LPDDR6 will become the next major component addressing the rising memory requirements in Agentic AI datacenters.
LPDDR6 Is The Leading Memory Standard For AI Datacenters Moving Forward, Expect A Sharp Increase In Demand
As Agentic AI workloads continue to gain adoption in an aggressive way, the demand for more memory capacity is rising across AI firms. This has driven demand for LPDDR memory, which has become the gold standard for data centers owing to its higher memory capacities and better efficiency characteristics. The next chapter in LPDDR memory is going to be LPDDR6.
Recently, we reported how Samsung, SK Hynix, and Micron were aiming for the commercialization of LPDDR6 by 2028-2029, though given the Agentic AI demand, this may happen sooner. JEDEC has also previewed the initial paper for LPDDR6, citing 512 GB capacities and SOCAMM2 modules, clearing the path for next-gen AI servers.
Speaking to EETimes, JEDEC said that datacenter customers are interested in LPDDR, and LPDDR6 will bring several new features that address their needs. Some of the notable changes are a narrower x6 per-die interface, which will lead to higher capacities. The bandwidth jump won't be as significant, stating 10-20% improvement over the current generation, but higher densities matter more in these servers.
Nagashima said the goal is to reach 512GB density, which is beyond the current LPDDR5/5X maximum, to address the ever-growing memory capacity requirements of AI training and inference workloads.
via EETimes
In that regard, the non-binary interface and the additional x6 sub-channel will enable DRAM makers to add higher capacities per integrated circuit. The goal is to reach 512 GB, which will double the current SOCAMM2 LPDDR5X capacities of 256 GB. LPDDR6 will also come in SOCAMM2 modules, the go-to choice for AI datacenters. NVIDIA has already adopted LPDDR5X SOCAMM2 for its Vera CPUs, & AMD has also announced LPDDR5X for its inference-optimized Verano CPUs.
LPDDR6, as well as SOCAMM, has caught the attention of data center customers as power consumption becomes an increasingly important metric. “AI and data centers are really driving—or straining—the ecosystem right now.”
via EETimes
JEDEC said that it is also nearing the completion of its LPDDR6-PIM (Programming In Memory) tech, which further complements the LPDDR6 roadmap. The PIM controller will let the memory perform calculations within the memory instead of just storing data, offloading some of the processing from the CPU directly to the memory. And in all of this, LPDDR standards are known for their better power characteristics that will lead to more efficient datacenters.
LPDDR6 is poised to become the dominant memory standard for next-generation AI datacenters. As Agentic AI drives explosive demand for higher capacity and efficiency, LPDDR6 delivers exactly what the industry needs: up to 512GB densities, superior power characteristics, and innovative features like Processing-in-Memory (PIM). With major players like NVIDIA and AMD already embracing the LPDDR ecosystem, LPDDR6 is set to power the future of AI infrastructure more efficiently than ever. The memory revolution for Agentic AI is well underway.
About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.
Follow Wccftech on Google to get more of our news coverage in your feeds.
