The article outlines the trend of AI compute center cooling shifting from air cooling to liquid cooling. Nvidia's Vera Rubin racks have eliminated fans, exceed 190 kilowatts of power per rack, and adopt 100% full liquid cooling, while air cooling can generally only handle about 20 to 30 kilowatts of heat dissipation per rack. The article points out that cold plates, CDUs, and quick connectors together account for about 85% of the cost of a liquid cooling system, that the high-end market has long been controlled by Taiwanese and European/American manufacturers, and it summarizes three ways Chinese manufacturers are entering: cross-sector restructuring, single-point efficiency, and trading time for market.
The article outlines the trend of AI compute center cooling shifting from air cooling to liquid cooling. Nvidia's Vera Rubin racks have eliminated fans, exceed 190 kilowatts of power per rack, and adopt 100% full liquid cooling, while air cooling can generally only handle about 20 to 30 kilowatts of heat dissipation per rack. The article points out that cold plates, CDUs, and quick connectors together account for about 85% of the cost of a liquid cooling system, that the high-end market has long
Xia Fanjie, a strategist at CSC Financial, said that as of October 24, the 848 A-share companies that had disclosed their third-quarter reports saw revenue and net profit attributable to parent companies grow 13.6% and 19.0% year on year, respectively, with earnings highlights concentrated in three major areas: AI computing power, some resource commodities and high-end manufacturing. The electronics sector's third-quarter net profit attributable to parent companies surged 63.6%, minor metals and industrial metals net profit attributable to parent companies rose 96.7% and 101.2%, respectively, and machinery and equipment net profit attributable to parent companies grew 41.9%. He recommended a "dividend defense, performance-driven" allocation approach, with the offensive side focused on optical communications, high-end PCBs, AI power supply supporting components and industrial metals.
Xia Fanjie, a strategist at CSC Financial, said that as of October 24, the 848 A-share companies that had disclosed their third-quarter reports saw revenue and net profit attributable to parent companies grow 13.6% and 19.0% year on year, respectively, with earnings highlights concentrated in three major areas: AI computing power, some resource commodities and high-end manufacturing. The electronics sector's third-quarter net profit attributable to parent companies surged 63.6%, minor metals and
At CASPA's 35th Annual Conference, SemiAnalysis founder Dylan Patel and Bernstein analyst Stacy Rasgon discussed the outlook for AI infrastructure investment, arguing that with the U.S. 10-year Treasury yield breaking above 5% and unsecured bonds from emerging cloud service providers yielding close to 15%, financing capacity is becoming a key constraint on expansion. The two judged that substrates and memory are bottlenecks harder to scale than power, that HBM requires roughly 3 to 4 times as many wafers as DDR5, and that if new capacity is concentrated in releases after 2028, memory prices and profit margins could come under pressure. Nvidia still supports its pricing through performance gains, but Google TPU, Amazon Trainium, and the custom chips of OpenAI and Anthropic are eroding its share and bargaining power.
At CASPA's 35th Annual Conference, SemiAnalysis founder Dylan Patel and Bernstein analyst Stacy Rasgon discussed the outlook for AI infrastructure investment, arguing that with the U.S. 10-year Treasury yield breaking above 5% and unsecured bonds from emerging cloud service providers yielding close to 15%, financing capacity is becoming a key constraint on expansion. The two judged that substrates and memory are bottlenecks harder to scale than power, that HBM requires roughly 3 to 4 times as ma
Meta's AI assistant Muse has accumulated 10.9 million downloads since its September launch, of which 10.2 million came from the United States, making it one of the mobile apps to surpass 10 million U.S. downloads within 40 days of launch. DeepSeek open-sourced its inference framework DeepInfer and disclosed that a single cluster serves a throughput of 155 trillion tokens per day, with revenue of $38 per card per hour. Anthropic launched Claude Duet, a personal agent based on Claude Opus 5.5, opening research preview waitlist applications to Claude Pro and Max subscribers in the United States.
Meta's AI assistant Muse has accumulated 10.9 million downloads since its September launch, of which 10.2 million came from the United States, making it one of the mobile apps to surpass 10 million U.S. downloads within 40 days of launch. DeepSeek open-sourced its inference framework DeepInfer and disclosed that a single cluster serves a throughput of 155 trillion tokens per day, with revenue of $38 per card per hour. Anthropic launched Claude Duet, a personal agent based on Claude Opus 5.5, ope
In the first three quarters, ChangXin Technology's revenue was approximately 265 billion to 270 billion yuan, with net profit attributable to parent company shareholders of approximately 131 billion to 132.5 billion yuan; in the same period last year, it posted a loss of 5.28 billion yuan. However, third-quarter revenue increased by 15.2 billion to 20.2 billion yuan quarter-on-quarter, while net profit attributable to parent company shareholders rose by only 550 million to 2.05 billion yuan, and the net profit margin attributable to parent company shareholders fell from 53.1% to about 46%. The quarter-on-quarter increase in DRAM contract prices has narrowed from 90% to 95% in the first quarter to 13% to 18% in the third quarter. The third-quarter report needs to focus on gross margin, income tax expense, and minority interests to determine whether this is an accounting issue or an operating inflection point.
In the first three quarters, ChangXin Technology's revenue was approximately 265 billion to 270 billion yuan, with net profit attributable to parent company shareholders of approximately 131 billion to 132.5 billion yuan; in the same period last year, it posted a loss of 5.28 billion yuan. However, third-quarter revenue increased by 15.2 billion to 20.2 billion yuan quarter-on-quarter, while net profit attributable to parent company shareholders rose by only 550 million to 2.05 billion yuan, and
Amazon CEO Andy Jassy said in an interview on the a16z podcast that the company's capital expenditure is expected to reach $220 billion in 2026, but compute capacity is already sold out for 2026, essentially fully booked for 2027, and a large share of 2028 quotas has been locked in. He described enterprise AI demand as dumbbell-shaped: a large number of production systems in the middle have yet to be deeply integrated because inference costs are too high, and costs need to fall by another order of magnitude before an explosion in adoption occurs. Jassy also noted that 99% of AWS's top 1,000 customers use Graviton, which offers about 40% better price-performance than x86, while Trainium has a 30% to 40% advantage over comparable hardware.
Amazon CEO Andy Jassy said in an interview on the a16z podcast that the company's capital expenditure is expected to reach $220 billion in 2026, but compute capacity is already sold out for 2026, essentially fully booked for 2027, and a large share of 2028 quotas has been locked in. He described enterprise AI demand as dumbbell-shaped: a large number of production systems in the middle have yet to be deeply integrated because inference costs are too high, and costs need to fall by another order
CNBC reports that Nvidia's role as an AI infrastructure financier is facing its first major stress test, with neoclouds such as CoreWeave, Nebius and IREN falling at least 30% this month and CoreWeave's credit default swap spreads widening past 11%. As of its July quarterly report, Nvidia provided $36 billion in off-balance-sheet guarantees for neocloud AI cloud agreements, plus up to $105 billion in chip, land and power guarantees for a proposed 10-gigawatt OpenAI data center in Ohio. Jensen Huang told CNBC this week during GTC in Berlin that neocloud demand is absurdly high and what's missing is the working capital to drive growth.
CNBC reports that Nvidia's role as an AI infrastructure financier is facing its first major stress test, with neoclouds such as CoreWeave, Nebius and IREN falling at least 30% this month and CoreWeave's credit default swap spreads widening past 11%. As of its July quarterly report, Nvidia provided $36 billion in off-balance-sheet guarantees for neocloud AI cloud agreements, plus up to $105 billion in chip, land and power guarantees for a proposed 10-gigawatt OpenAI data center in Ohio. Jensen Hu
The Sunday Times 2026 UK Rich List shows that XTX founder Gerko ranks eighth with a net worth of £16.006 billion, an increase of £7.261 billion in a year. XTX has about 300 employees; in 2025, its three UK entities had total revenue of about £3.93 billion and daily trading volume of about $250 billion. The company stores 400PB of data on a supercomputer in Iceland, has spent more than £150 million on 25,000 AI chips, uses machine learning to predict prices of more than 53,000 financial instruments, and plans to invest more than €1 billion to build five data centers in Finland.
The Sunday Times 2026 UK Rich List shows that XTX founder Gerko ranks eighth with a net worth of £16.006 billion, an increase of £7.261 billion in a year. XTX has about 300 employees; in 2025, its three UK entities had total revenue of about £3.93 billion and daily trading volume of about $250 billion. The company stores 400PB of data on a supercomputer in Iceland, has spent more than £150 million on 25,000 AI chips, uses machine learning to predict prices of more than 53,000 financial instrumen
AI startup Flapping Airplanes, founded about a year ago, has completed a $1 billion Series A funding round at a valuation of $6 billion, with participation from Nvidia, Sequoia Capital, GV and others; its seed round valued it at $1.5 billion. The company independently develops software that drives Nvidia GPUs, and its founder says its "giant kernel framework" can improve chip efficiency and lower the cost of AI training and operation. It currently has no commercial product, and independent tests, customer cases and a product timeline have not been made public.
AI startup Flapping Airplanes, founded about a year ago, has completed a $1 billion Series A funding round at a valuation of $6 billion, with participation from Nvidia, Sequoia Capital, GV and others; its seed round valued it at $1.5 billion. The company independently develops software that drives Nvidia GPUs, and its founder says its "giant kernel framework" can improve chip efficiency and lower the cost of AI training and operation. It currently has no commercial product, and independent tests
The union at Micron's Taoyuan plant in Taiwan, China, began a 24-hour warning strike at 9 a.m. on October 22, carried out by 11 union officials at the factory gate; production lines were not directly halted, and the core demand is to institutionalize profit sharing. The union is dissatisfied with the production-line employees' "two days on, two days off" schedule, 12-hour shifts per shift, and the work pattern of coordinating with the U.S. headquarters for early-morning and midnight meetings. Micron's fiscal 2026 revenue was $133.188 billion and net profit was $84.969 billion, both setting historical records.
The union at Micron's Taoyuan plant in Taiwan, China, began a 24-hour warning strike at 9 a.m. on October 22, carried out by 11 union officials at the factory gate; production lines were not directly halted, and the core demand is to institutionalize profit sharing. The union is dissatisfied with the production-line employees' "two days on, two days off" schedule, 12-hour shifts per shift, and the work pattern of coordinating with the U.S. headquarters for early-morning and midnight meetings. Mi
Danish investor Lars Tvede, author of "The Economic Cycle That Can't Be Escaped," believes the AI bubble will come sooner or later, but we are not currently in a bubble: Nvidia's forward P/E ratio is about 18 times, compared with about 200 times when Cisco peaked in 2000. He points out that the bottlenecks constraining AI development lie in upstream chip shortages and power grid connection queues, and believes that chipmakers' current prices are "absurdly low" and worth watching over the next three to five years.
Danish investor Lars Tvede, author of "The Economic Cycle That Can't Be Escaped," believes the AI bubble will come sooner or later, but we are not currently in a bubble: Nvidia's forward P/E ratio is about 18 times, compared with about 200 times when Cisco peaked in 2000. He points out that the bottlenecks constraining AI development lie in upstream chip shortages and power grid connection queues, and believes that chipmakers' current prices are "absurdly low" and worth watching over the next th
Warp Sciences, an AI scientific research startup founded by former OpenAI Chief Product Officer Kevin Weil, has emerged from stealth, completing a $215 million Series A funding round at a post-money valuation of $865 million, led by Lightspeed Venture Partners. The company, founded in San Francisco with a team of about 20 people, plans to use AI, robotics, and automated laboratories to help pharmaceutical companies screen molecular variants in drug development, without developing drugs itself. The approach is still at an early stage, and whether it can run stably and improve experimental efficiency remains to be verified.
Warp Sciences, an AI scientific research startup founded by former OpenAI Chief Product Officer Kevin Weil, has emerged from stealth, completing a $215 million Series A funding round at a post-money valuation of $865 million, led by Lightspeed Venture Partners. The company, founded in San Francisco with a team of about 20 people, plans to use AI, robotics, and automated laboratories to help pharmaceutical companies screen molecular variants in drug development, without developing drugs itself. T
At SpaceX's Q3 2026 earnings call, Musk said superintelligent compute capacity will face a long-term shortage, with power, cooling and infrastructure constraining supply more than chips. SpaceX's superintelligence business revenue for the quarter was $7.56 billion, surpassing the connectivity business led by Starlink; compute scale rose from 0.6 gigawatts a year earlier to 1.7 gigawatts, with plans to reach 13 gigawatts by the end of 2027. Musk said paid weekly active users of Grok Bot reached 800,000 in early October, and the company also plans to build its own gas turbine blade casting and chip factories to ease bottlenecks.
At SpaceX's Q3 2026 earnings call, Musk said superintelligent compute capacity will face a long-term shortage, with power, cooling and infrastructure constraining supply more than chips. SpaceX's superintelligence business revenue for the quarter was $7.56 billion, surpassing the connectivity business led by Starlink; compute scale rose from 0.6 gigawatts a year earlier to 1.7 gigawatts, with plans to reach 13 gigawatts by the end of 2027. Musk said paid weekly active users of Grok Bot reached 8
At Tesla's earnings call, Musk said humanoid robots will become the world's largest industry, and he expects that after solving robot AI, electromechanical design, reliability and lifespan issues by the end of next year, Optimus can be delivered to companies other than Tesla; Tesla will also build an Optimus Academy, planning to mass-produce 20,000 units for autonomous training. The domestic construction robot market reached 18.6 billion yuan in 2025 and is expected to exceed 80 billion yuan by 2030. DeepSeek is reported to have completed about 100 billion yuan in financing, with CATL and Tencent participating.
At Tesla's earnings call, Musk said humanoid robots will become the world's largest industry, and he expects that after solving robot AI, electromechanical design, reliability and lifespan issues by the end of next year, Optimus can be delivered to companies other than Tesla; Tesla will also build an Optimus Academy, planning to mass-produce 20,000 units for autonomous training. The domestic construction robot market reached 18.6 billion yuan in 2025 and is expected to exceed 80 billion yuan by
According to sources, Tesla and NVIDIA plan to jointly invest approximately 3 trillion KRW in Samsung Electro-Mechanics, earmarked for expanding FC-BGA capacity for AI semiconductor packaging substrates, with Tesla contributing approximately 1 trillion KRW and NVIDIA approximately 2 trillion KRW. Tesla's dedicated line will be located at the Sejong plant, producing AI semiconductor substrates for autonomous driving electric vehicles and the humanoid robot Optimus, with a supply contract period from 2029 to 2033; NVIDIA's dedicated line targets data centers, producing AI computing chipsets and network substrates, with the collaboration disclosed on October 21 at GTC 2026 in Germany. Samsung Electro-Mechanics had just announced on September 28 a packaging substrate capacity expansion plan totaling 6.78 trillion KRW, with mass production targeted for September 2028.
According to sources, Tesla and NVIDIA plan to jointly invest approximately 3 trillion KRW in Samsung Electro-Mechanics, earmarked for expanding FC-BGA capacity for AI semiconductor packaging substrates, with Tesla contributing approximately 1 trillion KRW and NVIDIA approximately 2 trillion KRW. Tesla's dedicated line will be located at the Sejong plant, producing AI semiconductor substrates for autonomous driving electric vehicles and the humanoid robot Optimus, with a supply contract period f
The Shanghai Municipal Commission of Economy and Informatization released the "Shanghai Action Plan for Accelerating the High-Quality Development of AI-Empowered Industrial Services (2027-2028)", proposing that by 2028 the penetration rate of AI applications in industrial services reach over 85%, striving to cultivate 10 digitally native unicorn enterprises. The plan provides funding support ranging from up to 5 million to 20 million yuan for stages including R&D and design, supply chain services, flexible manufacturing, embedded quality inspection, and predictive maintenance, and clearly identifies research directions such as industrial agent task planning and swarm collaboration, device-cloud collaborative inference and lightweight vertical small models, and industry high-quality corpus libraries.
The Shanghai Municipal Commission of Economy and Informatization released the "Shanghai Action Plan for Accelerating the High-Quality Development of AI-Empowered Industrial Services (2027-2028)", proposing that by 2028 the penetration rate of AI applications in industrial services reach over 85%, striving to cultivate 10 digitally native unicorn enterprises. The plan provides funding support ranging from up to 5 million to 20 million yuan for stages including R&D and design, supply chain service
At GTC Berlin, Nvidia launched the Open Agent Safety Platform, adding hardware-level controls around the open-source runtime OpenShell; a BlueField-4 DPU works with Sentry software to independently monitor agent behavior and, when it crosses boundaries, isolate and stop it within milliseconds. OpenShell supports third-party agent frameworks such as Anthropic's Claude Code and OpenAI's Codex, and Nvidia says more than 100 organizations already use the platform's technology. Nvidia also introduced the Agent Knowledge Foundation and the Auto-Ontology library, designed to let agents access structured and unstructured data without enterprise data leaving local premises, and expanded its Nemotron 3.5 Super VL, Nemotron 3.5 Bolt 74B-A9B and Kumo series models.
At GTC Berlin, Nvidia launched the Open Agent Safety Platform, adding hardware-level controls around the open-source runtime OpenShell; a BlueField-4 DPU works with Sentry software to independently monitor agent behavior and, when it crosses boundaries, isolate and stop it within milliseconds. OpenShell supports third-party agent frameworks such as Anthropic's Claude Code and OpenAI's Codex, and Nvidia says more than 100 organizations already use the platform's technology. Nvidia also introduced
DeepSeek has open-sourced its internal inference framework Deepinfer on GitHub, with the goal of reducing the compute and energy costs of deploying models on your own. During the inference stage, the framework distributes load more evenly across available processor cores, uses speculative decoding to speed up responses, and improves the structure of results before output; it currently supports only DeepSeek-V4.1-Flash, with more of its own models to be opened up later. On the hardware side, it requires Nvidia GPUs with the SM100 architecture or Huawei Ascend 950 NPUs, so it is not aimed at individual users.
DeepSeek has open-sourced its internal inference framework Deepinfer on GitHub, with the goal of reducing the compute and energy costs of deploying models on your own. During the inference stage, the framework distributes load more evenly across available processor cores, uses speculative decoding to speed up responses, and improves the structure of results before output; it currently supports only DeepSeek-V4.1-Flash, with more of its own models to be opened up later. On the hardware side, it r
At the Guoxin Exhibition, Dongfang Suanxin announced that its 128-card DF1000 training cluster is running stably and that it has achieved large-scale training cluster delivery, with plans to achieve thousand-card-level cluster delivery by the end of the year. The cluster trained continuously for 72 hours without interruption and has carried training tasks for models such as Llama2-70B and Step 130B; the hundred-billion-parameter-level Step 130B completed training adaptation in only 7 days. The DF1000 adopts 3D-stacked near-memory computing, with per-chip memory access bandwidth of 6.4TB/s, and has completed a closed loop across the domestic industrial chain from chip design and 14nm wafer manufacturing to 3D hybrid bonding packaging.
At the Guoxin Exhibition, Dongfang Suanxin announced that its 128-card DF1000 training cluster is running stably and that it has achieved large-scale training cluster delivery, with plans to achieve thousand-card-level cluster delivery by the end of the year. The cluster trained continuously for 72 hours without interruption and has carried training tasks for models such as Llama2-70B and Step 130B; the hundred-billion-parameter-level Step 130B completed training adaptation in only 7 days. The D