COMPARE / P100 VS V100

P100 vs V100

Specs from the chip makers' datasheets; prices from the live Compute Cafe panel, refreshed hourly. Every rate is an on-demand published price with a source receipt.

CHEAPEST P100
$0.045/hr
QuickPod · median $0.10 · 2 providers
CHEAPEST V100
$0.170/hr
Akash · median $3.70 · 11 providers
CHEAPER RIGHT NOW
P100
live panel, 2026-09-27
P100V100
MAKERNVIDIANVIDIA
ARCHITECTUREPascalVolta
LAUNCHED20162017
MEMORY16 GB HBM216 or 32 GB HBM2
MEMORY BANDWIDTH732 GB/s900 GB/s
TENSOR COMPUTE18.7-21.2 TFLOPS FP16 (first-generation FP16 support; no tensor cores)125 TFLOPS FP16 tensor (SXM2)
POWER DRAW250-300 W300 W (SXM2) / 250 W (PCIe)
INTERCONNECTNVLink 1.0 (SXM2) or PCIe 3.0NVLink 2.0 300 GB/s (SXM2) or PCIe 3.0

HOW TO CHOOSE

P100 vs V100
The V100 adds tensor cores and doubles everything that matters for usually a small premium. The P100 is only for code that specifically needs Pascal.
Workload fit
P100: legacy cuda and hpc, teaching and coursework, modern llm work. V100: budget training, batch inference, legacy cuda stacks.

P100 VS V100 - FAQ

Which is cheaper to rent, P100 or V100?
By today's published on-demand rates: P100 starts at $0.045/GPU/hr (median $0.10, 2 providers) and V100 starts at $0.170/GPU/hr (median $3.70, 11 providers). The cheapest live listing is the P100 at QuickPod. Figures recompute from the panel on every refresh.
P100 or V100 - which should I rent?
P100: The P100 was the first deep-learning datacenter GPU - 2016's Pascal flagship, the first with FP16 and HBM2. V100: The V100 was the first GPU with tensor cores and dominated deep-learning training from 2017 to 2020. The side-by-side spec table and live price rows above give the full picture.
Where can I rent P100 and V100?
2 providers publish P100 rates and 11 publish V100 rates on Compute Cafe. Every row links to the provider's own published price - committed-use and reserved rates are excluded.

Full pricing: P100 · V100 · all comparisons