IBM's New AI Brain-Inspired Chip is 22 Times Faster than Commercial Chips

IBM researchers in San Jose, California, have developed a brain-like computing chip named NorthPole, which enables faster and more energy-efficient AI operations. The NorthPole chip incorporates a core-with-memory design, reducing the need for frequent external memory access, thereby accelerating AI tasks such as image recognition compared to existing chips, while significantly lowering energy consumption. Researchers report that NorthPole delivers 22 times faster computational speeds in image recognition tests and improves energy efficiency by 25 times over commercial chips. Inspired by the structure of the human brain, NorthPole is expected to be applied in speed-sensitive AI scenarios such as autonomous driving, advancing the development of AI technology.

Broadcom is raising over $50 billion for an AI chip project tailored for OpenAI, with Apollo and Blackstone joining in. Oracle is also seeking funds to buy chips

According to foreign media, Broadcom is arranging financing of over $50 billion for a custom AI chip project in collaboration with OpenAI. Apollo Global Management and Blackstone Group are negotiating to participate in this large loan. At the same time, Oracle is discussing with Apollo and Goldman Sachs to raise funds for chip purchases. This financing chain connects Broadcom, OpenAI, Oracle, and several financial institutions, highlighting the huge capital demands behind the AI chip competition.

IBM Report: One Quarter of Data Breaches Caused by AI Attacks, with Individual Losses 20% Higher than Average

According to a report by IBM and the Ponemon Institute, AI-driven attacks have accounted for one quarter of malicious data breach incidents, increasing by 56% year-on-year. The average loss per incident is $6 million, 20% higher than the overall cost. Attacks are highly concentrated, with 62% targeting critical infrastructure, and financial services and energy institutions being the primary targets, posing serious threats to the economy, supply chains, and essential services.

Samsung and Broadcom Team Up with Over $200 Billion Investment, Fully Betting on Next-Generation AI Chips

Samsung and Broadcom have reached a long-term deep cooperation, covering the entire supply chain of memory chips, foundry, and advanced packaging. It is expected that by 2030, the scale of cooperation will break through $200 billion, marking that AI chips have entered a stage of 'design plus manufacturing' deep integration. The core technology is Broadcom's ASIC custom design capabilities combined with Samsung's manufacturing. The two companies will jointly tackle challenges in sub-2nm process technology and HBM high-bandwidth memory fields.

IBM's Quarterly Performance Misses Expectations, Stock Plummets; Executives Insist AI Won't Eliminate Mainframes

IBM's latest financial report shows revenue of $17.2 billion and net profit of $2.2 billion, but the performance fell far short of expectations. A decline in infrastructure business led to a record 25% drop in stock price in one day, and the full-year growth forecast has been downgraded. The main reason is a sharp 42% drop in mainframe sales, with rising costs squeezing budgets, and the contraction of core businesses triggering a chain reaction.

Google's TPU Strategy Shifts: From Internal Tool to Commercial Use, Directly Challenging NVIDIA's Dominance in AI Chips

Google's self-developed TPU transforms into commercial use, collaborating with Broadcom to provide AI infrastructure for customers such as Anthropic, directly targeting NVIDIA's 86% market share. With cost and system advantages, the commercialization process is accelerating, reshaping the competitive landscape of AI chips.