#ai-chips

#ai-chips

Developing in-house AI chips reduces dependence on outside designers like Nvidia, challenges Nvidia's market dominance, boosts performance, and saves energy costs. [ more ]

Microsoft's Copilot AI service will run locally on next-gen AI PCs with over 40 TOPS of power. This move aims to reduce lag, improve performance, and enhance privacy.

Intel is planning to enhance the NPU speeds in its chips, with the Lunar Lake chips expected to triple the current NPU speeds by 2025. [ more ]

Intel CEO Pat Gelsinger promises an epic comeback by focusing on core engineering and manufacturing plan.

Intel announces rebrand of its 'foundry' business, aiming to produce efficient AI-era chips to compete with TSMC. [ more ]

Nvidia reports massive revenue growth driven by demand for AI chips

Intel aims to be a top chip foundry by 2030 with $15 billion in foundry business orders

OpenAI CEO seeks trillions of dollars for new chip production effort [ more ]

Google is developing its own Arm-based CPU, Axion, for AI workloads in data centers.

Axion CPU offers better performance than general-purpose Arm chips and Intel's processors. [ more ]

Nvidia's dominance in AI stems from both hardware and software

Tech companies form coalition to challenge Nvidia's AI dominance [ more ]

Big tech companies like Amazon, Google, Meta, and Microsoft are building their own AI chips to reduce dependence on Nvidia and control their own destiny.

The demand for specialized AI chips has outpaced Nvidia's supply, leading to shortages and high costs for tech giants. [ more ]

Altman called for a $5 to 7 trillion global investment in AI chips.

Industry analysts are surprised by the investment amount needed for AI advancement. [ more ]

Nasdaq-100 Technology Sector index rose 67% in 2023, making it an excellent time to invest in tech

Microsoft is positioned to be a top tech stock to buy in 2024 with its strong growth and investments in AI [ more ]

OpenAI CEO Sam Altman is reportedly seeking trillions in funding to bolster global production of AI chips.

The company believes it can more than double its revenue to $2 billion in 2025 based on strong interest from business customers. [ more ]

OpenAI CEO Sam Altman aims to raise billions for an AI chip venture to develop a global network of chip fabrication factories.

The race to manufacture high-powered chips for AI systems has intensified, with limited fabs and deep-pocketed investors needed. [ more ]

AWS has unveiled new AI chips, Trainium2 and Graviton4, aimed at improving performance and energy efficiency in model training and general use.

Trainium2 chips offer up to 4x better performance and 2x energy efficiency for model training, while Graviton4 processors promise a 30% increase in general performance when using a trained AI model.

These advancements in AI chips aim to lower costs, improve time-to-results, and enhance scalability for AWS customers. [ more ]

Baidu is largely unaffected by US chip export bans and can continue to use silicon allowed to be sold in China.

Baidu has a substantial reserve of AI chips and other alternatives to support AI-native apps for the next year or two.

Baidu is seeking alternatives from elsewhere and modifying its deep learning framework to be compatible with different types of AI chips. [ more ]

Tech Spark AI is working on its own ChatGPT alternative and has acquired $1.4 million in funding.

Microsoft Ignite event announced new AI-focused chips for data centers, rebranded Bing Chat as Microsoft Copilot, and introduced Copilot Studio program.

YouTube is rolling out new rules for AI content to prevent misinformation, including requiring creators to disclose use of generative AI. [ more ]

Chinese technology giant Tencent Holdings has a stockpile of AI chips from U.S. maker Nvidia, but new U.S. restrictions on chip sales to China will eventually impact its cloud services.

Tencent will need to use its existing stock of Nvidia chips more efficiently and seek domestically produced alternatives due to the ban on chip exports to China.

The ban on chip exports will affect cloud services where Tencent sells computing power to clients. [ more ]

AMD estimates the data center AI processor market to be worth $45 billion this year, with expectations for it to grow to $400 billion by 2027.

AMD launched two new AI data center chips from its MI300 lineup, one focused on generative AI applications and another geared toward supercomputers. [ more ]

AWS announced new generations of its Graviton and Trainium chips, offering better compute performance and memory bandwidth for cloud workloads and AI model training.

Amazon Bedrock, the foundation model building service, introduced new features including Agents, Custom models, and Guardrails to help organizations customize and control their AI content. [ more ]