Home » General
Samsung reportedly raised some 4nm foundry prices by as much as 15%. The overlooked reason is HBM4: its logic base die also consumes Samsung’s SF4 foundry capacity.
Read MoreCalibration is priced by the work an instrument requires — bench time, test points, reference standards, paperwork. It is not priced as a fraction of what the instrument is worth. That sounds too obvious to matter. Follow it across a price ladder and it stops being obvious. A calibration bill that barely registers against a…
Read MoreOn July 15, ASML confirmed that Intel had become the first company to ship high-volume logic chips manufactured with High-NA EUV lithography. Some of the Core Ultra Series 3 processors going into laptops right now contain layers printed on a machine that, until this year, had never produced a commercial chip. It is a real…
Read MoreThe third-quarter memory numbers are starting to land, and they say the squeeze is nowhere near done. Contract negotiations point to another 20–30% increase for DRAM and 35–40% for NAND in Q3 2026. The more conservative survey puts it at 13–18% for conventional DRAM and 10–15% for NAND, quarter over quarter. The two forecasts disagree…
Read MoreTL;DR: Korea’s June 29 plan puts over $550B toward new memory fabs — a slice of a much larger $3.1T multi-decade Samsung/SK spending plan. Most new capacity targets high-margin server DRAM and HBM, not everyday PC/laptop memory. Micron is running its own parallel expansion; China’s CXMT is growing fast but not yet server/HBM-ready. Real relief…
Read MoreIf you run AI models in production, the bill has a shape you’ve probably noticed: a surprising amount of it traces back to memory. Every token a model generates has to be served out of fast memory sitting right next to the GPU, and that memory — High Bandwidth Memory, or HBM — is scarce,…
Read More