EagleCorp benchmarks Nemotron 3.5 Super VL at 2.74x vanilla single-request throughput
NVIDIAAI · x · 2026-10-29
NVIDIA shared EagleCorp's SPEED-Bench benchmarks of Nemotron 3.5 Super VL (4x B300, TP4, T=0):
- Single request: 647 tokens/s vs 236 vanilla, a 2.74x speedup
- At 512 concurrent requests: 33,480 vs 19,823 aggregate tokens/s, a 1.69x speedup
The numbers show major serving-stack throughput gains over the vanilla model, with the largest benefit in latency-sensitive single-request workloads.
More from Infra
- Mirendil Launches Lor, Agents That Run Hundreds of Parallel AI R&D Experiments — shayan_ · 2026-10-29
- Anthropic's S-1 inference economics changes Modal CEO's view on GPU prices — firstadopter · 2026-10-29
- Supabase cofounders on the Latent Space podcast: $0 to $10B, and why agents love it — Latent Space · 2026-10-29
- Agents push compute back to CPUs: $221B server CPU market seen by 2030 — BenBajarin · 2026-10-29
- AI server revenue forecast to hit $920B in 2027, pushing market past $1T — Beth_Kindig · 2026-10-29
- Oracle deploys Google TPUs for the first time and scales Vera Rubin NVL72 clusters to multi-gigawatt scale with Nvidia — Beth_Kindig · 2026-10-29