# Сопутствующие статьи по теме Computing Power

Новостной центр HTX предлагает последние статьи и углубленный анализ по "Computing Power", охватывающие рыночные тренды, новости проектов, развитие технологий и политику регулирования в криптоиндустрии.

Why Does the Term 'Year of AI Computing Power Realization' Have Pitfalls? —Understanding the Four Hurdles from Policy Signals to Actual Orders in One Article

This article critiques the phrase "The First Year of AI Computing Power Cashing In," arguing it oversimplifies a complex, multi-stage process. It proposes a "Four Gates" framework to assess the true commercialization of domestic AI computing power (like Huawei's Ascend chips): 1. **Policy Procurement:** Widely open in 2026. Significant government funding and large bulk orders from tech giants like Alibaba and Tencent exist. However, purchasing hardware is not the same as deploying it for real use. 2. **Real Deployment:** A crack has opened. The key evidence is DeepSeek V4, a top-tier AI model fully migrating from NVIDIA's CUDA to domestic computing platforms. This proves the capability for real, high-level tasks, but widespread adoption beyond leading tech firms is still nascent. 3. **Mature Software Ecosystem:** A narrow crack has opened. While frameworks like Huawei's CANN are progressing, they lag far behind NVIDIA's vast, established CUDA ecosystem in terms of supported models and developer ease-of-use. Building this middle-to-downstream developer environment is estimated to need 1-2 more years. 4. **Scalable Replication:** Essentially closed. This final gate, where thousands of mid-sized enterprises across various industries can easily adopt the technology without major migration costs, is not expected before 2027-2028. The core risk is conflating these stages. While 2026 marks a real turning point in policy-driven procurement and proving technical viability (Gates 1 & 2), the phrase "cashing in" is premature for the full industry. True, large-scale value realization depends on the later, slower-to-open gates of software maturity and scalable replication to the broader market. DeepSeek V4's shift is identified as the most critical 2026 signal, changing the narrative from "can it work?" to "when will supply meet demand?"

marsbit05/08 11:34

Why Does the Term 'Year of AI Computing Power Realization' Have Pitfalls? —Understanding the Four Hurdles from Policy Signals to Actual Orders in One Article

marsbit05/08 11:34

Computing Power Constrained, Why Did DeepSeek-V4 Open Source?

DeepSeek-V4 has been released as a preview open-source model, featuring 1 million tokens of context length as a baseline capability—previously a premium feature locked behind enterprise paywalls by major overseas AI firms. The official announcement, however, openly acknowledges computational constraints, particularly limited service throughput for the high-end DeepSeek-V4-Pro version due to restricted high-end computing power. Rather than competing on pure scale, DeepSeek adopts a pragmatic approach that balances algorithmic innovation with hardware realities in China’s AI ecosystem. The V4-Pro model uses a highly sparse architecture with 1.6T total parameters but only activates 49B during inference. It performs strongly in agentic coding, knowledge-intensive tasks, and STEM reasoning, competing closely with top-tier closed models like Gemini Pro 3.1 and Claude Opus 4.6 in certain scenarios. A key strategic product is the Flash edition, with 284B total parameters but only 13B activated—making it cost-effective and accessible for mid- and low-tier hardware, including domestic AI chips from Huawei (Ascend), Cambricon, and Hygon. This design supports broader adoption across developers and SMEs while stimulating China's domestic semiconductor ecosystem. Despite facing talent outflow and intense competition in user traffic—with rivals like Doubao and Qianwen leading in monthly active users—DeepSeek has maintained technical momentum. The release also comes amid reports of a new funding round targeting a valuation exceeding $10 billion, potentially setting a new record in China’s LLM sector. Ultimately, DeepSeek-V4 represents a shift toward open yet realistic infrastructure development in the constrained compute landscape of Chinese AI, emphasizing engineering efficiency and domestic hardware compatibility over pure model scale.

marsbit04/26 00:27

Computing Power Constrained, Why Did DeepSeek-V4 Open Source?

marsbit04/26 00:27

Jensen Huang Publicly Challenges Google and Amazon, Is the Chip Business Entirely Sustained by Anthropic?

In a candid interview, Nvidia CEO Jensen Huang challenged competitors like Google and Amazon, admitted past strategic errors, and criticized U.S. export controls on AI chips. He framed Nvidia’s role as turning “electricity in, tokens out,” emphasizing the complexity and value of AI inference. Huang dismissed rival custom chips like Google’s TPU and Amazon’s Trainium as inflexible and niche, claiming their growth relies heavily on clients like Anthropic. He also acknowledged missing early investment opportunities in OpenAI and Anthropic. On China, Huang warned that export restrictions risk pushing the country toward self-sufficiency and could cost U.S. leadership in AI. Finally, he explained Nvidia’s acquisition of Groq as a move to serve premium, low-latency token markets. Throughout, Huang emphasized ecosystem trust and Nvidia’s central role in global AI infrastructure.

marsbit04/21 07:00

Jensen Huang Publicly Challenges Google and Amazon, Is the Chip Business Entirely Sustained by Anthropic?

marsbit04/21 07:00

From "Silicon Valley's Sacred Shoes" to "GPU Computing Power": The Absurdity and Logic Behind Allbirds Renaming to NewBird AI

From "Silicon Valley's Favorite Shoe" to "GPU Computing Power": The Absurdity and Logic Behind Allbirds' Rebranding to NewBird AI On April 15, Allbirds, the maker of merino wool running shoes, announced a radical pivot from footwear to AI compute, rebranding as "NewBird AI." The move triggered a 582% surge in its stock price the same day. This followed the sale of its shoe business for $39 million—a fraction of its $4 billion IPO valuation in 2021. Allbirds rose to fame in 2016 with its comfortable, eco-friendly minimalist shoes, becoming a status symbol in tech circles. But after rapid expansion and failed attempts to attract Gen Z, revenue declined, losses mounted, and its value plummeted. By early 2026, all its U.S. stores had closed. Now, under CEO Joe Vernachio, the company is attempting a reboot. It secured $50 million in convertible notes from an undisclosed investor to purchase high-performance GPUs and offer "GPU-as-a-service" to AI developers. The company cites real market shortages in compute capacity, but questions remain about how a $50 million entry can compete in a capital-intensive industry dominated by giants like NVIDIA and CoreWeave. The move echoes past market frenzies, such as Long Island Iced Tea’s pivot to blockchain in 2017—a hype-driven strategy that ended in delisting and SEC action. While AI compute demand is real, NewBird AI’s operational capacity and execution plan remain unproven. The timing is suggestive: the stock soared based on a narrative, before any shareholder vote or operational results. The company plans a special dividend in Q3, raising questions about who benefits from the short-term market enthusiasm. NewBird AI exemplifies a broader trend: companies with broken business models turning to AI for revival. Whether this is a legitimate transformation or a market play remains to be seen.

marsbit04/16 04:52

From "Silicon Valley's Sacred Shoes" to "GPU Computing Power": The Absurdity and Logic Behind Allbirds Renaming to NewBird AI

marsbit04/16 04:52

Monthly Salary of 30K: Guarding DeepSeek's Data Center on the Inner Mongolia Grassland

Monthly salary of 30,000 RMB to operate and maintain DeepSeek’s data center in the grasslands of Inner Mongolia. DeepSeek is hiring data center operation and delivery managers instead of AI researchers or algorithm engineers, signaling a strategic shift toward physical compute infrastructure. The company, known for its cost-efficient AI models, is now expanding its efficiency-first philosophy from software to hardware—optimizing not just training and inference efficiency, but also the total cost of ownership (TCO), including cooling and electricity. The data center is located in Ulanqab, a key hub in China’s "East Data West Computing" project. With a cool climate enabling natural server cooling and a PUE under 1.2, plus over 65% clean energy usage, the region is ideal for large-scale computing. It already hosts major cloud providers like Huawei and Alibaba, reducing risks for DeepSeek’s expansion. This move comes ahead of DeepSeek’s rumored V4 model launch, expected in late April, which may feature trillion-scale parameters and a 1M+ context window. As algorithmic gains slow, physical compute power becomes a critical moat. The company has also recently introduced "Fast Mode" and "Expert Mode" on its web platform, hinting at more advanced capabilities ahead.

marsbit04/14 08:05

Monthly Salary of 30K: Guarding DeepSeek's Data Center on the Inner Mongolia Grassland

marsbit04/14 08:05

Giants Collectively Raise Prices, Is the AI Price Hike Wave Coming? Can We Still Afford Lobster Employees?

Major AI companies, including Alibaba Cloud, Baidu Intelligent Cloud, Tencent Cloud, and Zhipu, have recently announced significant price increases for AI computing and storage services, with hikes ranging from 5% to over 460% in some models. This trend follows similar moves by global giants like Amazon AWS and Google Cloud earlier this year. The price surge is driven by explosive demand for computing power, fueled by the rapid adoption of AI agents like OpenClaw (referred to as "Lobster" in the article), which consume tokens at rates dozens or even hundreds of times higher than traditional AI applications. This has created a severe supply-demand imbalance. Additionally, shortages in high-end hardware—such as AI chips and high-bandwidth memory (HBM)—have constrained computing capacity and raised operational costs. The industry is shifting away from loss-leading pricing strategies toward value-based models, prioritizing sustainable development over market-share competition. A new "token economy" is emerging, where pricing is increasingly based on token usage, complexity, and speed rather than flat fees. This reflects AI computing's evolution from a generic service to a specialized, high-value resource. Some companies are even considering token allowances as part of employee benefits, highlighting its growing role as both a production tool and a cost factor. The article concludes by questioning whether AI services will remain affordable as compute costs continue to rise.

marsbit04/13 04:20

Giants Collectively Raise Prices, Is the AI Price Hike Wave Coming? Can We Still Afford Lobster Employees?

marsbit04/13 04:20

All the AI News You Need to Know Is Here: Rhythm Officially Launches AI News Feed

By 2026, the AI industry has seen a significant increase in news volume compared to previous years. While earlier concerns focused on model upgrades and efficiency, this year, products like OpenClaw have captured widespread attention even without major model improvements, highlighting AI’s growing impact on daily life. This surge in activity has also intensified information overload and competition. With OpenClaw accelerating developments, both model providers and companies without in-house models are racing to attract users, creating a crowded and fast-moving landscape.

marsbit03/11 07:23

All the AI News You Need to Know Is Here: Rhythm Officially Launches AI News Feed

marsbit03/11 07:23

The Great Migration of Mining Companies: Some Already Hold $12.8 Billion in AI Orders

Bitcoin mining companies are undergoing a massive structural shift toward AI and high-performance computing (HPC), driven by compressed mining profits and the lure of massive AI infrastructure contracts. With Bitcoin’s mining difficulty rising and profitability declining, miners are leveraging their existing assets—power contracts, land, cooling systems, and data centers—to pivot into AI compute provisioning. Key players like TeraWulf have secured over $12.8 billion in HPC contracts, while IrisEnergy holds a $9.7 billion deal with Microsoft. Other firms, including Hut8, CipherDigital, and CoreScientific, have also announced multi-billion dollar AI agreements. Some, like BitDigital, are exiting Bitcoin mining entirely to focus on AI and Ethereum infrastructure. Financial reports show a clear trend: AI/HPC revenue is becoming a major income source, sometimes surpassing mining. Capital markets are selective, favoring companies with strong execution, resource advantages, and strategic positioning within the AI supply chain. The migration is accelerating—miners must adapt to the new reality or risk obsolescence.

marsbit03/06 05:58

The Great Migration of Mining Companies: Some Already Hold $12.8 Billion in AI Orders

marsbit03/06 05:58

The Great Migration of Mining Companies: Some Already Hold $12.8 Billion in AI Orders

Driven by rising Bitcoin mining difficulty, shrinking profits, and extreme capital pursuit of AI, major crypto mining firms are collectively pivoting towards AI and high-performance computing (HPC). They leverage existing assets—power, land, cooling systems, and data centers—to secure massive AI compute orders. Key catalysts include Bitcoin’s hashprice dropping to multi-year lows, pushing firms like Bitdeer to sell all BTC holdings and NFN8 into bankruptcy. Conversely, companies are securing enormous contracts: TeraWulf holds a $12.8B HPC deal, IrisEnergy a $9.7B order with Microsoft, and Hut8 a $7B agreement. Others like CipherDigital ($5.5B with AWS) and CoreScientific ($10B+ with CoreWeave) are rebranding and expanding data centers. Q1 2026 earnings show AI/HPC revenue now constitutes 40%-60% for some. While firms like Marathon Digital remain committed to Bitcoin long-term, others like BitDigital have fully exited BTC mining for ETH and AI. Market sentiment is volatile but selective, with capital backing execution-ready firms with strong power contracts and scalability. The migration is structural: adapt to AI or risk obsolescence.

Odaily星球日报03/06 05:48

The Great Migration of Mining Companies: Some Already Hold $12.8 Billion in AI Orders

Odaily星球日报03/06 05:48

From Spring Festival Gala Robots to the Computing Power Energy War: Why Does China Hold the 'Trump Card' in the AI Era?

China's 2026 Spring Festival Gala showcased a breakthrough in embodied AI, featuring robots from companies like Magic Atom, Unitree, and Galaxy General performing complex tasks such as dancing, martial arts, and comedy. This demonstrated China's advanced progress in robotics and AI physical integration. Meanwhile, the U.S. faces an escalating energy crisis, with electricity prices rising 36% by early 2026. Training AI models like GPT-4 consumes power equivalent to 100,000 households annually, and U.S. data centers are projected to use 600,000 GWh by 2028. Aging infrastructure, fragmented grids, and lengthy approval processes for new transmission lines exacerbate the problem. In contrast, China has built a strategic advantage through decades of infrastructure investment. It operates 45 ultra-high-voltage (UHV) power transmission projects, spanning 40,000 kilometers, efficiently delivering clean energy from the west to eastern data centers. Renewable energy accounts for over 60% of China’s power capacity, with 40% of electricity coming from green sources. China also dominates transformer production, holding 60% of global capacity. While the U.S. excels in AI algorithms, China’s robust energy infrastructure—UHV grids, renewable energy, and manufacturing capacity—provides a foundational edge in the AI era, turning energy into a critical competitive asset.

marsbit02/22 02:27

From Spring Festival Gala Robots to the Computing Power Energy War: Why Does China Hold the 'Trump Card' in the AI Era?

marsbit02/22 02:27

活动图片

Популярные категорииright-arrow

Популярные тегиright-arrow

Ethereum

Technology Trends