Marvell

TechCrunch pieces reveal hard funding/data-center economics around AI compute, highlighting XCENA’s memory-centric MX1 chip and multiple seed/Series B rounds, underscoring a shift in AI inference architecture alongside RSI/AGI discourse.

XCENA raises $135M Series B at a $570M valuation to back memory-centric AI hardware (MX1) with mass production expected by late 2026 and revenue in 2027.

AWS Trainium3 unveils switched-scale-up rack architectures (NL32x2 Switched and NL72x2 Switched) with Gen1/Gen2/Gen3 switch trays, advancing memory bandwidth and per-TCO performance while expanding software openness and ecosystem bets to challenge Nvidia and AMD in the datacenter AI race.