Intel2026-08-10 06:31:55Intel EMIB-T Packaging Faces Yield Hurdles, 50% Target for 2027 Mass ProductionIntel's next-generation EMIB-T advanced packaging technology is confronting serious yield challenges, according to Taiwanese media reports. Unimicron said at its July 29 earnings call that the technology is 'not yet mature,' with three EMIB-T substrate suppliers targeting just 50% yield for initial mass production by the end of 2027. Google's ninth-generation TPU is set to adopt EMIB-T for 2028 production, with MediaTek handling co-design.3890
Anthropic2026-08-07 03:03:43Anthropic confirms in-house chip team as it hires senior engineers for Claude hardwareAnthropic has confirmed that it is building an in-house chip team to design hardware for Claude, turning months of speculation into an official plan. Job postings on the company’s careers page show it is hiring for silicon engineer and technical program manager, silicon roles spanning front-end design, pre-silicon verification, physical design, DFT, analog and mixed-signal work, process and foundry engagement, packaging, and signal integrity. The silicon engineer position carries a salary range of $320,000 to $485,000 and specifically asks for candidates who have “shipped silicon,” pointing to a search for engineers with real tape-out and production experience. Anthropic also said it will keep a multi-chip approach, continuing to use outside hardware alongside any internal designs. The report places the move in a broader AI infrastructure shift already seen at OpenAI, Google, Meta, and reportedly Mistral, with the main attraction being lower inference costs for model-specific workloads, though any payoff would likely take years to materialize.2240
Anthropic2026-07-23 03:30:14Anthropic Hits $30 Billion Run Rate, Secures Google TPU Supply Through 2031Anthropic said its annualized revenue has reached $30 billion and disclosed a long-term compute agreement with Broadcom and Google that secures TPU-related supply through 2031.500
Anthropic2026-07-22 19:15:14Anthropic Weighs In-House AI Chip Effort as Revenue Tops $30 BillionReuters says Anthropic is evaluating whether to build its own AI chips, though the plan remains at an early stage. With annualized revenue above $30 billion and compute demand climbing, supply control has become a central issue.410
Google TPU2026-07-22 12:08:13Google Delivers Eighth-Gen TPU: Splits Training and Inference into Two Chips, Claims 80% Better Performance Per DollarAt Cloud Next 2026, Google unveiled its eighth-generation TPU, splitting training and inference into dedicated TPU 8t and TPU 8i chips. The inference chip claims 80% better performance per dollar over Ironwood. Anthropic and OpenAI have already adopted the TPU cluster.460
Samsung Elect2026-07-15 16:39:32Samsung Weighs Outsourcing Backend Design for Google’s 2nm TPU I/O Chip as Orders Stretch StaffingSamsung Electronics is considering outsourcing backend design work for the I/O chip in Google’s 10th-generation TPU, code-named Icefish, as a wave of 2nm orders puts pressure on its available engineering staff. The TPU is split into a Compute processor and an I/O chip: the Compute die is expected to be made by Taiwan Semiconductor Manufacturing Co. (TSMC) on 1.4nm, while Samsung would manufacture the I/O chip on 2nm to handle data transfer between the Compute processor and high-bandwidth memory, or HBM. Google is designing the chip with MediaTek, with mass production expected as early as 2028. Potential outsourcing partners mentioned in the report include ADTechnology, Gaonchips and Alphachips. ADTechnology and Gaonchips are already tied up with large projects and are said to have limited interest in low-value backend design contracts, which typically amount to tens of billions of won, far below the value of full ASIC projects that can run into the hundreds of billions or even 1 trillion won. Alphachips, by contrast, is described as more willing to participate and sees the Google TPU program as a growth driver.1560
SemiAnalysis2026-06-30 18:31:02SemiAnalysis: NVIDIA's CUDA Moat Slowly Eroding as Google TPU, AWS Trainium, AMD Gain GroundAccording to SemiAnalysis, Google's TPU, Amazon's Trainium, and AMD are accelerating their capture of AI chip market share, gradually eroding NVIDIA's long-standing CUDA ecosystem advantage. This subtle shift in the AI hardware competitive landscape could pose a long-term challenge to NVIDIA's market leadership.1360