Two Nvidia data center GPUs, one Blackwell Ultra and one Kepler: up to 288 GB HBM3e against 12 GB GDDR5, 33x the memory bandwidth on the B300, and NVLink 1,800 GB/s on the B300 where the K80 is PCIe only.
At a glance
Updated 4 minutes ago- Memory
- 288 GB
- HBM3e
- Memory bandwidth
- 8.00 TB/s
- From
- $6.50 /GPU/hr
- 31 providers
- Memory
- 12 GB
- GDDR5
- Memory bandwidth
- 240 GB/s
- From
- $0.90 /GPU/hr
- 2 providers
Price comparison
Cheapest listed price per GPU per hour by billing type, and the median across providers.
| Billing | B300 | K80 | Difference |
|---|---|---|---|
| On-demand | $6.50 | $0.90 | 7.2x more expensive on B300 |
| Spot | $4.30 | No spot listings | |
| Reserved | $4.25 | $0.04 | 95x more expensive on B300 |
| Median | $7.87 | $0.90 | 8.7x more expensive on B300 |
Median: Median across providers, one vote each, so a large catalog does not outweigh a small one.
Price history
B300 vs K80 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | B300 | K80 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 288 GB HBM3e | 12 GB GDDR5 | 24x more on B300 |
| Bandwidth | 8,000 GB/s | 240 GB/s | 33x higher on B300 |
| Bus width | 8,192-bit | 384-bit | 21x wider on B300 |
| Compute · dense | |||
| FP4 | 14,000 TFLOPS | Not supported | B300 only |
| FP8 | 4,500 TFLOPS | Not supported | B300 only |
| FP16 / BF16 | 2,250 TFLOPS | Not supported | B300 only |
| INT8 | 153.5 TOPS | Not supported | B300 only |
| FP32 | 75 TFLOPS | 4.4 TFLOPS | 17x higher on B300 |
| FP64 | 1.2 TFLOPS | 1.5 TFLOPS | 1.2x higher on K80 |
| Platform | |||
| Interconnect | NVLink 1,800 GB/s | PCIe only | NVLink on B300 only |
| Board power | 1,100 W | 150 W | 7.3x higher on B300 |
| Released | Q1 2025 | Q4 2014 | 10 years newer on B300 |
| Precision | |||
| Supported formats | FP4FP6FP8FP16BF16TF32FP32FP64INT8 | FP32FP64 | |
Compute: Vendor peak figures, dense and per GPU. Not measured throughput.
Source: B300 datasheet and K80 datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | B300 /GPU/hr | K80 /GPU/hr |
|---|---|---|
GPU.ai
|
On-Demand $7.89 | -- |
Runpod
|
On-Demand $7.89 | -- |
Vast.ai
|
On-Demand $8.75 | -- |
GPUaaS
|
Custom on request | -- |
Packet·ai
|
Custom on request | -- |
Impossible Cloud
|
Reserved (60mo) $4.32 | -- |
Canopy Wave
|
Custom on request | -- |
Bentaus
|
On-Demand $6.50 | -- |
AWS
|
-- | On-Demand $0.90 |
Database Mart
|
-- | Reserved (24mo) $0.0447 |
31 providers list the B300 and 2 the K80. Every listing is on the B300 page and the K80 page.
Similar GPUs
Other comparisons in the same class.
Our data for Nvidia B300 was last updated on Sept. 7, 2026, and for Nvidia K80 on Sept. 7, 2026.