SemiAnalysis posted data from 174,264 agentic coding sessions showing 42% of runtime on CPU tasks versus 58% on GPU inference and highlighted cloud pricing mismatches with per-token monetization
Median per-turn time measured 5.13 seconds.
#687Bojan Tunguz@TUNGUZVery important.
SemiAnalysis@SemiAnalysis_FACT ALERT 🚨 : In modern agentic coding, 42% of the time is spent on CPU doing tool use such as editing files, running Bash scripts, running lints, etc. The economy of traditional cloud computing charges at $ per cpu core. In the economy of agents, the business model is $ per token thus to increase token revenue, you need to increase the amount of CPUs power u have so that you can generate your tokens.
#687Bojan Tunguz@TUNGUZHere is one big reason why this matters. Time spent on non-LLM inference tasks is only going to increase. However, tools that these AI system use are *very* inefficient and have been built from the ground up for CPU and human use. There is a huge untapped opportunity there to significantly improve those processes with AI agents in mind from the ground up.
SemiAnalysis@SemiAnalysis_FACT ALERT 🚨 : In modern agentic coding, 42% of the time is spent on CPU doing tool use such as editing files, running Bash scripts, running lints, etc. The economy of traditional cloud computing charges at $ per cpu core. In the economy of agents, the business model is $ per token thus to increase token revenue, you need to increase the amount of CPUs power u have so that you can generate your tokens.
#687Bojan Tunguz@TUNGUZHere is one big reason why this matters. Time spent on non-LLM inference time is only going to increase. However, tools that these AI system use are *very* inefficient and have been built from the ground up for CPU and human use. There is a huge untapped opportunity there to significantly improve those processes with AI agents in mind from the ground up.
SemiAnalysis@SemiAnalysis_FACT ALERT 🚨 : In modern agentic coding, 42% of the time is spent on CPU doing tool use such as editing files, running Bash scripts, running lints, etc. The economy of traditional cloud computing charges at $ per cpu core. In the economy of agents, the business model is $ per token thus to increase token revenue, you need to increase the amount of CPUs power u have so that you can generate your tokens.