Huawei Ascend 910B
Huawei's workhorse AI chip in China since 2023: 64 GB of HBM2e at 1.6 TB/s on SMIC's 7 nm (N+2) process, rented through Huawei Cloud and bought by Baidu, Tencent and iFLYTEK. Sources and verification dates.
Every published figure
Source per value- Owner The company that specified the chip and runs it in its own cloud, data centers or products.
- Huawei 2
- Role What the owner says the chip is for: training, inference, both, or a general-purpose server CPU.
- Training and inference 2
- Design partner The ASIC house that co-designed or physically implemented the chip (Broadcom, Marvell, MediaTek, Alchip, GUC and others), as stated by the owner, the partner, a filing or named trade-press reporting. 'In-house' when the owner states it designed the chip itself, including through an acquired team such as Annapurna Labs.
- In-house (HiSilicon) 2
- Foundry The fab that manufactures the die (TSMC, Samsung, SMIC, Intel).
- SMIC 2
- Node Process node as the owner or foundry states it, for example 'TSMC N3' or '7 nm'.
- SMIC 7 nm (N+2) 2
- Memory On-package memory type (HBM2e, HBM3, HBM3E, LPDDR5) or, for CPUs, the memory the socket drives (DDR5).
- HBM2e 2
- Memory capacity GB On-package memory per chip in GB. Blank for CPUs, whose memory sits off-package.
- 64 1
- Memory bandwidth TB/s Peak memory bandwidth per chip in TB/s, as published.
- 1.6 2
- FP16 TFLOPS Published FP16 throughput per chip or accelerator module in TFLOPS. See the precision note for shared-format figures, sparsity and module scope.
- 400 1
- Precision note What the throughput figures mean: dense or sparse, which precision, per chip or per package, and any figure the owner states only in another unit.
- 400 FP16 TFLOPS and 64 GB HBM, stated in the Huawei Cloud co-authored DeepServe paper (USENIX ATC 2025, section 2). CSET independently estimates about 400 FP16 TFLOPS for the fastest variant; variants differ. 1
- Status Announced: public roadmap without verified deployment. Sampling: first silicon or limited internal use. Preview: customers can request early access, without a GA claim. Deployed: stated production use or general availability. Discontinued: program stopped.
- Deployed 2
- Announced Year the owner first announced the chip publicly.
- 2,022 2
- Deployment scale The largest deployment the owner or a named report states: chip counts, clusters, gigawatts or regions.
- Bought by Baidu, Tencent and iFLYTEK (CSET, 2024). 2
- How to get it How an outside buyer can use the chip: a named cloud instance or service, a product, or 'Internal only'.
- Huawei Cloud; Atlas servers 2
Not verified in the sources reviewed: partner basis, bf16, mxfp8, mxfp4, mx8, mx4, fp8, int8, fp4, power, largest pod.