SK hynix, the world's largest supplier of high-bandwidth memory, has published a technical roadmap for co-packaged optics (CPO) in Nature Electronics, signaling a strategic shift from competing on individual chip performance to developing integrated memory-and-optics systems.
This reflects the AI industry's broader move toward solving data-movement bottlenecks in server architecture, where optical connectivity integrated directly with memory chips is becoming central to next-generation AI infrastructure.
What happened
SK hynix, the world's largest supplier of high-bandwidth memory (HBM), has published a technical roadmap for co-packaged optics (CPO) in the journal Nature Electronics, positioning itself to compete on how memory and optical components work together as systems rather than on individual chip performance alone.
Why it matters
As AI computing architecture evolves beyond standalone chip improvements, memory makers are moving toward integrated systems that combine chips with optical connectivity. This shift reflects the industry's focus on solving bottlenecks in data movement between servers, where traditional electrical connections are becoming a limiting factor for large AI model training and inference.
What to watch
The roadmap's technical details in Nature Electronics will likely influence how other memory and chipmaker suppliers design their own system-level solutions, particularly for hyperscalers building the next generation of AI infrastructure.
Ask the AI about this article →
SK hynix's move into co-packaged optics marks a notable shift in competitive strategy for memory suppliers. Historically, companies in the memory space competed primarily on individual chip metrics—capacity, speed, power efficiency. The Nature Electronics publication signals that SK hynix recognizes a fundamental architectural change in how AI infrastructure is being built. Hyperscalers training and deploying large language models face severe constraints in moving data between GPU clusters; optical links integrated at the package level bypass some of those constraints. By publishing a technical roadmap, SK hynix is not only staking a claim in this emerging category but also establishing itself as a thought leader in system-level solutions rather than component-level optimization alone.
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytime
Ask AI anything about this article. Q&As are published on this page for other readers too.
Astromech, an AI startup co-founded by Ben Lamm and geneticist George Church, raised $20 million in funding le…
Alibaba is guiding its AI cloud revenue toward a US$10 billion run-rate in the next quarter, signaling that it…
Ode, a venture arm of Anthropic, has acquired Casper Studios to expand its enterprise artificial intelligence…
Alibaba Group's June quarter results show cloud AI revenue growing 45%, while capital expenditure (investment…
Anthropic launched Claude Academy on August 20, a free learning site that explains AI fundamentals and how to…
Marvell Technology disclosed an expanded custom chip partnership with Google that includes a warrant giving Go…
