BREAKING
NVIDIA Bets on 'Intelligence per Dollar'
Why Post-Training Is Continuous
1Pre-train fluency
2RL post-train skills
3Rollout = inference
4Repeat continuously
0PFLOPS
FP4 compute
0GB
HBM4
0TB/s
NVLink 6
Rubin vs Blackwell Gains
Inference5
Training3.5
NVLink BW2
0B
parameters
0B
active params
0%
SWE-bench
Verdict Waits on Q4 2026
AI NEWS BLITZ
NVIDIA is pitching its Vera Rubin platform around cost-per-token for AI agents.