Media Coverage
The AI Infrastructure Bottleneck No One Is Solving
October 20, 2026
James Spall
Every infrastructure roadmap for generative AI reads the same way: more GPUs, more high-bandwidth memory, faster interconnects, better token throughput. These investments matter, but specialization is only currently targeting one half of the inference pipeline. The other half, the stage where a model actually reads the prompt before it writes a word, still has no purpose-built hardware in production...
.jpg)
Light Over Silicon: Why Optical Compute Is AI Infrastructure's Next Chapter
AI's energy crisis can't be solved with more silicon. Lumai's Head of Engineering explains why optical compute is the only path forward — and how Iris Nova delivers up to 10× lower energy per inference today.
Why AI Inference Hardware Must Specialize - Starting with Prefill
The prefill and decode disaggregation debate is settled. Now the question is hardware specialisation. Explore why prefill demands purpose-built compute — and what that means for the future of AI inference infrastructure.
Why Disaggregated Inference Needs Two Different Solutions
Disaggregation solves one problem and exposes another. Separating prefill and decode into independent compute pools eliminates the resource contention that plagues coupled architectures, lets each pool scale independently, and allows you to tune batch sizes and parallelism strategies for the actual workload profile rather than a compromise between two conflicting ones.
