A hands-on test of the 256GB M5 Ultra Mac Studio found it capable of running large local AI models, including Qwen3.8-Flash-Next, fast enough to power personal assistant apps like Open Minis and Hermes Agent without cloud costs. The reviewer compared it against the prior M3 Ultra Mac Studio and a gaming PC with an RTX 5090, noting the Mac's smaller size, lower noise, and unified memory architecture as key advantages despite the 5090 retaining a memory-bandwidth edge.
Newegg is selling a 32GB (2x16GB) V-Color Manta XSky DDR5-6000 CL30 memory kit for $429 with promo code TEF366, undercutting comparable kits by roughly $70. A black variant, the Manta XFinity, is also available for $439 with code TEF364.
Apple's latest Mac mini keeps the same compact redesign introduced with the M4 generation but swaps in the new M6 chip. The entry configuration with 16GB RAM and 256GB storage now costs $899, a $300 jump from the $599 starting price of the 2024 M4 model with identical specs. Apple attributes the increase to a global memory shortage affecting RAM and storage pricing.
A new project called Lossless-memory offers a long-term memory system for personal AI assistants that stores complete conversation logs rather than summarizing or embedding them for similarity search. Every record carries a timestamp, forming what the creator calls a 'Temporal Backbone,' enabling time-first queries like 'yesterday evening, about the budget.' The system has reportedly been running daily since July 2026 for a single user as the memory layer of an AI assistant.
Acer chairperson Jason Chen said PC prices will keep rising 5-20% through late 2025, plateau in early 2027, then start falling as memory chip supplies normalize sooner than manufacturers claim. He pushed back on suppliers like SK hynix and Adata, who warn shortages could persist until 2030 or beyond, arguing these companies are inflating fears to protect profit margins.
A writer recounts memorizing the first 100 digits of pi three years ago after reading Joshua Foer's 'Moonwalking with Einstein.' They used a person-action-object (PAO) system, a memorization method that pairs pairs of digits with vivid mental images of people performing actions on objects, building on the classical method of loci or memory palace technique.
Prominent leaker Kopite7kimi now expects Nvidia's Rubin-based RTX 6000 series to launch in 2028, later than his earlier forecast of late 2027. This aligns with a similar prediction from leaker Kepler_L2, who also expects AMD's RDNA 5 Radeon 10000 series to slip to 2028, though AMD's smaller 'AT2' GPU remains on track for next year.
ChatGPT's memory feature has evolved from requiring users to explicitly say 'remember this' into an automatic system that continuously updates and revises what it retains about you, a process OpenAI calls 'dreaming.' This means the chatbot now decides on its own what details from your conversations to keep, edit, or discard, shaping how it responds to future prompts.
Cloudflare engineers found that its internal load-balancing service, Pingora Backend Router, was consuming far more memory than necessary in its pingora-ketama consistent-hashing library. By reworking the underlying algorithm with better math and Rust code, the team eliminated the excess memory usage across its global fleet.
According to Reuters sources, Chinese DRAM maker CXMT is considering building an experimental 3D NAND production line at its planned second facility near Beijing, which has yet to break ground. The company has also set up a Beijing research institute reportedly working on NAND flash development, and has discussed potential NAND supply with at least one prospective customer targeting AI storage products. CXMT has not officially confirmed any of these NAND ambitions.
Analysts at Morgan Stanley, JPMorgan, and TrendForce report that legacy memory types like NOR flash and SLC NAND are facing severe supply shortages as manufacturers redirect production capacity toward more profitable HBM and DRAM chips driven by AI demand. Contract prices for both chip types have already more than doubled in the first half of 2026, with SLC NAND expected to rise another 120-170% in the second half.
The jemalloc project shipped version 5.4.0, a maintenance-focused update spanning more than 160 commits aimed at reducing technical debt through refactoring, bug fixes, expanded test coverage, and portability fixes. Key additions include an EXTENT_ALLOC_FLAG_PINNED flag letting custom allocation hooks mark non-reclaimable mappings like HugeTLB pages for preferential reuse, plus new mallctl interfaces to report pinned-memory and mutex statistics. The update also replaces the old fixed tcache refill/flush policy with a dynamic system that adapts to observed demand, removing seven legacy tuning controls in the process.
GoKawiil