@meta
  v: 1
  route: /compare/h100/h100-nvl-2x
  generated: 2026-06-10T10:34:29.575Z

@intent
  purpose:    Head-to-head comparison of H100 80GB vs 2× H100 NVL for local LLM inference.
  audience:   hardware-shopper, ai-engineer, self-hoster
  capability: compare_specs, compare_model_fits, drill_into_card

@state
  a{slug,name,vram_gb,bandwidth_gbs,fp16_tflops,budget_8k_gb}: h100,H100 80GB,80,3350,989,63
  b{slug,name,vram_gb,bandwidth_gbs,fp16_tflops,budget_8k_gb}: h100-nvl-2x,2× H100 NVL,188,3938,1979,151

@actions
  - id: view_a
    method: GET
    href: /gpu/h100
  - id: view_b
    method: GET
    href: /gpu/h100-nvl-2x

@context
  > Side-by-side comparison: H100 80GB (80GB) vs 2× H100 NVL (188GB). Includes specs delta and fit difference across the 30-model catalog.

@nav
  self:      /compare/h100/h100-nvl-2x
  parents:   [/, /gpu/]
  peers:     [/gpu/h100, /gpu/h100-nvl-2x]