Share
Linkedin
Twitter
Facebook
Whatsapp
Copy Link
Amazon is negotiating with investors to transfer roughly $8 billion worth of Nvidia Grace Blackwell chips into a special-purpose vehicle. The cloud giant would lease the AI chips back while shifting ownership to external investors, marking a strategic shift toward an asset-light approach as tech companies seek creative ways to finance massive AI infrastructure spending.
Elon Musk announced Tesla cut AI5 chip memory from 144GB to 96GB and AI6 from planned capacity to 144GB amid severe memory supply constraints. The reduction aims to secure volume production for Optimus humanoid robots, though Micron Technology estimates each humanoid robot requires 200GB of DRAM.
Federal authorities arrested Greg Lui, CEO of Earthmade Computer, for allegedly smuggling over $300 million worth of export-controlled Nvidia AI chips to China. The scheme involved routing servers containing A100 and H100 GPUs through Malaysia and Singapore using fraudulent paperwork. Officials increasingly question Nvidia's compliance protocols as smuggling arrests mount.
Sam Altman-backed startup Volantis raised $88 million to develop photonic AI chips that use laser-based communication to connect 220 memory chiplets. The A-1 system targets 10 terabytes of memory at 240 terabits per second, aiming to serve 20 trillion parameter models at 10,000 tokens per second when it ships in 2027.
Google launched its first Project Suncatcher satellite carrying Tensor Processing Units into low-Earth orbit aboard a SpaceX rocket on October 1. The mission tests whether AI chips can survive space conditions and function in orbital data centers. Google's research indicates SpaceX needs approximately 1,800 Starship launches before space-based AI infrastructure becomes economically viable.
Apple's Mac Studio with M5 Ultra processor positions itself as a premium local AI workstation with 96 GB unified memory and unprecedented GPU performance. Starting at $5,499, it outperforms competitors like Nvidia's RTX 5090 in rendering workflows while enabling on-device AI tasks. The system's unified memory architecture and 1.2 TB/s bandwidth make it ideal for running large language models locally with robust privacy controls.
Synopsys shares surged 10% following the announcement of groundbreaking AI partnerships with OpenAI and Amazon at its investor day. The chip design software maker revealed a $1 billion-plus Amazon deal and joint AI model development with OpenAI, alongside fiscal 2027 revenue guidance exceeding analyst expectations.
Wall Street banks launched syndication of a $60 billion debt package to fund Anthropic's lease of Broadcom AI chips, marking the largest chip-financing deal to date. The financing includes $42 billion in senior secured loans and $18 billion in junior debt, with Blackstone committing $9 billion. Anthropic's IPO prospectus reveals potential conflicts of interest in Broadcom's dual role as supplier and lender.
Chinese tech giant Tencent signed a five-year lease with Oracle for access to 100,000 advanced AI chips across Southeast Asia data centers in a deal worth $7 billion. The arrangement, with 30% paid upfront, marks Tencent's largest overseas lease as Chinese firms race to secure offshore AI compute resources amid tightening US export controls.
Silicon Valley startup CScale raised $145 million in Series C funding backed by Nvidia and Intel to develop optical interconnect technology for AI data centers. Led by former Cisco executive Martin Lund, the company aims to replace copper cables with fiber-optic connections that link thousands of AI accelerators across dozens of racks by 2028.
OpenAI and Synopsys announced a multi-year revenue-sharing partnership to develop GPT-Synopsys, a specialized AI model for semiconductor design. The model will autonomously operate electronic design automation tools, enabling engineers to explore more design options and deliver sophisticated chips faster while maintaining verification standards.
Rivian clarified its autonomy hardware rollout timeline at an investor conference, revealing that Lidar-equipped R2 vehicles with the in-house RAP1 chip will go to employees in late 2026. Paying customers won't receive the Gen 3 autonomy platform until 2027, marking a shift from earlier communications that suggested a broader late-2026 launch.
Amazon and Synopsys announced a multi-year strategic collaboration worth over $1 billion to accelerate AI chip design and custom silicon development. Amazon becomes the lead customer for Synopsys' application-optimized IP while Synopsys adopts AWS compute services and Amazon Bedrock for AI-powered engineering workflows.
Meta has classified its AI data centers as experimental pilot models to claim billions in federal tax credits, cutting $3.9 billion from its 2025 tax bill. The aggressive tax strategy involves treating Nvidia chips as research supplies, but Meta's own accountants warn the IRS could overturn the savings, highlighting uncertainties in a legal gray area.
Samsung launched the Galaxy Tab S12 Ultra and Galaxy Tab S12+ with breakthrough 3nm processor technology, delivering up to 61% faster NPU processing and 19% faster CPU performance. The tablets come with Photoshop Mobile pre-installed and target creatives and professionals, though prices start at $1,199—a $200 increase over last year's lineup.
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.