Anthropic Develops Custom AI Chip with Samsung
@hardwarebriefcom.bsky.social
@hardwarebriefcom.bsky.socialAnthropic Develops Custom AI Chip with Samsung
@siliconsignalai.bsky.socialThis arXiv paper proposes a vehicle-integrated digital twin framework that merges physics-based modeling with machine learning to monitor bridges and roads via drive-by sensing, aiming to make structural health monitoring more…
#Semiconductors #AIInfrastructure #AI
https://arxiv.org/abs/2610.08822
@technology.hivewire.newsAI PCs face memory squeeze amid server demand #Semiconductors #Chips #News
@semiconductornews.bsky.socialSamsung profits expected to jump nine-fold thanks to AI chip boom
Enquire about sponsoring this channel. https://www.ProNetGrowth.com/?utm_source=bluesky&utm_medium=post&utm_campaign=news-semiconductor&utm_content=semiconductor
@semiconductornews.bsky.socialSamsung profits expected to jump nine-fold thanks to AI chip boom
Enquire about sponsoring this channel. https://www.ProNetGrowth.com/?utm_source=bluesky&utm_medium=post&utm_campaign=news-semiconductor&utm_content=semiconductor
@kautious.comTSMC’s AI-led Q3 points to record revenue—but risk appetite is fragile.
Monthly disclosures imply NT$1.49tn (+51% y/y), above the NT$1.46tn estimate; Sep fell 0.6% m/m. Brent: $104; U.S. 10Y: 5.33%.
Watch Oct. 15 results and guidance. $TSM #Semiconductors #AI
@news.karthihegde.devAI agent designs a complete RISC-V CPU from a 219-word spec sheet in 12 hours
Discussion | hackernews | Author: fork-bomber
@apexjjap.bsky.social[2025 OPEN ACCESS]
AlScN/GaN HEMTs with 4 A/mm on-current and maximum oscillation frequency >130 GHz
2025 Appl. Phys. Express 18 016506
iopscience.iop.org/article/10.3...
#APEX
#OpenAccess
#Physics
#FET
#Heterostructures
#Ohmic
#Alloys
#Nitrides
#Semiconductors
#Electrical
@jlpea-mdpi.bsky.social🎙️ New interview: JLPEA Editorial Board Member Dr. Kyeong‑Sik Min (Kookmin University).
He shares insights on low‑power VLSI, in‑memory computing & emerging memory hardware for edge‑AI.
🔗 Read full interview: www.mdpi.com/journal/jlpe...
#VLSI #MemoryDesign#CMOS
#FPGA #Semiconductors#ICDesign
@qmshahzad.bsky.socialReports say the AI inference-chip startup is weighing offers at a $40-50 billion valuation, roughly double its September round, though no new round is confirmed. #Etched #AIChips #Semiconductors
@news.karthihegde.devFrom Moore's Law to Eroom's Law: Power Constraints and the Productivity Slowdown
Discussion | hackernews | Author: paulpauper
@siliconsignalai.bsky.socialVisionWeave is proposed as a native elastic visual representation capability for MLLMs, learning where and at what granularity to allocate visual tokens rather than using fixed-size patch tokens. The arXiv paper targets…
#semiconductors #AI #MLLMs #inference
https://arxiv.org/abs/2610.07987
@semiconductornews.bsky.socialASML revenue hits $36.6B in FY25, up 15%. Dominates critical chip tech
Please sponsor this exciting channel. https://www.ProNetGrowth.com/?utm_source=bluesky&utm_medium=post&utm_campaign=news-semiconductor&utm_content=semiconductor
@hardwarebriefcom.bsky.socialIntel Joins Elon Musk's Terafab Project in Texas
@semiconductornews.bsky.socialAMD CEO warns AI chip demand outstrips supply, with bookings extending to 2028. Micron workers vote to strike. Full day recap
Support this channel today. https://www.ProNetGrowth.com/?utm_source=bluesky&utm_medium=post&utm_campaign=news-semiconductor&utm_content=semiconductor
@semiconductornews.bsky.socialAMD CEO warns AI chip demand outstrips supply, with bookings extending to 2028. Micron workers vote to strike. Full day recap
Support this channel today. https://www.ProNetGrowth.com/?utm_source=bluesky&utm_medium=post&utm_campaign=news-semiconductor&utm_content=semiconductor
@asiaai.bsky.socialSamsung's push to double 2027 HBM4 prices to Nvidia proves that upstream packaging complexity, not buyer pushback, dictates AI infra terms. If you control the 4nm base die, you hold the ultimate pricing power.
@siliconsignalai.bsky.socialTEE-based LLM inference is slow because GPUs handle most compute; new work proposes splitting inference with DP-protected intermediate representations to balance privacy and speed. Authors show 80% prompt-reconstruction accuracy…
#semiconductors #AI #GPU #privacy
https://arxiv.org/abs/2609.15039