Skip to main content

4 posts tagged with "MTT S5000"

摩尔线程 MTT S5000 全功能 GPU

View all tags

摩尔线程 MTT S5000 出货与生态最新进展

· 6 min read
AI Compute Cards Wiki Editorial
Industry Research Team

国产 GPU 厂商中，摩尔线程是 2026 年上半年财报最亮眼的一家。据其 2026 年半年度报告，MTT S5000 智算集群已实现规模化销售，营收 17.36 亿元、同比增长 147.42%，这一数字已超越 2025 年全年的 15.06 亿元。规模化的信号不是发布PPT，而是在无锡的智算中心投运两个多月就接近满载。

财务与出货：规模化销售的最硬证据​

据公开报道，摩尔线程 2026 年上半年实现营业收入 17.36 亿元，同比增长 147.42%；毛利总额 9.89 亿元，同比增长 103.78%，整体毛利率 56.95%；归属于上市公司股东的净利润为 -1156.31 万元，较上年同期 2.71 亿元的亏损大幅收窄 95.73%；扣非净利润 -1.51 亿元，亏损收窄 52.37%。

关键的一句话在半年报里：旗舰级 AI 训推一体智算卡 MTT S5000（PH100 芯片）智算集群已实现规模化销售，并在北京、无锡、杭州等多地完成部署与交付，且首批通过国家《安全可靠测评》。

指标2026 年上半年2025 年全年同比
营业收入17.36 亿元15.06 亿元+147.42%
毛利总额9.89 亿元—+103.78%
整体毛利率56.95%——
归母净利润-1156.31 万元—亏损收窄 95.73%
扣非净利润-1.51 亿元—亏损收窄 52.37%

这组数字的含义需要放在国产算力的语境里理解：国产 GPU 长期被质疑"只能做政府试点、无法支撑生产负载"。营收超过上年全年、亏损收窄 95%，意味着 S5000 正在从验证走向可重复的集群交付。

无锡智算中心：两个多月接近满载​

最有说服力的单一案例来自无锡。据报道，摩尔线程在无锡惠山区建设的 AI 算算中心于 2026 年 6 月 11 日投运，一期运行 2000 张 MTT S5000，搭建在摩尔线程自研的夸娥（KUAE）智算集群系统之上，投运一个多月后利用率接近满载。

该中心已有 20 多家客户在测试或使用集群，包括中国头部高校、国家实验室、科研院所与 AI 初创企业，其中无锡本地用户占比近半。基于该中心算力资源落地的 AI 智能体应用已有六类，涵盖政务、医疗、教育、芯片设计等场景——医疗领域无锡惠山区人民医院与南京邮电大学合作部署了近 30 个临床辅助模型；教育领域华东师范大学启创平台集成了 300 多个教育 AI 智能体；EDA 领域芯华章利用该算力加速低功耗无线芯片的端到端设计。摩尔线程副总裁方朝霞表示，目标规模是迈向万卡级算力。

训练与生态：国芯训国模的成果​

据公开报道，在 2026 年 9 月 3 日的半年度业绩说明会上，摩尔线程董事长、总经理张建中表示，公司基于 MTT S5000 智算集群已完成 MoE-236B 超大参数基础大模型从零训练，并完成智源具身大脑模型 RoboBrain 2.5、5D 世界模型 EvoPhys-World 等前沿模型的全流程训练。

这一成果的量级需要留意：MoE-236B 的语料超过 25 万亿 Token，是从零起步完整训练的；北大 EvoPhys 团队则在 S5000 全功能 GPU 上完成了 5D 世界模型 EvoPhys-World 的全栈原生训练。摩尔线程联合创始人张钰勃进一步介绍，公司已打通具身智能从合成数据生成、大模型训练、仿真验证到端侧硬件部署的完整链路。

产品侧的三个动作值得关注：MTT C256 超节点架构规划采用计算与交换一体化的高密设计，服务万卡乃至十万卡级训练与推理集群；PD（Prefill-Decode）异构分离方案把高算力需求的 Prefill 计算池（S5000）与高带宽的 Decode 生成池分池部署，支持 KV Cache FP8 与 1M 超长上下文；此外摩尔线程联合中国移动研究院、之江实验室等产业伙伴，通过 OISA 协同创新平台正式发布了《OISA 高密超节点参考设计技术规范》。

需要注意的约束​

规模化的另一面是产能瓶颈。据公开报道的行业调研，S5000 在阿里仅有少量采购测试，字节做了小规模灰度测试（几百片），原因是客户担忧量产后供应；2026 年 4-5 万卡的出货目标仍可能有之江实验室、运营商及地方算力平台项目支撑。此外，高阶双带版本尚未量产，主要卡在双带性能和带间互联指标，需重新流片优化；S6000 目前未回片，预计 2026 年底提供小批量样品。

生态侧也有坦率之处：S5000 尚未发布足够多的独立复现基准，官方性能指标多为厂商声明。同时 PD 异构分离方案中的 Decode 池使用国际主流 GPU，这说明纯国产方案在长上下文解码环节仍有工程缺口。

研判​

摩尔线程 2026 年上半年交出的答卷，是国产 GPU 商业化进程中少见的硬数据：营收翻倍以上、集群接近满载、模型训练跑通闭环、拿到安全可靠测评。但从投资与采购视角看，仍有三个问题待解：营收集中度与复购比例尚未披露、产能是否支撑 4-5 万卡的年度目标、以及 S6000 回片后能否补上 HBM 容量短板。

《OISA 高密超节点参考设计技术规范》的发布是一个容易被忽视的信号——摩尔线程正在从卖卡转向输出系统标准，这与昇腾走超节点架构、百度走天池超节点是同一方向。国产算力的竞争维度，已经从单卡参数转向系统与标准的话语权争夺。

摩尔线程双线作战：MTT S5000 智算卡与"庐山"图形 GPU 的AI+游戏并举

· 6 min read
AI Compute Cards Wiki Editorial
Industry Research Team

当多数国产 AI 芯片公司all in 大模型算力时，摩尔线程依然坚持"两条腿走路"：一手是 MTT S5000 智算卡撑起的夸娥万卡集群，一手是瞄准 3A 游戏的"庐山"图形 GPU。9 月初的半年度业绩说明会给出了两条线的最新坐标——这条"AI+游戏"并举的路线，正在从权宜之计变成差异化壁垒。

一、MTT S5000：平湖架构的训推一体旗舰​

MTT S5000 基于第四代 MUSA"平湖"架构，搭载 PH100 芯片、4096 个自研 MUSA Core，2025 年 2 月公开参数。核心规格如下：

项目参数
架构MUSA 第四代（平湖），PH100 芯片
显存80 GB（颗粒类型官方未公开）
显存带宽1.6 TB/s
FP8 稠密算力1000 TFLOPS（液冷版）/ 920 TFLOPS（风冷版）
BF16/FP16400 TFLOPS
FP32100 TFLOPS
INT8800 TOPS
互联MTLink 784 GB/s（8 卡节点内全互联）
TDP500 W（市场口径）
形态OAM / PCIe，液冷 + 风冷双版本
单价~$4000–6000（OAM）

几点澄清值得注意：其一，"单卡 AI 稠密算力 1000 TFLOPS"特指 FP8 精度（液冷版），并非 FP16；其二，80GB 显存的颗粒类型官方从未公布，市场流传的"GDDR6X"说法与 1.6 TB/s 带宽自相矛盾，本站已订正为"类型未公开"；其三，S5000 配备硬件级 FP8 Tensor Core，支持 FP8→FP64 全精度，并已基于 FP8 完成 DeepSeek-V4 的 Day-0 适配。加上《安全可靠测评结果公告（2026 年第 2 号）》的通过——AI 训推芯片首次纳入该体系——S5000 同时拿到了信创市场的入场券。

二、万卡落地：夸娥集群与京东云合作​

S5000 的成色，最终要靠集群说话。基于夸娥（KUAE）万卡智算集群的实测数据：

指标实测值
集群算力10 EFLOPS
Dense 大模型训练 MFU60%
MoE 大模型训练 MFU40%
训练线性扩展效率95%（64 卡→1024 卡保持 90%+）
有效训练时间占比>90%
第三方对齐智源 RoboBrain 2.5 与 H100 集群 loss 差异仅 0.62%

95% 的线性扩展效率背后是 ACE 异步通信引擎——把通信任务从计算核心卸载，实现计算与通信零冲突并行。商业侧，2026 年 3 月签下 6.6 亿元夸娥集群订单；软件侧，MiniMax M3、智谱 GLM-5.2、阿里 Qwen3.5、美团万亿参数 LongCat-2.0 相继完成 Day-0 适配。最关键的落地信号来自 9 月 19 日：京东云与摩尔线程合作的万卡级集群已部署上线服务——从"能跑通"到"规模化服务外部客户"，国产 GPU 集群跨过了商用验证的分水岭。随后的数贸会（9 月 23–27 日，杭州）上，摩尔线程集中展示了 AI 智算卡、万卡级集群，以及"模型训练工厂、词元生产工厂、智能体生产工厂"三层能力，产品覆盖云-边-端全场景。

三、"庐山"GPU：从兼容显卡到渲染范式革命​

图形产品线上，9 月 3 日半年度业绩说明会披露了新一代"庐山"GPU 的指标——基于花港架构，预计 2026 年底推出：

  • 3A 游戏渲染性能提升 15 倍；
  • 光线追踪性能提升 50 倍；
  • AI 性能提升 64 倍。

比倍数更重要的是两项渲染技术的组合：一是硬件光线追踪，原生支持 DirectX Raytracing，让国产 GPU 第一次在光追管线层面与主流标准对话；二是AI 生成式渲染 AGR，基于全自研 MTAGR 1.0，把渲染范式从"逐像素计算"推向"生成"——用 AI 直接生成画面内容，而非暴力算出每个像素。前者补齐了传统管线的硬件短板，后者则是在渲染范式切换窗口期的卡位：当英伟达 DLSS、AMD FSR 都在走向 AI 化时，AGR 让摩尔线程有机会在下一代渲染标准里占据一席，而不是永远追赶。

四、游戏反哺 AI：双架构复用的独特路线​

把两条产品线放在一起看，摩尔线程的打法浮现出独特逻辑：图形与 AI 共享同一套 MUSA 架构底座——SIMT 核心、张量单元、光追核心、统一软件栈全部复用。游戏 GPU 的价值因此不只是营收 diversification：海量游戏场景是最好的架构压力测试与现金流来源，图形侧打磨的调度、显存与光速能力，反向反哺 AI 卡的通用计算质量；而 AI 业务积攒的先进制程流片经验，又拉高图形芯片的天花板。

已登陆科创板的摩尔线程，如今可以用上市平台的融资能力同时养两条产品线。这与纯 AI 芯片公司"单点押注大模型"的路线形成鲜明对比——在国产算力需求旺盛但波动性尚存的市场里，双线作战既是对冲，也是壁垒。

小结​

摩尔线程的 2026 年验证了"全功能 GPU"路线的可行性：MTT S5000 以 FP8 稠密 1000 TFLOPS、80GB 显存撑起夸娥万卡集群，并通过京东云万卡级集群部署完成了规模化商用验证；"庐山"GPU 则以硬件光追与 AI 生成式渲染 AGR 完成了从"国产兼容显卡"到"渲染范式探索者"的技术跃迁。图形+AI 双架构复用的独特路线，让摩尔线程在国产 AI 芯片阵营中占据了一个难以复制的生态位。2026 年底"庐山"如期推出与否，将是这条双线故事的下一次大考。

Domestic GPU IPO Wave: The "Four Little Dragons" Assemble on Capital Markets, Moore Threads MTT S5000 Benchmarks Against H100

· 5 min read
AI Compute Cards Wiki Editorial
Industry Research Team

From December 2025 to July 2026 — just half a year — at least 6 AI chip companies have listed or are about to list on capital markets. Together with already-listed Cambricon, Hygon, and Iluvatar, the domestic GPU corps' total market cap is approaching ¥2 trillion. This marks the critical climb from domestic GPUs being "usable" to "useful."

1. The "Four Little Dragons" assemble on capital markets​

CompanyListing statusRaise / issue priceSponsor
Moore ThreadsListed (STAR Market sh688795, 2025-12-05)Issue price ¥114.28, raised ¥8BCITIC Securities
MetaXIPO accepted (2026-06-30)¥3.904B (total investment ¥5B)Huatai United
EnflamePassed review (2026-06-15)¥6B—
BirenHKEX / sprinting——

Already-listed camp: Cambricon (sh688256, STAR Market 2020-07-20), Hygon, Iluvatar (HKEX). Moore Threads turned a book profit of ¥29.35M in Q1; MetaX narrowed losses 57.7% and gave a 2026 breakeven timeline.

2. Moore Threads MTT S5000: benchmarking against H100​

Moore Threads announced its flagship AI train+inference GPU MTT S5000 successfully completed full-pipeline adaptation validation of Zhipu's new-generation large model GLM-5 — measured performance "breaks the domestic compute ceiling":

MetricMTT S5000
Architecture4th-gen "Pinghu" architecture
FP8 compute1 PFLOPS (1,000 TFLOPS)
Memory bandwidth1.6 TB/s
PositioningFull-function train+inference GPU, benchmarks against NVIDIA H100
ProductionMass-produced; clusters online supporting trillion-parameter training

Deployment validation: jointly completed full-pipeline training of embodied-brain model RoboBrain 2.5 with BAAI; partnered with SiliconFlow for high-performance DeepSeek-V3 inference, single-card speed near international top products. IPO funds go to three directions: next-gen AI train+inference chip, next-gen graphics chip, next-gen AI SoC chip.

WAIC 2026 new progress: Moore Threads showcased the MTT C256 SuperNode (first-of-its-kind single-layer Scale-up 256-card full interconnect, sub-microsecond latency) and three AI-factory solutions — "model training factory / token production factory / agent production factory"; the company pre-announced H1 2026 revenue of ¥1.65B-1.75B, up 135%-149% YoY.

3. Cambricon: dual flagships MLU590/690​

ChipProcessComputeMemoryCustomer / status
MLU590 (思元590)7nm ChipletINT8 512 TOPS / FP16 345 TFLOPS96 GB HBM2eByteDance inference mainstay, ~80% of A100 overall, mass shipments early 2026
MLU690 (思元690)5nm-class (SMIC N+2)FP16 700+ TFLOPS / INT8 2800+ TOPS196 GB HBM3 (3.35 TB/s)Dual-die packaging, MLU-Link 890 Gbps; ~70% of H100 (80-90% pure inference); ByteDance largest customer, mass production early 2026

Cambricon is the only domestic AI chip vendor with a "unified edge-cloud architecture" — one MLU instruction set spans 思元 220 (edge) → 370 (border) → 590/690 (cloud), with one NeuWare toolchain across compute tiers.

Capital and performance double explosion: Cambricon's total market cap exceeded ¥1 trillion on June 30, 2026, becoming the STAR Market's first "trillion-yuan stock," up 75%+ YTD. On performance, Q1 2026 revenue ¥2.885B (+160% YoY), deducted net profit ¥934M; full-year 2025 revenue ¥6.497B (+453% YoY), net profit attributable to parent ¥2.059B, ending long-term losses. ByteDance has cumulatively deployed over 100k 思元 590/690, its largest customer.

4. DeepSeek-V4 effect: changing the expectation coordinate system​

On April 24, 2026, DeepSeek released the trillion-parameter flagship DeepSeek-V4. Unlike a year earlier when V3's launch sparked debate over "can domestic chips even run large models," this time multiple domestic chips — Huawei Ascend, Cambricon, Hygon, MetaX, Moore Threads, Kunlun, T-Head, Iluvatar — completed adaptation on launch day.

The evaluation coordinate system is shifting: from "what percentage of NVIDIA's same-generation product performance" to "can it carry the real workloads of top-tier large models."

Industry interpretation​

  1. Capital ammunition in place: dense IPOs provide ample funding for domestic GPU R&D iteration and capacity expansion, moving from "technology breakthrough" to "commercial virtuous cycle."
  2. Train+inference becomes the mainstream route: Moore Threads takes the full-function GPU route (graphics+AI+general compute), differentiating from Huawei Ascend's "AI-focused."
  3. Software ecosystem is the decider: Day-0 adaptation and the maturity of unified software stacks (MUSA / NeuWare / MXMACA) are replacing raw peak compute as the core yardstick of domestic GPU "usability."

References​

This article continuously tracks the domestic GPU listing process and product iteration.

China's Domestic AI Chip Triopoly (2026): Ascend, Cambricon, Moore Threads — Who Is the "China H100"?

· 7 min read
Charles Qing
AI Hardware Analyst

Against the backdrop of U.S. export controls, China's AI chip market is forming a "three-way standoff." This article compares the technical routes, product specs, software ecosystems, and commercial progress of the three major domestic AI chip vendors: Huawei Ascend, Cambricon MLU, and Moore Threads MTT.

Key Points​

  • Huawei Ascend: leader in domestic AI training chips; Ascend 950 in mass production; most mature software ecosystem
  • Cambricon MLU690: the "China H100," compute close to H200, clear efficiency advantage
  • Moore Threads MTT S5000: full-function GPU route; achieved Day-0 support for Qwen3.5 and GLM-5.2 in June 2026
  • Shared challenge: affected by U.S. export controls, primarily aimed at the Chinese market, limited internationally

I. Vendor Overview​

VendorFoundedFounderListed2025 RevenueMain Customers
Huawei Ascend2018 (division)Ren Zhengfeiprivate (wholly owned by Huawei)~¥20B (est.)Chinese gov, SOEs, military
Cambricon2016Chen Tianshi (CAS)2020-07 (STAR Market 688256)~¥5.2BByteDance, Alibaba, Baidu
Moore Threads2020Zhang Jianzhong (ex-NVIDIA China)2023-12 (STAR Market 688495)~¥1.5B (est.)gov, SOEs, gaming cos.

Strategic Positioning​

VendorTech routeCore strengthMain challenge
Huawei AscendAI-training-specific (Da Vinci)co-optimized HW/SW, carrier channelssanctions, process limits
CambriconAI-training-specific (MLUarch)high efficiency, competitive priceimmature ecosystem
Moore ThreadsFull-function GPU (MUSA)graphics + AI + general compute, Day-0 supportcompute below dedicated AI chips

II. Flagship Product Comparison​

1. Huawei Ascend 950DT (2026 flagship)​

ItemSpec
BF16 compute1,000 TFLOPS
Memory144GB HiZQ 2.0 (in-house HBM)
Memory bandwidth4 TB/s
TDP400W
ProcessN+2 (improved 7nm)
Released2026-04
Mass production2026-Q2
Unit price~¥80,000 (est.)

Strengths:

  • ✅ High large-model inference throughput: 144GB memory friendly to DeepSeek R1 (671B MoE)
  • ✅ Most mature ecosystem: CANN ~85% operator coverage, supports PyTorch, TensorFlow
  • ✅ Strong carrier channel: China Mobile, China Telecom large purchases

Weaknesses:

  • ❌ Process limited: N+2 below TSMC 4nm
  • ❌ Mediocre efficiency: 400W TDP, 2.5 TFLOPS/W

2. Cambricon MLU690 (2026 flagship)​

ItemSpec
BF16 compute600 TFLOPS
Memory64GB HBM3
Memory bandwidth2 TB/s
TDP280W
ProcessTSMC 7nm
Released2025-Q4
Mass production2026-Q1
Unit price~¥140,000 (est.)

Strengths:

  • ✅ Best efficiency: 280W TDP, 2.14 TFLOPS/W (1.5x H100)
  • ✅ Competitive price: ~$20,000, 33% cheaper than H100
  • ✅ Top-tier customer orders: ByteDance, Alibaba, Baidu

Weaknesses:

  • ❌ Small memory: 64GB limits large-model training scale
  • ❌ Immature ecosystem: NeuWare ~75–85% coverage; complex LLMs need manual tuning

3. Moore Threads MTT S5000 (2025 flagship)​

ItemSpec
FP16 compute~1,000 TFLOPS (est.)
Memory80GB GDDR6X
Memory bandwidth1.6 TB/s
TDP~350W
ProcessTSMC 4nm (est.)
Released2025-02
Mass production2025-Q2
Unit price~¥50,000 (est.)

Strengths:

  • ✅ Full-function GPU: graphics + AI + general compute, broader scenarios
  • ✅ Strong Day-0 support: June 2026 Day-0 support for Qwen3.5, GLM-5.2, MiniMax M3
  • ✅ Lowest price: ~¥50,000, high cost-performance

Weaknesses:

  • ❌ Compute below dedicated AI chips: FP16 ~50% of H100
  • ❌ Low memory bandwidth: 1.6 TB/s (48% of H100), limits large-model training

III. Compute Comparison (BF16/FP16)​

ChipBF16 computeMemoryBandwidthTDPEfficiency
Huawei Ascend 950DT1,000 TFLOPS144GB4 TB/s400W2.5 TFLOPS/W
Cambricon MLU690600 TFLOPS64GB2 TB/s280W2.14 TFLOPS/W
Moore Threads MTT S5000~1,000 TFLOPS80GB1.6 TB/s~350W~2.86 TFLOPS/W
NVIDIA H100989 TFLOPS80GB3.35 TB/s700W1.41 TFLOPS/W
NVIDIA H200989 TFLOPS141GB4.8 TB/s700W1.41 TFLOPS/W

Key insights:

  1. Ascend 950DT has the highest compute (1,000 TFLOPS) but mediocre efficiency
  2. Cambricon MLU690 has the best efficiency (2.14 TFLOPS/W), TDP only 280W
  3. Moore Threads MTT S5000 wins on full-function versatility but low bandwidth

IV. Software Ecosystem​

VendorStackFramework supportCoverageMaturity
Huawei AscendCANNPyTorch, TensorFlow, MindSpore~85%⭐⭐⭐⭐ (4/5)
CambriconNeuWarePyTorch-Cambricon, TensorFlow-Cambricon~75–85%⭐⭐⭐ (3/5)
Moore ThreadsMUSIFYPyTorch, TensorFlow, ONNX~70%⭐⭐⭐ (3/5)
NVIDIACUDAall~99%⭐⭐⭐⭐⭐ (5/5)

Ecosystem Maturity Assessment​

Huawei Ascend CANN:

  • ✅ Strength: highest operator coverage, supports MindSpore (in-house framework)
  • ❌ Weakness: steep learning curve, incomplete docs

Cambricon NeuWare:

  • ✅ Strength: PyTorch/TensorFlow compatible, low migration cost
  • ❌ Weakness: complex LLMs need manual tuning

Moore Threads MUSIFY:

  • ✅ Strength: strong Day-0 support, ONNX support
  • ❌ Weakness: lowest operator coverage, dual graphics+AI engine complexity

V. Commercial Progress​

Vendor2026 commercial progressMain customersShipments
Huawei AscendAscend 950 mass production; China Mobile large purchaseChina Mobile, China Telecom, gov~100K/yr (est.)
CambriconMLU690 mass production; ByteDance, Alibaba ordersByteDance, Alibaba, Baidu~50K/yr (est.)
Moore ThreadsMTT S5000 mass production; Day-0 Qwen3.5gov, SOEs, gaming cos.~30K/yr (est.)

Latest as of June 2026​

Huawei Ascend:

  • ✅ Ascend 950DT fully ramping
  • ✅ ¥1B procurement agreement with China Mobile

Cambricon:

  • ✅ MLU690 in volume shipment
  • ✅ ByteDance order ~20K units

Moore Threads:

  • ✅ Day-0 support for Qwen3.5, GLM-5.2, MiniMax M3
  • ✅ MTT S5000 2nd-gen released

VI. Selection Advice​

Scenario 1: Trillion-parameter training (GPT-4 class)​

Recommended: Huawei Ascend 950DT

  • ✅ 144GB large memory supports super-large models
  • ✅ Most mature ecosystem (~85% coverage)
  • ✅ Strong carrier channel, Chinese government backing

Alternative: Cambricon MLU690 (high efficiency, but small memory)

Scenario 2: Tens-to-hundreds-of-billions parameter training​

Recommended: Cambricon MLU690

  • ✅ Best efficiency (2.14 TFLOPS/W), low TCO
  • ✅ Competitive price (~$20,000)
  • ✅ Validated by top customers (ByteDance, Alibaba)

Alternative: Huawei Ascend 920 (more compute, mediocre efficiency)

Scenario 3: Cloud AI inference​

Recommended: Huawei Ascend 950PR (inference-specific)

  • ✅ Well-optimized inference throughput
  • ✅ 128GB memory friendly to MoE models
  • ✅ Mature stack, low deployment cost

Alternative: Moore Threads MTT S5000 (full-function GPU, inference + graphics)

Scenario 4: Edge AI / on-device inference​

Recommended: Moore Threads MTT S5000

  • ✅ Full-function GPU, graphics + AI
  • ✅ Lowest price (~¥50,000)
  • ✅ Strong Day-0 support

Alternative: Huawei Ascend 310 (low power, 8W TDP)

Scenario 5: Domestic substitution (gov, SOEs)​

Recommended: Huawei Ascend 950DT

  • ✅ Chinese government first choice, carrier bulk buys
  • ✅ Co-optimized HW/SW, stable performance
  • ✅ Supported by national semiconductor fund

Alternative: Cambricon MLU690 (high efficiency, competitive price)

VII. Future Roadmap​

Vendor2026 H220272028
Huawei Ascend950DT ramp960 (FP8 ~2 PFLOPS)970 (N+3 process)
CambriconMLU690 rampMLU790 (5nm, BF16 ~1,000 TFLOPS)MLU890 (3nm)
Moore ThreadsMTT S5000 2nd-genMTT S6000 (HBM3, FP16 ~1,500 TFLOPS)MTT S7000

VIII. Summary: Who Is the "China H100"?​

DimensionAscend 950DTMLU690MTT S5000
Compute⭐⭐⭐⭐⭐ (5/5)⭐⭐⭐ (3/5)⭐⭐⭐ (3/5)
Memory⭐⭐⭐⭐⭐ (5/5)⭐⭐ (2/5)⭐⭐⭐ (3/5)
Efficiency⭐⭐⭐ (3/5)⭐⭐⭐⭐⭐ (5/5)⭐⭐⭐⭐ (4/5)
Ecosystem⭐⭐⭐⭐ (4/5)⭐⭐⭐ (3/5)⭐⭐⭐ (3/5)
Price⭐⭐⭐ (3/5)⭐⭐⭐⭐ (4/5)⭐⭐⭐⭐⭐ (5/5)
Overall⭐⭐⭐⭐ (4/5)⭐⭐⭐ (3/5)⭐⭐⭐ (3/5)

Final conclusion:

  • Huawei Ascend 950DT is the domestic AI training chip closest to H100, strongest overall
  • Cambricon MLU690 is the most efficient domestic AI chip, lowest TCO
  • Moore Threads MTT S5000 is the cheapest full-function GPU, suited to edge AI and graphics+AI

References​

Disclaimer: Data based on public sources; actual specs per vendor official. MirrorFrog continuously updates domestic AI chip data — corrections welcome.

Changelog: 2026-06-23 initial release