{"slug": "gpu-compute-tightness-index", "title": "GPU Compute Tightness Index", "summary": "The GPU Compute Tightness Index fell to 43.8/100 as of June 29, 2026, entering Loose territory after a monthly drop of 13.7 points, indicating abundant idle AI infrastructure and supply exceeding near-term demand. Per-GPU readings show H100 at 37, B200 at 47, and H200 at 57, with spot discounts of 55%, 47%, and 31% below on-demand pricing.", "body_md": "The Compute Tightness Index sank to **43.8/100** as of **2026-06-29**, pushing the blended gauge into Loose territory after a **-13.7** pts monthly drop that pulled it under the prior **44–64** range. Per-GPU readings of **37** for H100, **47** for B200, and **57** for H200 correspond to spot discounts of **55%**, **47%**, and **31%** below on-demand, signaling abundant idle AI infrastructure and supply that currently exceeds near-term training and inference demand. It is the input-side companion to the [Token Demand Index](/token-demand) (the *output* — tokens consumed).\n\n← looser · 45 balanced · 60 tight · tighter →\n\nAs of **2026-06-29** the blended index sits at **43.8** (**Loose**) and has been **loosening** over the past 30 days (-13.7 pts). Listing-weighted across H100, H200 and B200 by on-demand market depth.\n\n| GPU | On-demand $/hr | Spot $/hr | Spot discount | Listings | CTI | 30d |\n|---|---|---|---|---|---|---|\nB200 | $6.42 | $3.43 | 47% | 46 | 46.6 | ▼ -5.9 |\nH200 | $4.13 | $2.84 | 31% | 69 | 56.9 | ▼ -20.9 |\nH100 | $3.62 | $1.64 | 55% | 145 | 36.8 | ▼ -13.5 |\nMI355Xthin · excl. from blend | $8.60 | $0.00 | 100% | 2 | — |\n\nSpot discount = how far below on-demand the spot rate trades. A **shrinking** discount (rising CTI) means spare capacity is drying up; a **widening** discount (falling CTI) means idle GPUs are being dumped on spot. MI355X is tracked for coverage but excluded from the blended index — its listing count is too thin to weight reliably.\n\nFor each GPU on each day: `CTI = 100 × ( 0.8 · spot_avg/on_demand_avg + 0.2 · unavailable/(unavailable + on_demand_listings) )`\n\n. The scarcity term is the share of listed capacity that has gone dark (clamped to 1). The blended index is the listing-weighted mean across B200, H200 and H100 (weighted by on-demand listing count, a proxy for market depth). Inputs are the daily GetDeploying aggregate of public GPU rental listings — on-demand, reserved and spot averages plus availability — one snapshot per day.\n\n**Reading it:** <45 loose · 45–60 balanced · 60–80 tight · >80 acute scarcity. The spot-premium term dominates day to day; the availability term only bites when providers actually go dark, so it spikes the index during genuine capacity crunches. Young series (data since 2026-05-07); thresholds will calibrate as history accrues. Full daily history is available as [CSV](/compute-tightness.csv).\n\nThis is the **input** side of the AI-compute economy. For the **output** side, see the [Token Demand Index](/token-demand) (tokens consumed + the effective price of inference). Want the underlying rows or to query from an agent? Grab the [CSV](/compute-tightness.csv), or use the [Bargo MCP endpoint](/mcp) (`get_spot_pricing`\n\n). See also the [research notes](/research) and [Congress trades](/politicians).", "url": "https://wpnews.pro/news/gpu-compute-tightness-index", "canonical_source": "https://www.bargo.ai/compute-tightness", "published_at": "2026-06-30 22:07:07+00:00", "updated_at": "2026-06-30 22:20:21.816203+00:00", "lang": "en", "topics": ["ai-infrastructure", "ai-chips", "ai-research"], "entities": ["H100", "B200", "H200", "MI355X"], "alternates": {"html": "https://wpnews.pro/news/gpu-compute-tightness-index", "markdown": "https://wpnews.pro/news/gpu-compute-tightness-index.md", "text": "https://wpnews.pro/news/gpu-compute-tightness-index.txt", "jsonld": "https://wpnews.pro/news/gpu-compute-tightness-index.jsonld"}}