Blended GPU compute costs remained stable 0.4% this week across major cloud providers, led by ALVEO_U30's 12.8% decline.
Top Movers
| GPU | Blended Price | WoW Change | Class |
|---|---|---|---|
| ALVEO_U30 | $0.04/hr | ▼ 12.8% | General |
| RTX PRO 6000 | $0.54/hr | ▼ 2.0% | General |
| VIRTEX_VU9P | $0.91/hr | ▲ 1.7% | General |
| VIRTEX_VU47P | $0.52/hr | ▼ 1.2% | General |
| V100 32GB | $1.57/hr | ▼ 1.1% | Training |
| B200 | $7.52/hr | ▲ 1.0% | Training |
| A100 40GB | $1.79/hr | ▼ 0.6% | Training |
Blended pricing = average of spot, on-demand, and 1-year reserved rates across major cloud providers.
Training vs Inference
Training-class GPU pricing held steady this week (avg $4.31/hr, -0.0% WoW), while inference-class pricing held steady (avg $0.52/hr, -0.0% WoW).
The training-to-inference price ratio stands at 8.3x — stable compared to last week. The elevated spread suggests strong demand for training compute relative to inference, consistent with ongoing large model training activity.
Regional Spotlight: North America
North America trades at a 9% discount to global averages this week, with 38 GPU types available across the region. The most expensive GPUs in the region are GB200 ($12.47/hr), GB300 ($12.32/hr), B300 ($10.31/hr). The 9% discount makes North America one of the more cost-effective regions for GPU deployment this week.
For detailed pricing data across all North America sub-regions, see the full regional profile.
Implications
For cloud buyers: Europe continues to offer the lowest average GPU pricing ($2.25/hr blended average). For workloads with regional flexibility, the gap between Europe and Middle East is $1.84/hr — a 82% premium. Compare regional pricing →
For semiconductor analysts: GPU pricing trends remain broadly stable this week. H100 (+0.3% WoW) and MI300X (-0.0% WoW) are tracking within normal ranges. Blackwell (B200) blended pricing at $7.52/hr (+1.0% WoW) provides an early read on next-generation adoption curves. View all GPU profiles →
For GPU investors: Stable pricing supports predictable returns for existing deployments. Model scenarios with the GPU ROI Calculator →