An SK hynix research team published a paper at the 2025 IEEE International Electron Devices Meeting (IEDM) proposing co-optimization of cell, non-cell, and page schemes to improve the energy efficiency of analog compute-in-memory (A-CiM) based on 3D FeNAND.
Memory supply tightness will persist through 2027, with DRAM lead times at 20 weeks versus a balanced level of 8 weeks; vendors are shifting PC DRAM capacity to server DRAM, and HDD lead times remain at 50 weeks.
TrendForce's weekly report shows DRAM lead times have reached 20 weeks (versus a balanced level of 8 weeks), and the shortage is limiting the ability of some NAND vendors to support enterprise SSD solutions.
TrendForce has observed that some cloud service providers are adding QLC SSDs to SSD PODs, replacing the default TLC SSD plus HDD configuration. In AI inference, the Prefill and Decode stages repeatedly read and write KV Cache, and as context windows and batch sizes grow, KV Cache capacity needs multiply, prompting vendors to move caching from costly, capacity-limited HBM down to cheaper tiers.
NAND suppliers are shifting capacity toward enterprise SSDs on recovering demand, but eSSD lead times still stretch to 16 weeks, far above the 8 weeks of a balanced market, and overall supply remains insufficient. Other highlights this week include reports that SK hynix is in early talks with Intel about producing memory chips in the US, and Huawei moving its Ascend 960DT up three quarters to early 2027 to catch up with NVIDIA.
At FMS 2026, Marvell unveiled a new-generation AI memory infrastructure portfolio spanning server-level AI storage, rack-level CXL memory expansion and pooling, and cabinet-level optical shared memory.
As test-time scaling rises, the AI industry's focus is shifting from training to inference, bringing structural changes to the memory hierarchy. In the first half of 2026, NVIDIA and Google respectively launched CMX, Rubin integrated Groq LPX, and the TPU v8i designed for inference (with 72 GB more HBM3e and 384 MB SRAM than the TPU v8t).