Micron CTO downplays new memory chip design as 'complex and costly' while Chinese rival CXMT advances DRAM technology

Micron Technology Inc. MU is rejecting an advanced memory architecture gaining attention after Chinese rival ChangXin Memory ...
Tech Times on MSN

Samsung outsources Nvidia AI memory base dies to TSMC as foundry yields fall short

Samsung TSMC HBM base die outsourcing confirms a roughly 40-percentage-point yield gap between Samsung's SF2 gate-all-around ...
Wccftech

While Micron Calls The 4F² Memory Cell Architecture A “Myth,” YMTC Has Followed CXMT In Lining Up A Prototype That Operates At 9,600 MT/s

Micron's Scott DeBoer says 4F² has 'fundamental problems' with scaling, even as CXMT and YMTC appear to be going all in on this technology.
Wccftech

CXMT To Leapfrog Samsung By ~2 Years, As It Gears Up To Introduce A DDR5 RDIMM Product That Leverages The 4F² Memory Cell Architecture Before Year-End

CXMT published its first 4F² paper back in 2021, though its initial version will likely be a partial derivative on the new ...
TechNode

CXMT plans next-generation DDR5 server memory using denser DRAM architecture

Chinese memory maker CXMT plans to launch a new generation of DDR5 registered memory modules by the end of 2026 using its 4F² DRAM architecture. President ...

SanDisk Shares Fall 2.8% as Chinese Memory Chip Advances Raise Competition Concerns

SanDisk Corporation (NASDAQ:SNDK) shares declined 2.8% to $1,445.50 in pre-market trading, extending recent losses as investors assessed the potential impact of new Chinese memory chip technology and ...
VentureBeat

GAM takes aim at “context rot”: A dual-agent memory architecture that outperforms long-context LLMs

For all their superhuman power, today’s AI models suffer from a surprisingly human flaw: They forget. Give an AI assistant a sprawling conversation, a multi-step reasoning task or a project spanning ...
VentureBeat

Google's new neural-net LLM architecture separates memory components to control exploding costs of capacity and compute

A new neural-network architecture developed by researchers at Google might solve one of the great challenges for large language models (LLMs): extending their memory at inference time without ...