Two Nvidia data center GPUs, one Grace Blackwell and one Pascal: 186 GB HBM3e against 24 GB GDDR5, 23x the memory bandwidth on the GB200, and NVLink 1,800 GB/s on the GB200 where the P40 is PCIe only.

At a glance

Updated 12 hours ago
Memory
186 GB
HBM3e
Memory bandwidth
8.00 TB/s
From
$10.50 /GPU/hr
8 providers
Memory
24 GB
GDDR5
Memory bandwidth
346 GB/s
From
--

Price comparison

Cheapest listed price per GPU per hour by billing type, and the median across providers.

Billing GB200 P40 Difference
On-demand $10.50 --
Spot $27.04 No spot listings
Reserved $10.58 No reserved listings
Median $16.00 --

Median: Median across providers, one vote each, so a large catalog does not outweigh a small one.

Price history

GB200 vs P40 price history

Weekly median on-demand price per GPU per hour

Last 13 months
Aggregating historical prices...

Specs comparison

Spec GB200 P40 Difference
Memory
Capacity 186 GB HBM3e 24 GB GDDR5 7.8x more on GB200
Bandwidth 8,000 GB/s 346 GB/s 23x higher on GB200
Bus width 384-bit
Compute · dense
FP4 10,000 TFLOPS Not supported GB200 only
FP8 5,000 TFLOPS Not supported GB200 only
FP16 / BF16 2,500 TFLOPS
INT8 5,000 TOPS 47 TOPS 106x higher on GB200
FP32 80 TFLOPS 12 TFLOPS 6.7x higher on GB200
FP64 40 TFLOPS Not supported GB200 only
Platform
Interconnect NVLink 1,800 GB/s PCIe only NVLink on GB200 only
Board power 1,200 W 250 W 4.8x higher on GB200
Released Q1 2024 Q3 2016 8 years newer on GB200
Precision
Supported formats FP4FP6FP8FP16BF16TF32FP32FP64INT8 FP16FP32INT8

Compute: Vendor peak figures, dense and per GPU. Not measured throughput.

Source: GB200 datasheet and P40 datasheet.

Where to rent them

Providers listing these GPUs, those with both first, so you can switch without moving clouds.

Provider GB200 /GPU/hr P40 /GPU/hr
CoreWeave logo CoreWeave On-Demand $10.50 --
Oracle Cloud logo Oracle Cloud On-Demand $16.00 --
Microsoft Azure logo Azure On-Demand $27.04 --
Amazon Web Services logo AWS Reservation $10.58 --
CUDO Compute logo CUDO Custom on request --
Canopy Wave logo Canopy Wave Custom on request --
Nebius logo Nebius Custom on request --
Together AI logo Together AI Custom on request --

Similar GPUs

Other comparisons in the same class.

Our data for Nvidia GB200 was last updated on Sept. 7, 2026, and for Nvidia P40 on Sept. 7, 2026.