DeepSeek Eyes $45B Valuation as AI Funding Surge Continues

A wave of massive funding rounds and strategic partnerships is reshaping the global AI landscape this week, with DeepSeek reportedly approaching a $45 billion post-money valuation, Moonshot AI closing in on $2 billion in new financing, and Corning striking a $500 million equity deal with NVIDIA to deepen their infrastructure collaboration.

These developments arrive alongside AMD's dramatic decision to double its 2030 server market forecast to $120 billion — a signal that even chip giants are recalibrating their expectations upward as AI demand continues to surge.

Key Takeaways

• DeepSeek is reportedly in talks that could value the company at $45 billion, though sources say the valuation has not been finalized
• Moonshot AI (月之暗面) is set to complete a $2 billion funding round, pushing its valuation past $20 billion
• Corning and NVIDIA signed a $500 million equity financing agreement to strengthen their long-term partnership
• AMD doubled its 2030 server market size projection from $60 billion to $120 billion
• SpaceX proposed a $55 billion 'Terafab' manufacturing project in Texas
• Samsung announced a full withdrawal of home appliance sales from the Chinese market
• China's Ministry of Industry and Information Technology approved the country's first commercial satellite IoT trial

DeepSeek's Valuation Could Reach $45 Billion

The Chinese AI lab DeepSeek, which stunned the global tech industry earlier this year with its highly efficient open-source models, is reportedly in discussions that could place its post-money valuation at approximately $45 billion. Sources familiar with the matter confirmed that negotiations are underway but stressed that the final valuation has not yet been locked in.

This potential valuation represents a staggering rise for a company that gained international attention by demonstrating that competitive large language models could be trained at a fraction of the cost incurred by Western labs like OpenAI and Anthropic. DeepSeek's R1 model, released in January 2025, triggered a brief selloff in NVIDIA stock as investors questioned whether the massive capital expenditures planned by hyperscalers were truly necessary.

If confirmed at $45 billion, DeepSeek would rank among the most valuable AI startups globally, trailing only OpenAI (valued at roughly $300 billion) and potentially rivaling Anthropic's most recent valuation. The figure underscores how Chinese AI companies are attracting significant capital despite ongoing U.S. export controls on advanced chips.

Moonshot AI Nears $20 Billion Valuation With $2B Round

Moonshot AI, the Beijing-based startup behind the popular Kimi chatbot, is on track to close a $2 billion funding round that would push its valuation beyond $20 billion. The round represents one of the largest single fundraises by a Chinese AI company in 2025.

Founded in 2023 by former Google researcher Yang Zhilin, Moonshot AI has rapidly scaled its consumer-facing AI assistant to become one of the most-used chatbots in China. Kimi gained particular traction for its ability to process extremely long context windows — a technical capability that differentiated it from competitors early on.

The $2 billion raise comes at a time when Chinese AI startups are benefiting from renewed investor enthusiasm following DeepSeek's proof that cost-efficient model development is viable. Compared to U.S. AI startups, which have attracted over $30 billion in venture funding in the past 12 months alone, Chinese counterparts are now closing the capital gap.

• Moonshot AI's valuation has grown roughly 10x since its Series A
• The Kimi chatbot competes with Baidu's Ernie Bot and Alibaba's Qwen
• Long-context processing remains a key differentiator for Moonshot's models
• The round signals continued investor appetite for Chinese AI consumer products

Corning and NVIDIA Forge $500M Infrastructure Alliance

In a move that highlights the growing importance of physical infrastructure in the AI boom, Corning Incorporated and NVIDIA have entered a $500 million equity financing agreement designed to deepen their long-term strategic partnership. The deal focuses on Corning's optical connectivity solutions, which are critical for linking GPU clusters in large-scale AI data centers.

As AI training runs grow larger and more distributed across thousands of GPUs, the demand for high-bandwidth, low-latency optical interconnects has skyrocketed. Corning, best known for its Gorilla Glass used in smartphones, has become an increasingly important player in AI infrastructure through its fiber optic and optical cable products.

The partnership reflects NVIDIA's strategy of securing its supply chain and ecosystem partners well in advance. CEO Jensen Huang has repeatedly emphasized that the 'AI factory' model requires not just GPUs but also networking, cooling, power delivery, and optical connectivity to function at scale. By investing directly in Corning, NVIDIA is ensuring that a critical bottleneck — the physical links between compute nodes — does not constrain the buildout of next-generation data centers.

This deal follows a pattern of NVIDIA making strategic investments in ecosystem partners, similar to its earlier collaborations with companies in the networking and cooling spaces.

AMD Doubles Its 2030 Server Market Forecast to $120 Billion

AMD has revised its projected total addressable market for servers in 2030, raising the estimate from $60 billion to $120 billion. The revision represents a doubling of the company's previous forecast and signals AMD's confidence that AI-driven workloads will fundamentally transform the data center market.

The updated projection likely factors in several trends:

• Explosive growth in AI inference workloads, which are expected to surpass training in total compute demand
• Enterprise adoption of AI accelerators moving from pilot projects to production deployments
• Increasing demand for sovereign AI infrastructure as governments worldwide invest in domestic compute capacity
• The emergence of new AI workloads, including agentic AI systems and real-time multimodal processing

AMD has been steadily gaining market share in the data center GPU segment with its MI300X and upcoming MI350 accelerators. While NVIDIA still dominates with an estimated 80-90% share of the AI accelerator market, AMD's aggressive pricing and growing software ecosystem through its ROCm platform have attracted major cloud customers including Microsoft and Meta.

The $120 billion forecast also implicitly accounts for the growing convergence of CPU and GPU workloads, where AMD's dual presence in both markets gives it a structural advantage over pure-play competitors.

SpaceX Proposes $55 Billion Terafab Project in Texas

SpaceX has proposed investing $55 billion to build a massive manufacturing facility in Texas, dubbed the 'Terafab' project. While details remain limited, the scale of the proposed investment would make it one of the largest single manufacturing commitments in American history.

The Terafab concept aligns with Elon Musk's broader vision of dramatically scaling Starship production to support both Starlink satellite constellation expansion and eventual Mars missions. For the AI industry, the connection is indirect but significant: SpaceX's Starlink network is increasingly viewed as critical infrastructure for edge AI deployment and global connectivity.

This announcement comes alongside China's Ministry of Industry and Information Technology approving the country's first commercial satellite IoT trial — a milestone that signals the growing convergence of space-based communications and AI-powered edge computing worldwide.

Samsung Exits Chinese Home Appliance Market Entirely

Samsung Electronics has announced a complete withdrawal from selling home appliances in the Chinese market, marking another chapter in the South Korean giant's strategic retreat from the world's second-largest economy. The decision covers all categories including refrigerators, washing machines, and air conditioners.

The move reflects intensifying competition from Chinese domestic brands like Haier, Midea, and Xiaomi, which have gained dominant market share through aggressive pricing and rapid integration of smart home and AI features. Samsung's premium positioning proved increasingly difficult to maintain as Chinese manufacturers closed the quality gap while offering significantly lower prices.

For the broader tech industry, Samsung's exit serves as a cautionary tale about the challenges foreign brands face in China. However, Samsung continues to maintain a significant presence in China through its semiconductor and display businesses, which remain critical links in the global AI supply chain.

What This Means for the AI Industry

The confluence of these developments paints a clear picture: the AI industry's infrastructure buildout is accelerating, not decelerating. AMD's decision to double its server market forecast, NVIDIA's strategic investment in optical connectivity, and the massive funding rounds for Chinese AI startups all point to sustained momentum.

For developers and businesses, several practical implications emerge. First, competition between Chinese and Western AI labs is intensifying, which benefits end users through lower costs and faster innovation. Second, infrastructure investments are shifting from pure compute to the full stack — networking, connectivity, power, and cooling are becoming equally important bottlenecks.

Third, valuations for top-tier AI companies continue to climb, suggesting that private market investors see a long Runway for growth despite concerns about a potential AI bubble.

Looking Ahead: The Race Intensifies

The next 6-12 months will be critical in determining whether these massive capital commitments translate into sustainable business models. DeepSeek's potential $45 billion valuation and Moonshot AI's $20 billion-plus mark set new benchmarks for Chinese AI companies, but both will face pressure to demonstrate revenue growth commensurate with these figures.

Meanwhile, the hardware ecosystem is evolving rapidly. Corning's deeper ties with NVIDIA, AMD's bullish market projections, and SpaceX's manufacturing ambitions all suggest that the physical infrastructure layer of AI will attract as much attention — and capital — as the software layer in the coming years.

Investors, builders, and policymakers should watch closely as these threads converge. The AI race is no longer just about building better models — it is about building the entire stack, from silicon to satellites, that makes those models useful at planetary scale.