Val Bercovici

Apr 27, 2026

• "you're able to get out of the same CapEx and OpEx, the same GPUs and energy costs 6.5 times more, so 550% more tokens"

  cited in Developer · from NVIDIA GTC 2026 | Val Bercovici, WEKA & Daniel Kearney, Firmus

• "March 2026 is a fundamentally different 180 degree different conversation where there's huge demand... placing acute pressure on this thing called KV cache, which is very, very memory centric."

  cited in Deep Tech · from NVIDIA GTC 2026 | Val Bercovici, WEKA & Daniel Kearney, Firmus

• "you're able to get out of the same CapEx and OpEx, the same GPUs and energy costs 6.5 times more, so 550% more tokens. So it's as if in a macro scenario, you just created five and a half new data centers out of thin air to serve agents"

  cited in Deep Tech · from NVIDIA GTC 2026 | Val Bercovici, WEKA & Daniel Kearney, Firmus

• "you're able to get out of the same CapEx and OpEx, the same GPUs and energy costs 6.5 times more, so 550% more tokens. So it's as if in a macro scenario, you just created five and a half new data centers out of thin air"

  cited in C-Suite · from NVIDIA GTC 2026 | Val Bercovici, WEKA & Daniel Kearney, Firmus