XCENA Raises $135 Million for AI Memory Chip as It Targets 10-to-1 Server Cuts
Updated
Updated · TechCrunch · May 29
XCENA Raises $135 Million for AI Memory Chip as It Targets 10-to-1 Server Cuts
2 articles · Updated · TechCrunch · May 29
Summary
$135 million in Series B funding valued XCENA at $570 million, giving the four-year-old startup $185 million raised in total to commercialize its AI memory-bottleneck chip.
The company says its MX1 chip moves compute closer to DRAM, handling preprocessing, KV-cache management and data caching inside the memory module instead of shuttling data repeatedly through CPUs and GPUs.
XCENA claims that architecture could shrink workloads that now need 10 servers to one, a pitch aimed at hyperscalers spending tens of billions of dollars a year on AI infrastructure.
The MX1 is still a prototype, with mass-production chips due from Samsung's foundry by end-2026 and revenue expected in 2027, while talks with global memory vendors remain at an early stage.
Investor interest comes as demand for memory solutions has surged since late last year and chipmakers increasingly frame AI inference as a memory-scaling problem, not just a compute race.
Valued at over half a billion dollars, will XCENA's radical 'compute-in-memory' chip win over hyperscalers from established giants?
XCENA integrates thousands of cores into memory. Is this the key to AI efficiency or a manufacturing nightmare in waiting?
XCENA Secures $135M to Lead AI Memory Revolution with MX1 Computational Memory Chip
Overview
XCENA has secured a $135 million Series B funding round, marking a pivotal moment that highlights its growth and strong investor confidence. The company is tackling the AI bottleneck by introducing the MX1 computational memory chip, which integrates advanced CXL technology and Near-Data Processing to move computation closer to where data resides. This innovation reduces data movement between CPU and memory, significantly improving efficiency for AI inference and analytics. As AI drives unprecedented demand for high-performance memory, XCENA’s approach offers a timely solution to the memory crisis, positioning the company as a key player in the evolving AI infrastructure landscape.