Topic: custom silicon
-
Intel Goes Custom Chip Foundry, Nvidia Named as Key Customer
Intel has appointed a dedicated executive and secured a multi-year agreement to produce custom Xeon CPUs for Nvidia, positioning itself as a competitor in the custom chip design market. The company's custom silicon business, led by Srini Iyengar, aims to meet the growing demand for tailored proce...
Read More » -
Anthropic designs custom hardware to power Claude
Anthropic is assembling an in-house team to design custom silicon for its AI models, with job listings for semiconductor roles now live and confirmed by the company. The company will adopt a multi-chip strategy, continuing to use external vendors' processors alongside its own custom designs rathe...
Read More » -
Anker develops custom chip to embed AI across product lineup
Anker unveiled its custom Thus chip, the world's first neural-net compute-in-memory AI audio chip, which is smaller and more power-efficient than conventional chips by performing computations where the model is stored. The Thus chip will first appear in Soundcore's flagship earbuds, enabling larg...
Read More » -
Why AI giants like OpenAI and SpaceX are building their own chips
OpenAI announced Jalapeño, a custom AI chip developed with Broadcom, to reduce reliance on Nvidia as a single supplier. The custom chip strategy is a hedge to gain better performance, efficiency, and control over hardware, similar to Apple's shift from Intel processors. This move places OpenAI am...
Read More » -
Google embeds Gemini AI directly into new processor chip
Google is developing a processor code-named “Frozen v2” where the Gemini AI model is permanently etched into the silicon, rather than being loaded as software, making the chip itself the model. This hardwired design could offer massive gains in efficiency, speed, and power consumption by eliminat...
Read More » -
SiFive raises $400M in final pre-IPO funding round
On April 9, 2026, SiFive secured $400 million in its final private funding round, valuing the RISC-V chip IP company at $3.65 billion and setting the stage for an IPO. The investment, led by Atreides Management and backed by Nvidia, highlights surging demand for its open-standard, customizable CP...
Read More » -
Rivian's Bold Autonomy Plan: Custom Chips, Lidar, Robotaxi Vision
Rivian announced a major push into autonomous driving, unveiling plans for new custom hardware and software with a long-term vision that includes entering the robotaxi market to compete with leaders like Tesla. A key product is the "Universal Hands-Free" driver-assistance system launching in earl...
Read More » -
Microsoft's 132-Core Azure Cobalt 200 CPU Targets Performance Boost
Microsoft has launched the Azure Cobalt 200, a 132-core Arm-based CPU built on TSMC's 3nm process, designed to enhance performance and efficiency for its cloud services. The processor offers over 50% higher performance than its predecessor, integrates hardware accelerators for compression and cry...
Read More » -
Majestic Labs skips GPUs to break Nvidia's memory wall
Majestic Labs, a Tel Aviv startup, is challenging the AI hardware focus on raw compute by arguing that memory bandwidth is the real bottleneck, and has designed a GPU-less server called Prometheus to address this. Prometheus replaces GPUs with Ignite AI Processing Units and pools up to 128TB of l...
Read More » -
Cerebras to partner with all AI hardware players except NVIDIA
Cerebras CEO Andrew Feldman declared the company will partner with every major AI hardware maker except NVIDIA, positioning the firm as the "connective tissue" for buyers building diverse AI stacks. The company is selling itself as a hedge against the financial risk of vendor concentration, argui...
Read More » -
Apple's N1 Wi-Fi Chip Outperforms Broadcom in Every Test
Apple introduced its first proprietary wireless chip, the Apple N1, replacing third-party suppliers and enhancing control over core technologies in its devices. The N1 chip significantly improved Wi-Fi download and upload speeds on the iPhone 17 compared to the iPhone 16, with the most notable ga...
Read More » -
Meta Unveils 4 New AI Chips for Its Systems
Meta has introduced four new custom AI chips, the MTIA family, to power generative AI and content ranking for its platforms, marking a significant move toward developing in-house silicon and reducing reliance on external suppliers. Developed with Broadcom and TSMC using RISC-V architecture, the c...
Read More » -
Nvidia's Q4: A Make-or-Break Moment for AI Hardware
Nvidia's upcoming quarterly report is a key indicator for the AI hardware sector, with its dominant role in powering AI models having fueled its massive valuation, but the focus is now shifting to the sustainability of the broader AI investment cycle. Analysts expect another quarter of staggering...
Read More » -
Rivian Spin-Out Raises $500M for AI-Powered Industrial Robots
Mind Robotics, a spin-out from Rivian, has raised $500 million in Series A funding, valuing the company at around $2 billion and bringing its total funding to $615 million. The company aims to develop more dexterous industrial robots by leveraging real-world manufacturing data from Rivian's facto...
Read More » -
GlobalFoundries Boosts Chip IP with Synopsys ARC and RISC-V Acquisition
GlobalFoundries expands its processor IP portfolio by acquiring Synopsys's ARC and RISC-V IP business, strategically broadening its offerings for automotive, embedded, and industrial applications. The acquisition provides customers with a comprehensive, integrated suite of silicon-proven IP and d...
Read More » -
Google Bets on Data Center Expert to Lead AI Race
Google has elevated Amin Vahdat to a new executive role reporting directly to the CEO, signaling a massive strategic focus on AI infrastructure and major capital investment. Vahdat is a key architect behind Google's critical AI hardware, including its custom TPU chips and the high-speed Jupiter d...
Read More » -
Nvidia's Vera Rubin AI Chips Now in Full Production
Nvidia's new Vera Rubin AI platform is now in full production, promising to cut AI operational costs to roughly one-tenth of the current leading system and making it harder for clients to leave its ecosystem. The platform offers major performance gains, requiring far fewer chips for training mode...
Read More »