{
  "docId": "019dd923-605d-726d-81a5-622535f47574",
  "docSlug": "bi-091c3284a8780078",
  "documentTitle": "Spectral Compute raises $6 million to crack Nvidia's chip lock-in",
  "authorId": "Pitchdecks",
  "authorName": "Spectral Compute",
  "documentKindSlug": "pitchdeck",
  "documentKindLabel": "Pitch deck",
  "sourceTypeSlug": "strategy_consulting",
  "sourceTypeLabel": "Strategy consulting",
  "presentationDate": null,
  "orientation": "landscape",
  "aspectRatio": 1.776,
  "pageNumber": 11,
  "pageCount": 13,
  "prevPage": 10,
  "nextPage": 12,
  "slideType": "timeline",
  "function": "plan_implementation",
  "density": "balanced",
  "nDataPoints": 2,
  "notes": null,
  "elementsJson": [
    "headline_text",
    "callout_box"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-605d-726d-81a5-622535f47574/11",
  "deckHref": "/decks/019dd923-605d-726d-81a5-622535f47574",
  "deckJsonHref": "/decks/019dd923-605d-726d-81a5-622535f47574.json",
  "deckAnchorHref": "/decks/019dd923-605d-726d-81a5-622535f47574#slide-11",
  "components": [
    {
      "bbox": {
        "h": 0.7,
        "w": 0.42,
        "x": 0.53,
        "y": 0.2
      },
      "kind": "list",
      "text": "September 2026 – \"General Availability\". Features include: Performance optimisations, matrix multiply optimisations, language extensions. Supported projects include: PyTorch, TensorRT-LLM, DALI, cuGraph, GROMACS, AMGX, vLLM, Llama.cpp. Expected spend (Jan 2026 - Sep 2026): £1.15m (excl. marketing spend)",
      "attrs": null,
      "subkind": "timeline-item",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "447d95b9-1248-43b3-810d-23730677b6b5",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.7,
        "w": 0.42,
        "x": 0.05,
        "y": 0.2
      },
      "kind": "list",
      "text": "January 2026 – first widely-available commercialised release. Features include: Multi-GPU support, performance benchmarking, MFMA hardware support, math libraries optimisations, dynamic parallelism, cooperative groups, NVPTX support, CUDA mempools. Libraries include: CUTLASS (without tensormaps), cuBLASLt, cuFFT, CV-CUDA. Expected spend 2025: £1.077m (excl. marketing spend)",
      "attrs": null,
      "subkind": "timeline-item",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "9d7e19e8-d5a3-445a-b0db-3ed87dec2754",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.1,
        "w": 0.3,
        "x": 0.05,
        "y": 0.05
      },
      "kind": "title",
      "text": "Key Milestones",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "482f3492-e08b-4699-8327-dd389171cef8",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [],
  "frameworks": [
    {
      "name": "timeline",
      "slug": null,
      "matchId": "290843ab-97c9-4348-8efc-bf47de314379",
      "evidence": "Chronological presentation of product milestones",
      "confidence": 1
    }
  ],
  "arcBeats": [],
  "loops": [
    {
      "to": 13,
      "from": 11,
      "name": "Quick Win Big Bet",
      "slug": "47-quick-win-big-bet",
      "bestFor": "Transformation planning, 100-day plans, resource allocation",
      "matchId": "56218a5f-147a-4ade-8704-e6454a89ecb4",
      "evidence": "Slides 11-13 present key milestones and the investment opportunity.",
      "position": 1,
      "objective": "Can Spectral Compute achieve quick wins and make a big bet with SCALE?",
      "structure": "The Full List -> Quick Wins (Low effort, High impact) -> Big Bets (High effort, High impact) -> Sequenced Roadmap",
      "confidence": 0.6,
      "description": "Separate initiatives into immediate wins and longer-term strategic bets"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}