Blended GPU compute costs rose 1.4% this week across major cloud providers, led by ALVEO_U30's 50.9% gain.
Top Movers
| GPU | Blended Price | WoW Change | Class |
|---|---|---|---|
| ALVEO_U30 | $0.04/hr | ▲ 50.9% | General |
| TPU_V5LITEPOD | $0.18/hr | ▲ 3.2% | General |
| V100 32GB | $1.60/hr | ▼ 3.0% | Training |
| V100 | $1.84/hr | ▲ 1.9% | Training |
| L4 | $0.51/hr | ▲ 1.3% | Inference |
| VIRTEX_VU9P | $0.89/hr | ▲ 1.1% | General |
| VIRTEX_VU47P | $0.54/hr | ▲ 0.6% | General |
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 $4.30/hr, -0.0% WoW), while inference-class pricing held steady (avg $0.51/hr, +0.2% WoW).
The training-to-inference price ratio stands at 8.4x — 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: Middle East
Middle East trades at a 20% premium to global averages this week, with 13 GPU types available across the region. The most expensive GPUs in the region are GB200 ($21.55/hr), H200 ($12.37/hr), H100 ($7.33/hr). The 20% regional premium reflects emerging infrastructure and limited provider competition.
For detailed pricing data across all Middle East sub-regions, see the full regional profile.
Implications
For cloud buyers: Europe continues to offer the lowest average GPU pricing ($1.98/hr blended average). For workloads with regional flexibility, the gap between Europe and Middle East is $2.06/hr — a 104% premium. Compare regional pricing →
For semiconductor analysts: GPU pricing trends remain broadly stable this week. H100 (-0.1% WoW) and MI300X (+0.0% WoW) are tracking within normal ranges. Blackwell (B200) blended pricing at $7.33/hr (+0.2% WoW) provides an early read on next-generation adoption curves. View all GPU profiles →
For GPU investors: Rising rental rates are improving returns for existing GPU deployments. Model scenarios with the GPU ROI Calculator →