Alibaba unveils its own AI chip for training and inference
Alibaba has announced an in-house AI chip designed for both training and inference. The chip is general-purpose: it can be used to train large language models in the lab and to run workloads in production. The move is a key step in the company's vertical integration strategy, which aims to control every part of its cloud service. It also reduces Alibaba's dependence on the US company Nvidia.
AI-processed from Bloomberg Tech; edited by Hamidun News
Alibaba Group has added a new chip to its growing AI stack, continuing a strategy of covering every aspect of artificial intelligence development.
About the new processor
The company did not disclose the chip's exact name, but said it is designed for both training large language models and inference. That means the hardware is general-purpose: it can be used in a data center for training and then repurposed for production workloads.
The processor is built with a focus on energy efficiency. Alibaba stressed that the chip performs best with its own cloud service, Aliyun, and integrates with the company's AI frameworks.
Why Alibaba is building hardware
The Chinese giant's strategy is clear: control the entire AI development chain. Until recently, the company had to rely on Nvidia GPUs, but American chips are expensive and their export is restricted by sanctions. An in-house processor reduces that dependence.
- Operating costs lowered by 15-20%
- Full control over the cloud stack
- An edge over competitors using the same technology
- Fast optimization for Aliyun
- A fallback strategy in case of further sanctions
Alibaba is already investing in other parts of the ecosystem: its own training framework, the Qwen models and hardware accelerators.
The competitive context
Competition in custom AI chips is growing. Google has released the TPU, and Meta is developing its own accelerators. Even startups such as Groq are building specialized processors.
But chip development demands huge investment and semiconductor industry experience. Alibaba has both: its own cloud expertise (Aliyun) and access to manufacturing chains through its contacts in China. That gives it an advantage over purely cloud-focused companies in the West.
What this means
The announcement shows that the big cloud providers no longer want to depend on a single hardware supplier. Each is building its own stack: software, models and chips. For Alibaba, this is a step toward independence; for users, it may mean more choice and, possibly, lower prices for AI infrastructure.
*Meta has been designated an extremist organization and is banned in Russia.
- Nov 1, 202616 models tested on 7 search tools: Exa averaged 10.5 out of 14, SearXNG about 3
- Nov 1, 2026Google's StreamVibe video agent beats other models by 10.7% on response timing and quality
- Nov 1, 2026OpenAI Rewrites GPT Behavior Rules: Five Values and a Chain of Command
- Nov 1, 2026Voice bots built on LLMs turn to Jev, a fast classifier, for decisions in 200–500 ms
Key Questions Answered
What are the key details on About the new processor?
The company did not disclose the chip's exact name, but said it is designed for both training large language models and inference. That means the hardware is general-purpose: it can be used in a data center for training and then repurposed for production workloads. The processor is built with a focus on energy efficiency. Alibaba stressed that the chip performs best with its own cloud service, Aliyun, and integrates with the company's AI frameworks.
Why Alibaba is building hardware?
The Chinese giant's strategy is clear: control the entire AI development chain. Until recently, the company had to rely on Nvidia GPUs, but American chips are expensive and their export is restricted by sanctions. An in-house processor reduces that dependence. Operating costs lowered by 15-20% Full control over the cloud stack An edge over competitors using the same technology Fast optimization for Aliyun A fallback strategy in case of further sanctions Alibaba is already investing in other parts of the ecosystem: its own training framework, the Qwen models and hardware acc
What are the key details on The competitive context?
Competition in custom AI chips is growing. Google has released the TPU, and Meta is developing its own accelerators. Even startups such as Groq are building specialized processors. But chip development demands huge investment and semiconductor industry experience. Alibaba has both: its own cloud expertise (Aliyun) and access to manufacturing chains through its contacts in China. That gives it an advantage over purely cloud-focused companies in the West.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.