AI/ML
How to Overcome the Pain Points of AI/ML Hardware Design
Join Achronix for a live Webinar December 16th: 10-11 AM Pacific and Recorded On-Demand After the Event AI/ML hardware faces three common pain points: memory bandwidth, computational throughput and on-chip data movement. Next-generation FPGA technology includes a 2D network on chip, GDDR6 memory interfaces and high performance machine learning processors, which present new capabilities to… How to Overcome the Pain Points of AI/ML Hardware Design
Memory Bandwidth Races Higher with HBM3
With the formal release of the HBM3 specification, memory bandwidth for AI/ML and HPC shifts to a higher gear. Terabytes of bandwidth are possible using HBM3’s 2.5D/3D architecture. Join memory expert Frank Ferro as he discusses what changes come with the new generation of HBM, and how the Rambus HBM3 memory subsystem can help designers… Memory Bandwidth Races Higher with HBM3