Marvell extends CXL ecosystem leadership with Structera interoperability across major memory and CPU platforms

Marvell announced its Structera Compute Express Link memory-expansion controllers and near memory compute accelerators passed interoperability testing with DDR4 and DDR5 from Micron Technology, Samsung Electronics, and SK hynix. The company says this makes Structera the only CXL 2.0 product family validated across both major CPU architectures and all three memory suppliers.

Marvell Technology announced that its Structera Compute Express Link memory-expansion controllers and near memory compute accelerators have completed interoperability testing with DDR4 and DDR5 memory solutions from Micron Technology, Samsung Electronics, and SK hynix. The company framed the milestone as an extension of its leadership in data infrastructure semiconductor solutions, and noted the tests cover both major memory generations used in server and data-center designs.

This interoperability milestone follows a recent announcement that Structera achieved successful interoperability with AMD EPYC processors and 5th Gen Intel Xeon Scalable platforms. Combined, the memory and CPU validation makes Structera the only CXL 2.0 product family with completed interoperability testing across both leading CPU architectures and all three major memory suppliers, according to Marvell.

Marvell highlighted the practical implications of the validation for system builders. Validation with DDR4 and DDR5 is intended to enable scalable system design, reduce integration risk, and streamline qualification cycles. The company positioned those outcomes as especially important as data-centric applications grow in complexity and memory increasingly contributes to overall system performance. Marvell said the validated interoperability gives original equipment manufacturers and cloud providers flexibility to optimize system designs while maintaining supply chain flexibility.

72

Impact Score

Apertus: Swiss teams release fully open multilingual large language model

EPFL, ETH Zurich and the Swiss National Supercomputing Centre have released Apertus, a fully open multilingual large language model with its architecture, weights and training recipes published. The model is intended to support research, commercial adoption and public oversight of Artificial Intelligence.

32 GB of RAM could become the new standard for gamers

Steam´s hardware survey shows 32 GB of RAM rose to 36.46% of surveyed systems in August, up 1.31 percentage points from July and closing on 16 GB at 41.88%. Cheaper DDR5, broader OEM memory options, and local Artificial Intelligence and streaming workloads are cited as drivers.

Artificial intelligence sharpens humidity maps to improve forecasts

Researchers at Wrocław University of Environmental and Life Sciences used a super-resolution approach powered by Artificial Intelligence and NVIDIA GPUs to turn low-resolution GNSS snapshots into high-resolution 3D humidity maps, cutting retrieval errors in test regions.

Can an artificial intelligence doppelganger help me do my job

Digital clones combine hyperrealistic video, lifelike voice cloning, and conversational models to mimic a person. Startups promise that an artificial intelligence doppelgänger can scale personal interactions, but practical limits and safety concerns remain.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.