AI & Tech

Marvell and Penguin Solutions aim to end the RAMpocalypse with photonic memory fabric

Celestial AI’s dream of high-bandwidth memory (HBM) at a distance will soon be a reality as part of a new collaboration between Marvell Technology and system integrator Penguin Solutions to build a photonic memory fabric.

✦ Extractive summary — built from the sources' own reporting•From 1 source•1 min read•Updated November 16, 2026, 2:00 PM UTC
Key points
  • Marvell and Penguin are positioning the new memory appliance as an alternative to storage-based key value (KV) cache offload.
  • It comprises a memory expansion module with two stacks of HBM3e totaling 72 GB along with support for eight channels of DDR5 stacked atop a silicon photonic interposer with 7.2 Tbps (900 GB/s) of bidirectional bandwidth.
  • Through a software layer, which we’re told was developed by Penguin, each system can access the entire shared memory space, similar to how systems connecting to network-attached storage would have access to the appliance's full capacity.

“What we provide in this appliance is what we call a Photonic Fabric memory module… We provide these modules to Penguin, and they build the box, do all the power management, do the software, and all the management layer,” Preet Virk, senior VP of Marvell’s photonic fabric biz, tells El Reg.

Additional details
  • Penguin’s first appliances based on the tech will support 16 of these memory modules for up to 16 TB of DDR5 and 1 TB of HBM, with future systems capable of scaling to twice the DDR5 capacity.
  • Marvell contends that by offloading KV caches from HBM directly to these memory appliances, customers can achieve better performance and higher memory utilization.
  • With 57.6 Tbps of bandwidth distributed via a Marvell-designed optical fiber shuffle, each appliance supports up to 16 servers either via a network interface card, or via co-packaged (CPO) or near-packaged optical (NPO) modules.
  • NPO is expected to play a major role in the next generation of rack-scale systems from Nvidia and AMD, as it provides better serviceability and lower blast radius than baking photonic chiplets directly into CPUs or GPUs.
  • For the most part, these follow typical CXL memory tiering conventions, with the exception of an HBM cache-through mode that Virk says will enable users to extract the performance of HBM with the capacity of DDR5.

How this summary was made: extractive summary — built from the sources' own reporting. It is drawn only from the reporting of the sources listed below — read the originals for full detail. How we summarize →

Advertisement

Every article behind this summary — 1 article from 1 outlet