Nvidia K80 vs Nvidia L4

Two Nvidia data center GPUs, one Kepler and one Ada Lovelace: 12 GB GDDR5 against 24 GB GDDR6 and 25% more memory bandwidth on the L4.

At a glance

Updated 6 minutes ago

Nvidia K80

Nvidia L4

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

Billing	K80	L4	Difference
On-demand	$0.90	$0.32	2.8x more expensive on K80
Spot	No spot listings	$0.23
Reserved	$0.04	$0.32	7.1x more expensive on L4
Median	$0.90	$0.90	Within 5%

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

Weekly median on-demand price per GPU per hour

Last 12 months

Aggregating historical prices...

Spec	K80	L4	Difference
Memory
Capacity	12 GB GDDR5	24 GB GDDR6	2x more on L4
Bandwidth	240 GB/s	300 GB/s	1.2x higher on L4
Bus width	384-bit	192-bit	2x wider on K80
Compute · dense
FP8	Not supported	242 TFLOPS	L4 only
FP16 / BF16	Not supported	121 TFLOPS	L4 only
INT8	Not supported	242 TOPS	L4 only
FP32	4.4 TFLOPS	30.3 TFLOPS	6.9x higher on L4
FP64	1.5 TFLOPS	Not supported	K80 only
Platform
Interconnect	PCIe only	PCIe only
Board power	150 W	72 W	2.1x higher on K80
Released	Q4 2014	Q1 2023	8 years newer on L4
Precision
Supported formats	FP32FP64	FP8FP16BF16TF32FP32INT4INT8

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

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