Open App →
Back to News Feed
Digitimes September 17, 2026 By DIGITIMES neutral

SK Hynix looks beyond HBM as inference workloads diversify

SK HynixAI / LLMNVIDIA / GPUMemory
SK Hynix is expanding its memory strategy beyond the conventional GPU-HBM pairing as inference workloads split between latency-sensitive services and capacity-heavy, long-context applications.
Read original article ↗

Related Articles

SK Hynix’s Intel Liaisons: What You Need to Know

The deal between Intel and SK Hynix seems imminent not because of technology business imperatives, but because of geopol

EE Times · September 17, 2026

SK Hynix weighs US memory options as buyers lock in 2027 supply

SK Hynix is keeping its US memory options open as customers move earlier to secure 2027 supply, with tightening availabi

Digitimes · September 17, 2026

SK Hynix reportedly eyes Intel Ohio plant for US memory production

SK Hynix is reportedly lining up US memory chip production and is in talks with Intel over possible cooperation, includi

Digitimes · September 17, 2026

SK Hynix shifts supply chain planning for AI capex boom

AI-driven capacity expansion is pushing up demand for semiconductor equipment, and SK Hynix is strengthening supply chai

Digitimes · September 17, 2026