Forbes October 28, 2024
The pace of AI continues to be relentlessly fueled by new software, hardware, and learning paradigms, and it remains a challenge to meet the required exponential increases in performance. Gone are the days when a 10x to 20x generational increase is enough to meet increasing performance needs. AI requires 100x to 1000x performance increases from generation to generation, especially as we move from traditional AI to generative AI. And this doesn’t even account for what will be required for Artificial General Intelligence (AGI). The problem is compounded with the slowing of Moore’s Law. Increases in hardware performance are now driving power and thermal energy/heat to new levels. These challenges are changing how the industry thinks about and designs solutions, from...
Today's Sponsors
Today's Sponsor