Samsung hikes memory chip prices by up to 60% as shortage worsens, sources say (www.reuters.com)

Samsung has raised prices for server memory chips by roughly 30% to 60% amid worsening supply tightness, Reuters reports. The hikes apply to server-oriented memory components used in data centers—where demand from AI training and inference workflows has driven heavy consumption—and are adding near‑term cost pressure on cloud providers, hyperscalers and data infrastructure firms that procure large volumes of DRAM/NAND. Steeper memory costs can raise the total cost of ownership for AI clusters, slow procurement cycles, and push customers toward capacity optimization measures (e.g., model sparsity, quantization, tighter resource scheduling). For Samsung, the price surge is a welcome revenue boost after trailing rivals in the race to supply advanced AI accelerators; stronger memory margins can free cash to invest in next‑gen chips and manufacturing. For the AI/ML community, the move underscores how hardware supply dynamics still materially affect model training economics and deployment timelines. Expect short‑term inflation in compute costs, a scramble by operators to renegotiate contracts or shift sourcing, and renewed focus on software optimizations and chip-level efficiency as ways to blunt the impact of higher memory prices.