cd /news/ai-infrastructure/marvell-adds-six-memory-channels-to-… · home › topics › ai-infrastructure › article
[ARTICLE · art-149709] src=runtimewire.com ↗ pub= topic=ai-infrastructure verified=true sentiment=· neutral

Marvell adds six memory channels to CXL expanders for AI servers

Marvell announced the Structera X 3406 and 3506 CXL memory expander controllers, which add six memory channels and support up to 9TB of DDR4 or 6TB of DDR5 for AI servers, with sampling expected to begin in Q1 2027. The controllers are designed to expand memory capacity and share it across multiple hosts, extending Marvell's data-center silicon line into CXL memory expansion. Marvell's capacity and inference gains remain company claims, and customer adoption and workload-specific results will determine whether the design translates into meaningful business.

read1 min views3 publishedOct 12, 2026
Marvell adds six memory channels to CXL expanders for AI servers
Image: Runtimewire (auto-discovered)

Marvell's new controllers support up to 9TB of DDR4 or 6TB of DDR5, but are expected to begin sampling in Q1 2027.

        By [RuntimeWire Staff](https://runtimewire.com/author/runtimewire-staff)
        · Published 

Primary source: [Marvell](https://www.marvell.com/company/newsroom/marvell-structera-x-cxl-3-2-memory-expanders-ai-inference.html)

Why it matters #

The launch extends Marvell's bet on purpose-built data-center silicon into CXL memory expansion. Its headline capacity and inference gains remain company claims, and the chips are not expected to sample until Q1 2027, so customer adoption and workload-specific results will determine whether the design translates into a meaningful business.

Marvell is adding six memory channels to its CXL expanders for AI servers. Sehat Sutardja, Weili Dai and Pantas Sutardja founded Marvell in 1995 around specialized silicon for hard drives. The new Structera X 3406 and 3506 controllers are designed to expand memory capacity and share it across multiple hosts; Marvell expects them to begin sampling in Q1 2027, rather than shipping broadly today.…

── more in #ai-infrastructure 4 stories · sorted by recency
── more on @marvell 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/marvell-adds-six-mem…] indexed:0 read:1min 2026-10-12 · —