Chips
32 articles RSS
Etched Exits Stealth With Working Sohu Transformer ASIC, $800 Million Raised at a $5 Billion Valuation and Over $1 Billion in Chip Orders
Etched unveiled working A0 silicon for Sohu, its transformer-only inference ASIC, disclosing $800 million raised at a $5 billion valuation and more than $1 billion in signed customer contracts.
OpenAI and Broadcom Unveil 'Jalapeño,' a Custom LLM-Inference Chip Designed in Nine Months and Targeting Deployment by the End of 2026
OpenAI's first custom silicon, the Broadcom-built Jalapeño inference accelerator, went from design to tape-out in nine months and targets deployment by the end of 2026.
Tensordyne Sends Its Logarithmic-Math AI Chip Napier to TSMC, Claiming Four Times the Speed at a Fifth the Power of an Nvidia GB300 System
The startup says its 3nm Napier inference chip turns matrix multipliers into adders, with commercial sales of a 72-chip system slated for the second half of 2027.
Qualcomm Announces Snapdragon C, Pushing Windows on Arm and an On-Chip NPU Into $300 Budget Laptops
Qualcomm unveiled the Snapdragon C Platform for entry-tier Windows laptops starting at $300, with Acer, HP, and Lenovo as launch partners and devices expected later this year.
d-Matrix Moves Its Corsair Inference Accelerator Into Full Production, Betting On-Chip SRAM Beats the GPU at Decode
d-Matrix says its Corsair inference accelerator is in full production, pairing on-chip SRAM with GPUs to cut AI response times by more than 10x.
SpacemiT Demos a 'Usably Quick' RISC-V Mini Desktop on Its 16-Core K3, as the Open Architecture Inches Toward Everyday Computing
At the Ubuntu Summit, SpacemiT showed its 16-core Key Stone K3 running Ubuntu responsively, with a Banana Pi board priced at roughly twice a Raspberry Pi 5.
Intel's Diamond Rapids Xeon 7 Drops Hyperthreading for Up to 192 Performance Cores, Slated for 2027 on 18A-P
At Computex, Intel detailed its Diamond Rapids Xeon 7: up to 192 P-cores on 18A-P, 16 DDR5 channels and PCIe 6.0, but no Hyperthreading, arriving 2027.
Intel Launches Xeon 6+ 'Clearwater Forest' on 18A, Its First Data Center CPU on the New Node, With Up to 288 Cores
Intel's first 18A server chip ships with up to 288 Darkmont E-cores and 576 MB of L3 cache, claiming a 30% per-thread lead over AMD's 192-core EPYC 9965.
NVIDIA RTX Spark Officially Launches at Computex 2026: 20-Core Arm CPU, Blackwell GPU, and 128GB Unified Memory Coming From Six OEMs This Fall
NVIDIA officially unveiled the RTX Spark Superchip at Computex 2026 — a 20-core Arm Grace CPU paired with a Blackwell GPU and up to 128GB of unified memory, launching in laptops from six OEMs in fall 2026.
AMD Begins Production Ramp of EPYC Venice on TSMC 2nm, First HPC Chip on Advanced Node
AMD's 6th Gen EPYC 'Venice' is ramping on TSMC 2nm — the first HPC chip on the node — with 256 Zen 6 cores and 1.6 TB/s memory bandwidth.
Fractile Raises $220 Million Series B to Build In-Memory AI Inference Chips, Reaches Unicorn Valuation
London-based Fractile closes a $220M Series B led by Accel, Founders Fund, and Factorial Funds to tape out SRAM-based inference chips promising radically faster and cheaper AI output generation.
NVIDIA Set to Unveil First Arm Laptop Chip at GTC Taipei, Entering a Market Dominated by Qualcomm, Intel, and AMD
NVIDIA will reveal the N1/N1X SoC at Jensen Huang's June 1 GTC Taipei keynote, a Blackwell-GPU Arm chip co-developed with MediaTek targeting Dell and Lenovo laptops.