Jihan Wu’s Vision for AI Computing: Why the Push Toward 10,000, 100,000, and Even 1 Million GPUs Matters

Who Is Jihan Wu and Why His GPU Comments Matter

Jihan Wu is one of the most influential figures in the cryptocurrency mining industry. He co-founded Bitmain, which became a dominant manufacturer of Bitcoin mining machines during the global crypto mining boom.

In recent years, Wu has increasingly shifted attention toward AI infrastructure through Bitdeer. His latest comments about scaling GPU deployment from 10,000 to 100,000 and eventually 1 million units highlight how rapidly the artificial intelligence industry is evolving.

The statement is important because GPUs have become one of the most valuable strategic resources in the AI economy. Large AI systems require enormous computing power, and the companies that control GPU infrastructure may gain significant advantages in cloud computing, AI model training, and enterprise services.

Wu’s comments also reflect a larger industry trend: crypto mining companies are increasingly transforming into AI computing infrastructure providers.

Why AI Companies Are Racing to Deploy More GPUs

Modern AI models require vast amounts of parallel computing power. Graphics Processing Units (GPUs), originally designed for gaming and graphics rendering, are now widely used for machine learning and AI model training.

AI systems such as large language models, image generators, and autonomous agents rely heavily on GPU clusters because they can process enormous datasets simultaneously.

The scale of deployment is growing rapidly:

• Smaller AI startups may operate with several thousand GPUs
• Major cloud providers now deploy tens of thousands of GPUs
• Future hyperscale AI data centers could potentially require hundreds of thousands or even millions of GPUs

The reason is simple: larger AI models require exponentially more computational resources.

Training advanced AI models can consume massive amounts of electricity, cooling capacity, networking bandwidth, and semiconductor supply. As competition intensifies, companies are racing to secure GPU access before shortages become even more severe.

Industry reports and public filings show that firms connected to AI infrastructure are expanding aggressively into high-performance computing (HPC) and AI cloud services.

From Bitcoin Mining to AI Infrastructure

One of the most interesting developments in recent years is the convergence between crypto mining infrastructure and AI infrastructure.

Bitcoin mining companies already possess several assets that are useful for AI operations:

• Large-scale data centers
• Access to electricity
• Cooling systems
• Experience operating high-density computing hardware
• Relationships with energy providers

As a result, many mining firms are exploring AI hosting and GPU cloud services as new revenue streams.

Bitdeer has publicly discussed expanding into AI cloud services powered by NVIDIA hardware and high-performance computing systems. SEC filings and industry reports show that the company has invested in AI-related infrastructure while continuing mining operations.

This transition reflects changing economics within the crypto mining industry. Bitcoin mining revenues can fluctuate significantly depending on token prices, mining difficulty, and energy costs. AI infrastructure, by contrast, is viewed by some companies as a potentially longer-term growth market.

However, the transition is not simple. AI infrastructure requires different customer relationships, software ecosystems, networking architecture, and service-level reliability compared with crypto mining.

What Does 1 Million GPUs Actually Mean?

Wu’s mention of “1 million GPUs” may sound extreme today, but it reflects how some industry leaders are thinking about long-term AI demand.

To understand the scale:

• A single advanced GPU can cost tens of thousands of dollars
• A 100,000-GPU deployment could represent billions of dollars in hardware investment
• A 1 million GPU ecosystem would require enormous energy infrastructure and supply chain coordination

The operational complexity would also be massive.

Large AI clusters require:

• Advanced liquid cooling systems
• Ultra-fast networking
• Stable power grids
• AI-optimized data center architecture
• Massive semiconductor manufacturing capacity

Electricity demand could become one of the biggest limiting factors. AI data centers already consume large amounts of power, and future hyperscale deployments may place additional pressure on regional power infrastructure.

This is why energy access has become increasingly important in the AI race.

The Challenges Behind Massive GPU Expansion

Although AI demand remains strong, large-scale GPU deployment also comes with significant risks and uncertainties.

Supply Chain Constraints

Advanced AI chips remain difficult to manufacture because they depend on cutting-edge semiconductor fabrication processes. Supply shortages could continue if demand grows faster than production capacity.

High Capital Costs

Building AI data centers is extremely expensive. Companies may need billions of dollars in financing for hardware, land, networking, cooling, and electricity infrastructure.

Reports connected to Bitdeer suggest that expansion into AI infrastructure has involved substantial financing activity and debt issuance.

Energy Consumption

Power availability may become one of the most important bottlenecks for AI growth. Governments and regulators may also increase scrutiny of energy-intensive computing operations.

Rapid Technology Changes

AI hardware evolves quickly. GPUs purchased today could become less competitive within a few years as newer architectures emerge.

This creates risks for companies investing heavily in current-generation hardware.

How AI Demand Is Changing the Data Center Industry

The global data center industry is already adapting to the AI boom.

Traditionally, data centers focused on cloud storage, web hosting, and enterprise software. AI workloads are different because they require:

• Much higher power density
• Faster networking
• Advanced cooling technologies
• Specialized AI accelerators

As AI adoption expands, many infrastructure providers are redesigning facilities specifically for AI computing.

Some analysts believe AI could become one of the largest drivers of global data center investment during the next decade. Industry discussions increasingly focus on “AI factories” — large facilities dedicated entirely to AI training and inference workloads.

This trend may also reshape semiconductor manufacturing, electricity markets, and global technology competition.

Risks Investors and the Industry Should Watch

While enthusiasm around AI infrastructure remains strong, investors should approach the sector carefully.

Several important risks remain:

• AI demand growth may slow unexpectedly
• GPU oversupply could eventually pressure margins
• Rising electricity costs could reduce profitability
• Regulatory scrutiny may increase
• Heavy debt financing could create financial stress during market downturns

The AI infrastructure industry is still developing rapidly, and long-term winners have not yet been fully determined.

Investors should also remember that infrastructure expansion does not automatically guarantee sustainable profits. Many companies may compete aggressively for market share, which could pressure returns over time.

The Long-Term Outlook for AI Compute Infrastructure

Jihan Wu’s comments about scaling GPU deployment toward 10,000, 100,000, and eventually 1 million units reflect the growing belief that AI computing demand may continue expanding for years.

Whether the industry ultimately reaches those numbers remains uncertain. However, the broader trend is clear: computing power is becoming one of the most strategically important resources in the global technology sector.

The next phase of AI competition may depend not only on software and algorithms, but also on who can secure the largest and most efficient computing infrastructure.

For companies like Bitdeer and other former crypto mining firms, the shift toward AI infrastructure could represent both an opportunity and a major financial gamble.

As the AI race accelerates, the balance between technological ambition, capital investment, and sustainable economics will likely determine which companies succeed in the long run.