Latest about Performance
AI Fees Up to 15x Cheaper for English Than Other Languages
Tokenization process gives unfair advantage to English.
EeroQ Tapes Out "Wonder Lake" Quantum Processing Unit Chip
Based on scalable CMOS technology for rapid iteration and manufacturing ease.
Stable Diffusion Runs in Only 260MB of RAM on Raspberry Pi
Vito Plantamura has successfully managed to run Stable Diffusion on a Raspberry Pi Zero 2 by creating what he calls OnnxStream.
Intel Strives to Make Path Tracing Usable on Integrated GPUs
Intel is presenting several new papers to three universities focusing on optimizing path-tracing performance, to the point where integrated GPUs could run path-tracing in real-time.
Apple M2 Ultra GPU Outpaces RTX 4070 Ti in Early Compute Benchmarks
In early compute benchmarks, Apple's M2 Ultra chip appears to perform somewhere between an RTX 4070 Ti and RTX 4080 in Geekbench 6 and GFXBench testing.
Sony Project Q Handheld Rumored to Offer Just 3 or 4 Hours Battery Life
Streaming-only devices offload processing, so they should offer extended battery life, but a tipster suggests this isn't true for the Sony Project Q handheld.
Imec Reveals Sub-1nm Transistor Roadmap, 3D-Stacked CMOS 2.0 Plans
Imec, the most advanced semiconductor research firm in the world, recently shared its silicon and transistor roadmap at its ITF World summit in Antwerp, Belgium.
AMD and Intel Battle for Windows 11 AI Acceleration Lead
AMD and Intel talk about how they will be best to accelerate AI on Windows 11, and highlight the AI processing hardware in their latest SoCs.
