs Atlas 350 full specs Estimate cluster cost in the calculator →

Huawei Ascend 950DT vs Huawei Atlas 350

Huawei Ascend 950DT

What is the Huawei Ascend 950DT?

Can I rent the Ascend 950DT in the cloud?

How does the Ascend 950DT compare with other accelerators?

Where do the Ascend 950DT figures come from?

Huawei Atlas 350

What is the Huawei Atlas 350?

Can I rent the Atlas 350 in the cloud?

How does the Atlas 350 compare with other accelerators?

Where do the Atlas 350 figures come from?

Specifications

Performance (TFLOPS)

FLOPS by Precision

What actually differs

Frequently asked questions

Is the Ascend 950DT faster than the Atlas 350?

Which should you choose, the Ascend 950DT or the Atlas 350?

Which has more memory, the Ascend 950DT or the Atlas 350?

Get Comparison Updates

Stay Updated on GPU Specs