Rivian's Full Autonomy Gamble Takes to Palo Alto Streets
A firsthand ride in the R1S shows Rivian pushing its self-driving strategy into real-world suburban traffic.
Compute & Hardware
Chips, datacenters, interconnect, and the power they draw.
Reported $120B Google-Marvell Chip Deal: A Closer Look
The Next Platform breaks down the reported $120 billion chip agreement between Google and Marvell, questioning its true scope.
Hot Chips 2026: Intel and Fujitsu Detail New CPUs
Chips and Cheese's Hot Chips 2026 coverage spans four processors, with Intel fielding three designs and Fujitsu extending its HPC-focused Monaka line.
IBM Built a Secret Cold War Codebreaker for the NSA
A specialized IBM cryptographic machine, hidden for decades, ran jobs up to 200 times faster than other computers of its era.
Memory Prices Are No Longer Cheap, The Next Platform Argues
A new analysis from The Next Platform highlights how much more expensive memory has become compared with its historically low prices.
Samsung Builds HBM Base Dies on Logic Node for AI Memory
Samsung's logic-node base die for HBM could give AI memory stacks new headroom as ML demand grows.
Cerebras Overclocks WSE-3 for Faster Inference in Nexus CS-4
Cerebras is pushing its waferscale WSE-3 chip beyond its normal clock speed to improve inference performance in the new CS-4 system.
Cisco's Supercomputing Push Finally Gains Traction
Cisco can now sell high-performance computing systems at volume, bundling the network fabric that ties them together.
Qarakal's Pangaea Architecture Applies Classical Design Lessons to Quantum
A new quantum startup is borrowing proven ideas from classical computing to shape its Pangaea architecture.
Neoclouds Amass Cash Faster Than They Earn Revenue
New cloud providers are stockpiling capital at a rate that outpaces their income, signaling a race to build infrastructure before demand catches up.
IBM Focuses on System Design to Reach Ultra-Cold Temperatures for Quantum
IBM's latest system-design efforts target the extreme cooling required for quantum processors.
Open Source, Open Weight, Closed: AI's 'Open' War Explained
The Next Platform examines the widening battle over what 'open' really means for AI models.
Supermicro's GenAI Tailwind Is Not Exclusive, Analysis Says
A new analysis from The Next Platform argues that the AI infrastructure buildout will lift Supermicro and equally benefit other compute vendors.
Nvidia Targets AI Cost Efficiency With New Model and Router
Nvidia's latest releases focus on improving price-to-performance for generative AI workloads.
Arista Aligns with Three-Directional Growth in AI Networks
A new analysis ties Arista's trajectory to the expanding scale of AI networks in three distinct dimensions.
Synopsys Talks Physics of Chip Design at DAC 2026
A Chips and Cheese interview with Synopsys examines how physical constraints shape modern chip design.
AMD to Bake AI Inference Into Chips With Taalas Technology
The Next Platform reports that AMD is moving to embed AI inference directly into its processors, a shift that could reshape how inference workloads are handled in data centers.
Why Binary Translation Is Central to PC Market Battles
A Chips and Cheese analysis explains why the PC market is so hard for CPU designers and what binary translation means for challengers.