Written by Haider Saleem
Journalist and Political Analyst | LinkedIn / X
As Nvidia becomes the world’s first $4 trillion company, China’s Huawei is quietly trying to establish itself in the same AI-driven future – with a very different approach.
Rather than chasing performance benchmarks, Huawei is exporting small batches of AI chips and offering remote access to its computing infrastructure. Its goal: to build strategic relationships in markets like the Middle East, Southeast Asia, and Latin America, where digital infrastructure is growing and Western tech isn’t always dominant.
This article explores:
1. What Huawei’s Ascend chips are and how they compare to Nvidia’s
2. Where Huawei is expanding globally
3. Why the U.S. is concerned
4. And how Huawei’s real play may be in infrastructure—not just silicon
1. Inside Huawei’s AI Chips: The Ascend Series
Huawei’s Ascend chip family includes the 910B and 910C models. These chips support tasks such as image classification, text generation, and neural network training. But unlike Nvidia’s state-of-the-art H100 or upcoming GB200 GPUs, Huawei’s chips are better suited to mid-tier workloads.
– Ascend 910B: Huawei’s main export chip, one generation behind Nvidia’s leading GPUs.
– Ascend 910C: Reserved for domestic use and powers Huawei’s CloudMatrix 384 AI compute platform.
Glossary:
Training – Teaching an AI model using large datasets.
Inference – When a trained model makes predictions or generates outputs.
Huawei’s chip production is capped at 200,000 units in 2025, most of which stay in China. It has also stockpiled 2.9 million 910B dies, originally sourced from TSMC.
2. Target Markets: Modest Shipments, Strategic Locations
Huawei is selectively reaching out to global partners:
- United Arab Emirates – MBZUAI denied Huawei engagement, but the outreach reflects ambition.
- Saudi Arabia – Talks with SDAIA are reportedly advanced.
- Malaysia – A proposed 3,000-chip rollout is unofficial; the government has distanced itself from the project.
- Brazil – Huawei is offering data center services—connectivity, storage, and energy—not chips.
3. Why Washington Is Watching Closely
The U.S. Commerce Department has warned that using Huawei’s Ascend 910B, 910C, or 910D chips may violate export restrictions due to U.S. technology used in their development.
Though global restrictions were softened, unlicensed use still risks penalties. The U.S. is weighing expanded export controls on Malaysia and Thailand, and has delayed license approvals in the Gulf.
“Huawei’s small shipments today could lock in tomorrow’s partnerships.” – U.S. Official
4. The Nvidia Benchmark: $4 Trillion and Climbing
While Huawei seeks influence, Nvidia continues to dominate:
– Market Cap: $4 trillion (as of July 2025)
– Q1 Revenue: $44.1 billion, 69% YoY increase
– Q2 Forecast: $45 billion (±2%)
Nvidia powers most major AI platforms: ChatGPT, Meta Llama, Microsoft Copilot. Analysts call it the “foundation of the AI revolution” and its chips the “new gold and oil.”
AMD is also competing with lower-cost chips and has featured in several Gulf-region deals, though it trails in ecosystem integration.
Start Your Halal Stock Screening Journey
New to halal investing? Musaffa makes it easy to screen stocks, check Shariah compliance, and purify your portfolio — all in one place. Make informed, ethical decisions every step of the way.
5. Beyond Silicon: Huawei’s CloudMatrix and the China Stack
Huawei’s AI strategy extends to infrastructure via CloudMatrix 384, which provides remote access to AI compute services from within China.
This is part of a broader “China Stack”:
– Ascend chips – AI hardware
– MindSpore – Open-source AI framework
– CloudMatrix – Compute infrastructure
Glossary:
· CUDA – Nvidia’s software platform for programming GPUs.· DGX – Nvidia’s high-performance AI training server.
This bundled model positions Huawei as a sovereign AI partner for countries seeking alternatives to U.S.-dominated infrastructure.
6. Conclusion: Quiet Expansion with Long-Term Intent
It argued that Huawei isn’t trying to outpace Nvidia in performance. Its chips are older, production is limited, and no major export deals have closed.
But Huawei is showing up – infrastructure first. By building ties with governments and offering alternatives in underserved regions, it is setting the foundation for broader AI influence in the future.
Huawei’s low-volume, high-leverage strategy may prove more lasting than it seems today.
References
1. Bloomberg, Huawei Seeks AI Chip Clients in Middle East, Southeast Asia, July 10, 2025.
2. Reuters, Huawei Eyes Greater Role in Brazil Data Center Market, July 10, 2025.
3. BBC News, Nvidia Becomes World’s First $4tn Company, July 9, 2025.
4. Reuters, Nvidia’s Stock Market Value Hits $4 Trillion on AI Dominance, July 9, 2025.
Disclaimer: The content is for informational purposes only and does not constitute legal, investment or financial advice.
It is important to conduct your own research or consult with a financial or investment advisor. Past performance is not indicative of future results. All logos or brands are referenced for identification purposes only and do not constitute an endorsement of any kind. This information is accurate as of the date of publication and may not reflect recent changes. Access our comprehensive legal disclaimers at https://musaffa.com/disclaimer.
The post Huawei’s Global AI Chip Strategy: Can It Compete in a Nvidia-Dominated World? appeared first on Musaffa Academy.
0 comments