what david is referring to is that the energy spent (in this case with llm queries for optimization) eventually saves energy and compute later down the line (far more than it costs in some cases). this is a well known energy efficiency strategy for companies with high compute needs known as "GARG"
January 30, 2025 at 1:34 AM
what david is referring to is that the energy spent (in this case with llm queries for optimization) eventually saves energy and compute later down the line (far more than it costs in some cases). this is a well known energy efficiency strategy for companies with high compute needs known as "GARG"