OpenAI has made a significant breakthrough in chip development with its new 'Jalapeño' chip, which outperforms Nvidia's flagship GPU. Co-developed with Broadcom, the chip claims up to 1.9x throughput per kilowatt and 3.6x lower latency. In benchmarks against Nvidia's GB300 and Blackwell, OpenAI's chip showed impressive results, with 1.5x to 1.9x more AI work at peak throughput and 1.7x to 3.6x lower end-to-end latency.
This development comes as Nvidia is set to receive up to $105 billion in financing for OpenAI-leased data centers. OpenAI is also working on a second-generation chip and concept work on a third generation, signaling its commitment to advancing AI hardware.
Meanwhile, AMD's stock has risen after a bullish research report from Raymond James, with expectations that the company will take market share from rival Intel and close the gap with industry leader Nvidia. Taiwanese furniture-component maker King Slide Works is also benefiting from the AI boom, producing rails used in AI server racks.
Other developments in the AI space include Mate Security's launch of Gamebooks, a tool for governing AI investigations, and a study on using large language models as judges for evaluating conversational voice agents. Lumentum's stock has surged on the back of the AI optics trade, while Nvidia has reported strong sales growth driven by AI demand.
Australia is also pursuing a national AI framework, with a central office in the Department of the Prime Minister and Cabinet overseeing the effort. Peter Thiel's hedge fund is shifting its focus towards AI-related electricity and power infrastructure, highlighting the importance of energy supply and grid capacity in the AI industry.
Key Takeaways
• OpenAI's new 'Jalapeño' chip outperforms Nvidia's flagship GPU with up to 1.9x throughput per kilowatt and 3.6x lower latency. • OpenAI's chip is set to be deployed later in 2026 with volume production in 2027. • Nvidia is set to receive up to $105 billion in financing for OpenAI-leased data centers. • AMD's stock has risen on AI optimism, with expectations of taking market share from Intel and closing the gap with Nvidia. • Mate Security has launched Gamebooks, a tool for governing AI investigations. • A study found that LLM-based evaluation can be effective in evaluating conversational voice agents. • King Slide Works is benefiting from the AI boom, producing rails used in AI server racks. • Lumentum's stock has surged on the back of the AI optics trade. • Nvidia has reported strong sales growth driven by AI demand. • Australia is pursuing a national AI framework to attract investment and protect social license.OpenAI's Jalapeño Chip Outpaces Nvidia's Flagship GPU
OpenAI has developed a new chip called Jalapeño that outperforms Nvidia's flagship GPU. The chip, co-developed with Broadcom, claims up to 1.9x throughput per kilowatt and 3.6x lower latency. OpenAI's chip was tested against Nvidia's GB300 in three open models and showed significant performance improvements. This comes as Nvidia is set to receive up to $105 billion in financing for OpenAI-leased data centers. OpenAI is also working on a second-generation chip and concept work on a third generation.
OpenAI's New Chip Outperforms Nvidia's Blackwell
OpenAI's new chip, Jalapeño, has been benchmarked against Nvidia's Blackwell and shown impressive results. The chip delivers 1.5x to 1.9x more AI work at peak throughput and 1.7x to 3.6x lower end-to-end latency. However, the comparison is limited to inference only and does not affect Nvidia's training dominance. OpenAI's chip is set to be deployed later in 2026 with volume production in 2027.
Mate Security Launches Gamebooks for AI Investigations
Mate Security has launched Gamebooks, a new tool for governing how AI agents run investigations. Gamebooks describe investigative intent and set out what needs to be investigated, what evidence must be established, and which conditions should redirect the investigation. The tool aims to provide a more structured approach to security investigations and prevent AI agents from taking unauthorized actions.
Benchmarking LLM Judges for Voice-Agent Evaluation
Researchers have investigated the use of large language models (LLMs) as judges for evaluating conversational voice agents. The study compared human judgments with GPT-4.1 and GPT-5 on telecom and retail voice-agent conversations. The results show that LLM-based evaluation can be effective but is metric- and configuration-dependent.
AMD Stock Rises on AI Optimism
AMD's stock has risen after a bullish research report from Raymond James. The analyst upgraded AMD's stock to a strong buy and boosted the share price forecast. AMD is expected to take market share from rival Intel and close the gap with industry leader Nvidia.
Taiwan's King Slide Works Thrives on AI Demand
King Slide Works, a Taiwanese furniture-component maker, is benefiting from the AI boom. The company makes rails used in AI server racks, which are crucial for the operation of artificial intelligence systems. King Slide Works' success highlights the importance of supply chain management in the tech industry.
Peter Thiel Shifts Focus to AI Power Assets
Peter Thiel, co-founder of PayPal, is shifting his hedge fund's focus towards AI-related electricity and power infrastructure. This shift could have significant implications for the AI industry, as it highlights the importance of energy supply and grid capacity.
Lumentum Climbs on AI Optics Trade
Lumentum's stock has surged on the back of the AI optics trade. The company's stock has more than doubled this year, driven by strong demand for AI-related products. However, the stock's rally has been driven by residual demand and momentum traders.
Nvidia Beats Estimates on AI Demand
Nvidia has reported strong sales growth driven by AI demand. The company's revenue rose to $7.1 billion, up from $6.4 billion in the same period a year ago. Nvidia's AI business has been a key driver of its growth, with AI revenue more than doubling in the quarter.
Australia Pursues National AI Framework
Australian Prime Minister Anthony Albanese has announced plans to pursue a single national framework for AI. The framework will be overseen by a central office in the Department of the Prime Minister and Cabinet. The goal is to attract investment in frontier AI while protecting social license.
Sources
- OpenAI’s 700W Jalapeño ASIC outpaces 1,400W Nvidia flagship GPU — claims up to 1.9x throughput per kilowatt and 3.6x lower latency, co-developed with Broadcom
- OpenAI Says Its New Chip Outperforms Nvidia's Blackwell As Nvidia Prepares Earnings Release
- Exclusive: Mate Security launches Gamebooks to govern how AI agents run investigations
- Benchmarking LLM Judges for Voice-Agent Evaluation: Reliability, Calibration, and Human Oversight
- AMD Stock Is Rising. Here's Why This Top AI Stock Is a Strong Buy Right Now
- One of Taiwan’s Biggest AI Winners Isn’t a Chip Maker
- Why Is Palantir (PLTR) In Focus After Thiel Turned To AI Power Assets?
- Lumentum Climbs 4% as the AI Optics Trade Extends a Second Session, Coherent and Applied Optoelectronics Drift
- Nvidia beats estimates as AI demand surges, projects 70% sales growth by 2028
- Australian prime minister wants AI investment that maintains social license | MLex | Specialist news and analysis on legal risk and regulation
Comments
Please log in to post a comment.