The post Huawei reveals giant new AI chip cluster as Nvidia’s China issues rise appeared on BitcoinEthereumNews.com. A person walks past a display of an Atlas 900 AI cluster at the Huawei stand during the World Artificial Intelligence Conference at the Shanghai World Expo and Convention Center in Shanghai on July 28, 2025. Hector Retamal | Afp | Getty Images BEIJING — Chinese telecommunications giant Huawei announced Thursday new computing systems for powering artificial intelligence with its in-house Ascend chips, as it steps up pressure on U.S. rival Nvidia. The company said it plans to launch its new “Atlas 950 SuperCluster” as soon as next year. The U.S. has sought to cut China off from the most advanced semiconductors for training AI models. To cope, Chinese companies have turned more to grouping large numbers of less efficient, often homegrown, chips together to achieve similar computing capabilities. Under Huawei’s AI computing infrastructure, a supercluster is connected to multiple superpods, which, in turn, are built from multiple supernodes. Supernodes, which form the base, are built on Ascend chips, using system design to overcome technical limitations imposed by U.S. sanctions. Huawei said its new Atlas 950 supernode would support 8,192 Ascend chips, and that the Atlas 950 SuperCluster would use more than 500,000 chips. A more advanced Atlas 960 version, slated for launch in 2027, would support 15,488 Ascend chips per node. The full supercluster would have more than 1 million Ascend chips, according to Huawei. It was not immediately clear how the systems compared with those powered by Nvidia chips. Huawei claimed in a press release that the new supernodes would be the world’s most powerful by computing power for several years. “Huawei’s announcement on its computing breakthrough is well timed with recent increasing emphasis by the Chinese government on self-reliance on China’s own chip technologies,” said George Chen, partner and co-chair, digital practice, The Asia Group. While he… The post Huawei reveals giant new AI chip cluster as Nvidia’s China issues rise appeared on BitcoinEthereumNews.com. A person walks past a display of an Atlas 900 AI cluster at the Huawei stand during the World Artificial Intelligence Conference at the Shanghai World Expo and Convention Center in Shanghai on July 28, 2025. Hector Retamal | Afp | Getty Images BEIJING — Chinese telecommunications giant Huawei announced Thursday new computing systems for powering artificial intelligence with its in-house Ascend chips, as it steps up pressure on U.S. rival Nvidia. The company said it plans to launch its new “Atlas 950 SuperCluster” as soon as next year. The U.S. has sought to cut China off from the most advanced semiconductors for training AI models. To cope, Chinese companies have turned more to grouping large numbers of less efficient, often homegrown, chips together to achieve similar computing capabilities. Under Huawei’s AI computing infrastructure, a supercluster is connected to multiple superpods, which, in turn, are built from multiple supernodes. Supernodes, which form the base, are built on Ascend chips, using system design to overcome technical limitations imposed by U.S. sanctions. Huawei said its new Atlas 950 supernode would support 8,192 Ascend chips, and that the Atlas 950 SuperCluster would use more than 500,000 chips. A more advanced Atlas 960 version, slated for launch in 2027, would support 15,488 Ascend chips per node. The full supercluster would have more than 1 million Ascend chips, according to Huawei. It was not immediately clear how the systems compared with those powered by Nvidia chips. Huawei claimed in a press release that the new supernodes would be the world’s most powerful by computing power for several years. “Huawei’s announcement on its computing breakthrough is well timed with recent increasing emphasis by the Chinese government on self-reliance on China’s own chip technologies,” said George Chen, partner and co-chair, digital practice, The Asia Group. While he…

Huawei reveals giant new AI chip cluster as Nvidia’s China issues rise

2025/09/18 14:02

A person walks past a display of an Atlas 900 AI cluster at the Huawei stand during the World Artificial Intelligence Conference at the Shanghai World Expo and Convention Center in Shanghai on July 28, 2025.

Hector Retamal | Afp | Getty Images

BEIJING — Chinese telecommunications giant Huawei announced Thursday new computing systems for powering artificial intelligence with its in-house Ascend chips, as it steps up pressure on U.S. rival Nvidia.

The company said it plans to launch its new “Atlas 950 SuperCluster” as soon as next year.

The U.S. has sought to cut China off from the most advanced semiconductors for training AI models. To cope, Chinese companies have turned more to grouping large numbers of less efficient, often homegrown, chips together to achieve similar computing capabilities.

Under Huawei’s AI computing infrastructure, a supercluster is connected to multiple superpods, which, in turn, are built from multiple supernodes. Supernodes, which form the base, are built on Ascend chips, using system design to overcome technical limitations imposed by U.S. sanctions.

Huawei said its new Atlas 950 supernode would support 8,192 Ascend chips, and that the Atlas 950 SuperCluster would use more than 500,000 chips.

A more advanced Atlas 960 version, slated for launch in 2027, would support 15,488 Ascend chips per node. The full supercluster would have more than 1 million Ascend chips, according to Huawei.

It was not immediately clear how the systems compared with those powered by Nvidia chips. Huawei claimed in a press release that the new supernodes would be the world’s most powerful by computing power for several years.

“Huawei’s announcement on its computing breakthrough is well timed with recent increasing emphasis by the Chinese government on self-reliance on China’s own chip technologies,” said George Chen, partner and co-chair, digital practice, The Asia Group.

While he cautioned that Huawei might exaggerate its technical capabilities, Chen pointed out that the Chinese company’s ambition to be a world AI leader “cannot be underestimated.”

Research firm SemiAnalysis found in April that Huawei’s self-developed CloudMatrix system was able to perform better than Nvidia’s — despite each Ascend chip delivering only about one-third the performance of an Nvidia processor. Huawei built its advantage by having five times as many chips.

“Computing power has and will continue to be the key for AI,” Eric Xu, vice chairman and rotating chairman of Huawei, said Thursday in a statement, translated by CNBC. He was speaking at the opening of the company’s annual Huawei Connect event in Shanghai. The event runs through Saturday.

Two years ago at the same event, Huawei announced its Atlas 900 SuperCluster. The company currently sells a “Atlas 900 AI Cluster” with “thousands” of Ascend chips.

Rising pressure on Nvidia

Huawei’s announcement comes as China promotes homegrown alternatives to Nvidia. Earlier this week, the two countries wrapped up trade talks in Spain that included a path toward resolving the long-contested U.S. operations of social media app TikTok, owned by Beijing-based startup ByteDance.

In another aggressive signal, China on Monday announced it was extending a probe into Nvidia over alleged monopolistic practices.

Pressure has only risen since on the U.S. chipmaker. Its shares fell more than 2% Wednesday after the Financial Times, citing sources, said China has ordered local tech giants to stop tests and orders of the Nvidia RTX Pro 6000D chip.

Nvidia CEO Jensen Huang told reporters he was “disappointed” to hear the news of the reported ban. He’s previously described Huawei as a “formidable” competitor.

Source: https://www.cnbc.com/2025/09/18/huawei-atlas-950-960-ai-chip-cluster-node-processor-nvidia-china-us-rtx-blackwell.html

Disclaimer: The articles reposted on this site are sourced from public platforms and are provided for informational purposes only. They do not necessarily reflect the views of MEXC. All rights remain with the original authors. If you believe any content infringes on third-party rights, please contact [email protected] for removal. MEXC makes no guarantees regarding the accuracy, completeness, or timeliness of the content and is not responsible for any actions taken based on the information provided. The content does not constitute financial, legal, or other professional advice, nor should it be considered a recommendation or endorsement by MEXC.
Share Insights

You May Also Like

Is the Dollar Losing Its Crown? How AI and Crypto Are Rewiring Global Finance

Is the Dollar Losing Its Crown? How AI and Crypto Are Rewiring Global Finance

The dollar’s dominance has long defined global finance. Yet as central banks trial crypto and AI reshape cross-border settlement, the system faces its first true structural test in decades. This shift could redefine how global liquidity and trust are priced. IMF COFER data place the dollar’s share of global reserves at 56.32% in early 2025 — the lowest since the euro’s birth. Meanwhile, 94% of monetary authorities are testing central-bank digital currencies. That signals diversification and digitalization of state money. AI’s arrival in financial infrastructure accelerates this shift. The Bank for International Settlements warns that autonomous trading and liquidity algorithms could magnify systemic risk. At the same time, new digital rails promise cheaper and faster transfers. Legacy networks built on the greenback are quietly eroding. Indicators of a Permanent Shift in Dollar Dominance BeInCrypto spoke with Dr. Alicia García-Herrero, Chief Economist for Asia-Pacific at Natixis and former IMF economist. Drawing on two decades of macro research, she explains how CBDCs, AI, and stablecoins may redraw global monetary power. She also outlines which metrics will reveal that pivot first. The dollar still anchors reserves, yet erosion has begun. COFER data show a steady slide since 2000. The question is no longer whether alternatives arise, but when the shift becomes measurable — a timeline investors can now watch in real time. Source: IMF COFER, Q2 2025 “From my IMF days analyzing COFER data, we tracked USD’s share of global FX reserves — now 56.32% in Q2 2025 — alongside RMB and EUR gains plus CBDC pilots where 94% of central banks are engaged. Crypto’s volatility could amplify AI-driven risks, as BIS warns. But CBDCs offer controlled shifts. I’d expect measurable erosion if USD dips below 55% by 2027, with $1B+ annual CBDC settlements signaling permanence. Stablecoins buttress dollar stability without wild swings.” Her threshold — a drop below 55% by 2027 plus billion-dollar CBDC flows — would mark a turning point for reserve structures. It shows when diversification stops being theory and becomes policy. Stablecoin Market Share and Emerging Bloc Risks Stablecoins remain an extension of dollar liquidity. Around 99% of circulation is USD-pegged, with USDT and USDC dominant. Non-dollar or commodity-backed tokens could spark bloc-based competition — a clear sign that liquidity may fragment along political lines. Source: Messari “USD-linked stablecoins like USDT and USDC command over 99% of the $300 billion market as of October 2025. A yuan-backed stablecoin hitting 10–15% share could ignite bloc tensions. Conflict only arises if it surpasses 20%, fracturing global liquidity.” García-Herrero argues that a rival stablecoin must capture over 20% of global settlements to trigger true bloc fragmentation. That marks the point where digital currencies start redrawing geopolitics, not just payments. On-chain settlement now tops $35 trillion annually — twice Visa’s throughput. Stablecore CEO Alex Treece calls it “a modern Eurodollar network” serving global USD demand beyond banks. It shows that digital rails still strengthen the dollar’s reach. IMF data show these tokens already handle about 8% of GDP-scale flows in Latin America and Africa. That proves stablecoins now act as informal policy instruments. “Stablecoins satisfy existing dollar demand. It’s market-driven, not state-driven. In the short term they reinforce dominance. In the long term, it depends on US policy and confidence.” Treece compares this digital-dollar system to the 1960s Eurodollar market, when offshore investors tapped US liquidity through parallel networks. Private innovation extended the dollar’s reach instead of replacing it. Stablecoins in High-Inflation Economies In inflation-hit economies like Argentina and Turkey, stablecoins serve as informal dollar rails. They act as a digital hedge against currency collapse and offer a parallel financial lifeline showing crypto’s real-world role. “In Argentina, stablecoins shield 5 million users and make up over 60% of crypto transactions. They become destabilizing at 20–25% of retail payments or 15% of FX turnover. In Turkey, similar adoption ranks it high globally. Overall, their stabilizing role outweighs risks at current levels.” Her rule of thumb: moderate use stabilizes. But when stablecoins exceed a quarter of payments, they threaten monetary sovereignty — the point where relief turns into risk. Tokenization and Sovereign Debt Tokenization has become a key theme in finance, though sovereign uptake lags. While BIS pilots move slowly, private firms advance faster. Franklin Templeton expects early adoption in treasuries and ETFs in Hong Kong, Japan, and Singapore. These pilots show where regulation and innovation already meet. “Institutions want vehicles that manage volatility and enhance liquidity. It starts with retail, but institutional flows follow once secondary markets mature.” — Max Gokhman, Franklin Templeton CoinGecko data show tokenized treasuries above $5.5 billion and stablecoins over $220 billion. The concept is shifting from pilot to practice as traditional assets quietly migrate on-chain. “RWA tokenization’s trillions-by-2030 projections feel ambitious, but tokenized bonds have already hit $8 billion by mid-2025. I foresee 5% of new sovereign issuance by 2028, led by Asia and Europe, while USD resilience will persist.” Her projection — 5% of sovereign issuance tokenized by 2028 — signals gradual reform led by Asia and Europe. It complements rather than replaces the dollar system. Digital finance often evolves through compliance, not rebellion. Both public and private efforts are converging. García-Herrero expects regulator-led uptake, while Franklin Templeton bets on market pull. Either way, traditional assets are migrating to blockchain rails — one bond and one fund at a time. China’s e-CNY and State-Led Crypto China’s e-CNY continues to expand under tight central control. By mid-2025 it had handled 7 trillion yuan in transactions. This shows Beijing’s ability to digitize money without private crypto and how centralized ecosystems can scale quickly. Study Times, the Central Party School’s journal, frames crypto and CBDCs as tools of “financial mobilization.” Beijing’s digital yuan and blockchain networks serve as strategic assets for liquidity control and sanction resilience — a “digital logistics front” merging finance and security. “China’s e-CNY exemplifies disciplined digital finance. It processed 7 trillion RMB by June 2025. A fully state-led model emerges when private blockchain FDI falls below 10% of fintech inflows. By late 2026, we’ll see clear dominance.” She defines state-led dominance as private blockchain investment under 10% of fintech inflows. That level may arrive by late 2026, when digital sovereignty becomes measurable, not rhetorical. Russia–China Trade and the “State-Led Web3 Bloc” Facing sanctions, Russia and China now settle most trade outside the dollar system. Their digital-asset experiments raise the question of when coordination becomes a formal bloc — a turning point that could reshape settlement geography. “Russia’s 2025 legalization of crypto for foreign trade, with non-USD/EUR flows now over 90% in yuan and ruble, shows how a ‘state-led Web3 bloc’ could emerge if 50% of trade shifts to digital assets. CBDC bridges might mitigate risk, and ironically, USD-pegged stablecoins could stabilize such flows.” Her 50% benchmark defines the threshold for a new clearing sphere. It could stabilize sanctioned trade yet deepen global fragmentation. Europe has already reacted. The EU’s recent ban on a ruble-backed stablecoin, A7A5, marked its first direct crypto sanction. It showed how digital assets have become both weapon and target in financial conflict. Proof of Personhood and Financial Inclusion Proof-of-Personhood systems like Worldcoin’s biometric model are reframing debates on identity and inclusion. Their economic value remains unproven, yet scalability could shape how fast AI-age trust frameworks evolve. “Proof-of-Personhood pilots like Worldcoin, with 200 million identities verified by mid-2025, could cut borrowing costs by 50–100 basis points or lift capital access by 20–30%. If achieved by 2027, it would validate PoP beyond hype.” The debate mirrors the wider digital-identity race. TFH’s Adrian Ludwig sees proof-of-human systems as a trust layer for an AI age. García-Herrero says only measurable impact will prove their worth. AI and Crypto Cross-Border Trade Dominance AI-driven finance now shapes liquidity, compliance, and settlement. The BIS says machine-learning copilots already automate AML reviews. Project Pine smart contracts let central banks adjust collateral in real time, signaling programmable compliance’s rise. BIS frames this as a programmable yet regulated financial core. Speculative outlooks like AI 2027 imagine AI systems directing liquidity, R&D, markets, and security policy. BIS calls for integrity-by-design before such systems fully emerge. “AI’s cross-border edge will surge, with 75% of payments becoming instant by 2027. China seems poised for over 30% share through state-backed sandboxes and nearly $100 billion in investments. Stablecoins could complement AI agents, curbing volatility.” Investments nearing $100 billion by 2027 favor that model. Stablecoins may serve as compliant, tokenized layers linking automated liquidity to programmable money — the next battleground for regulators. Sovereign Bitcoin Reserves and Resource Bottlenecks Bitcoin’s share in sovereign reserves remains small yet symbolic. Its link to risk assets and reliance on energy and chips may create new geopolitical choke points. Digital reserves could soon tie to physical supply chains. “Sovereign Bitcoin reserves remain under 1% of total FX. Hitting 5% by 2030 would spark a volatile ‘digital gold race.’ Energy and semiconductor supply could become choke points, while stablecoins offer a steadier reserve alternative.” Meanwhile, digital-asset treasury (DAT) firms manage over $100 billion in crypto, revealing how fragile balance sheets can mirror sovereign risk. Bitcoin-focused treasuries with strict liquidity buffers appear most resilient — a preview of challenges nations may face as adoption rises. Transparency of Crypto and Governance Advantage Public blockchains are entering government registries and procurement systems. For democracies, transparent ledgers offer accountability that directly strengthens fiscal credibility. “Blockchain procurement pilots boost transparency in democracies like Estonia, with government adoption markets jumping from $22.5 billion in 2024 to nearly $800 billion by 2030. At 15–20% of national spend on-chain, democracies gain a structural edge.” Her 15–20% benchmark marks the point when blockchain adoption becomes structural. It raises transparency scores and gives open societies a governance advantage. Conclusion Across ten domains — CBDCs, AI, stablecoins, tokenization, and blockchain — García-Herrero’s framework suggests evolution, not revolution. The dollar’s reach is diffusing, not disappearing, as digital money turns monetary power into a shared, data-driven system. Her analysis grounds speculation in measurable data: reserve ratios, settlement flows, and adoption thresholds. The future monetary order will hinge less on disruption than on governance — how transparency, trust, and control align in the digital age.
Share
2025/10/28 07:53