AI Silicon
Semiconductor intelligence — ai silicon coverage from Futurum Advisory
NVIDIA-Marvell NVLink Fusion Partnership Reshapes the Custom Silicon Landscape
NVIDIA and Marvell’s NVLink Fusion partnership marks a pivotal shift in the custom silicon ecosystem, enabling hyperscalers to integrate their own XPUs directly into NVLink fabrics — a move that fundamentally changes the competitive dynamics for custom ASIC vendors.
analysis · Apr 2, 2026 · 10 min
Will NVIDIA Investment Accelerate Marvell’s XPU Growth?
NVIDIA’s strategic investment in Marvell raises critical questions about the future of the XPU market and whether this relationship accelerates or constrains Marvell’s custom silicon ambitions.
analysis · Apr 2, 2026 · 10 min
GPU Alternatives Poised to Outgrow GPUs in 2026 as Enterprises Ramp XPU Spending
New Futurum Research data shows enterprise XPU adoption accelerating sharply, with alternative GPU architectures on track to capture a larger share of AI infrastructure spend in 2026.
analysis · Jan 26, 2026 · 12 min
At CES, NVIDIA and AMD Made Memory the Future of AI
Both NVIDIA and AMD used CES 2026 to position memory architecture — HBM3E, CXL, and next-generation packaging — as the critical differentiator for AI workload performance, not just raw compute.
analysis · Jan 12, 2026 · 12 min