DRAM shortage to persist until 2027; NVIDIA reassesses Rubin Ultra HBM configuration

iconKuCoinFlash
Share
AI summary iconSummary
TrendForce reports that DRAM shortages will extend into 2027, with HBM4e certification still uncertain. NVIDIA is reconsidering the HBM layout for Rubin Ultra, shifting from a 12-layer HBM4e to alternatives such as an 8-layer HBM4e or 12-layer HBM4. The daily market report highlights that wafer capacity and yield issues are key factors. HBM4e production delays could limit I/O speeds to 11–12 Gbps instead of 14–16 Gbps. Altcoins to watch may respond to shifts in GPU supply. HBM bit shipments are expected to increase by 50–60% in 2027 but will still fall short of demand.

Odaily Planet Daily reports: Analyst Jukan from Citrini posted on X that, according to TrendForce’s latest storage industry research, DRAM supply shortages are expected to persist until 2027, and the certification timeline for HBM4e remains uncertain; NVIDIA has been reassessing the HBM configuration for Rubin Ultra since the third quarter of 2026.

The original plan was to use a 12-layer stacked HBM4e configuration, but NVIDIA is currently evaluating multiple designs in parallel, including 8-layer stacked HBM4e, 12-layer stacked HBM4, and 8-layer stacked HBM4, with the final specifications yet to be confirmed. In addition to NVIDIA, certain CSPs are reportedly also considering reducing the HBM capacity for their next-generation custom ASICs.

NVIDIA has designated the 12-layer stacked HBM4e as the baseline design for Rubin Ultra during the period from 2025 to the first half of 2026. However, beginning in the third quarter of 2026, NVIDIA began reviewing lower-specification alternatives. TrendForce attributes this shift to two primary supply-side constraints: first, an anticipated overall DRAM shortage in 2027 will limit the wafer capacity that memory manufacturers can allocate to HBM production; second, the certification timeline for the 12-layer stacked HBM4e and the pace of yield improvement remain uncertain.

TrendForce states that NVIDIA's primary focus for the Rubin Ultra generation is increasing I/O speed, with expanding GPU shipment volumes as a secondary priority. If NVIDIA ultimately decides to reduce HBM specifications, it is expected to achieve this by decreasing the number of DRAM stacking layers. Whether HBM4e can be certified and enter mass production as planned will determine whether Rubin Ultra’s I/O speed can increase from the previous generation’s 8 to 11.7 Gbps to 14 to 16 Gbps, or remain at 11 to 12 Gbps through optimized HBM4 design. Within the same product generation, the number of DRAM stacking layers dictates the trade-off between HBM capacity per GPU and the number of GPUs that can be shipped.

TrendForce also believes that the final configuration will depend on memory manufacturers' wafer allocation decisions. Regarding supply and demand, HBM bit shipments are expected to increase by 50% to 60% year-over-year in 2027, but are still projected to fall short of demand growth. With supply constraints continuing, HBM suppliers are expected to maintain pricing power throughout 2027. The industry widely anticipates significant HBM price increases, leaving AI chip manufacturers under dual pressure from limited HBM supply and rising procurement costs, further reinforcing the incentive to adopt lower HBM capacity configurations.

Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information. Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.