Flopper.io Flopper.io

NVIDIA Tesla K80 24GB vs NVIDIA Tesla P4 8GB

The Tesla K80 and the Tesla P4 deliver near-identical FP32 throughput (5.6 vs 5.5 TFLOPS dense). The Tesla K80 also carries 16 GB more memory (24 GB vs 8 GB).

Tesla K80: Kepler, 2014 Tesla P4: Pascal, 2016
Memory
24 vs 8 GB
16 GB more for the Tesla K80
Bandwidth
480 GB/s vs 192 GB/s
Tesla K80 moves data faster
TDP
300 W vs 75 W
Tesla P4 draws 225 W less
Download comparison sheet

Share this page

NVIDIA Tesla K80 24GB vs NVIDIA Tesla P4 8GB

A post with this link shows the card on the Image tab as its preview.

Post on X Share on LinkedIn
 Printable version →

NVIDIA Tesla K80 24GB

Full specs →

What is the NVIDIA Tesla K80?

The Tesla K80 is a GPU from NVIDIA on the Kepler architecture, an accelerator for AI training, inference and high-performance computing. It is a PCIe add-in card, so it fits standard servers and workstations. Tesla K80 full specifications →

Can I rent the Tesla K80 in the cloud?

Flopper is not tracking any live rental listings for the Tesla K80 yet. The provider directory lists the clouds that rent comparable accelerators on demand. GPU cloud providers →

How does the Tesla K80 compare with other accelerators?

This page sets the Tesla K80 against the Tesla P4. The Tesla K80 is the older generation of the two. Both are NVIDIA parts, and both ship as a PCIe card, so they fit the same class of system. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the Tesla K80 figures come from?

Every Tesla K80 figure on this page traces to NVIDIA's own datasheet or specification table. How Flopper verifies specifications →

NVIDIA Tesla P4 8GB

Full specs →

What is the NVIDIA Tesla P4?

The Tesla P4 is a GPU from NVIDIA on the Pascal architecture, an accelerator for AI training, inference and high-performance computing. It is a PCIe add-in card, so it fits standard servers and workstations. Tesla P4 full specifications →

Can I rent the Tesla P4 in the cloud?

Flopper is not tracking any live rental listings for the Tesla P4 yet. The provider directory lists the clouds that rent comparable accelerators on demand. GPU cloud providers →

How does the Tesla P4 compare with other accelerators?

This page sets the Tesla P4 against the Tesla K80. The Tesla P4 is the newer generation of the two. Both are NVIDIA parts, and both ship as a PCIe card, so they fit the same class of system. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the Tesla P4 figures come from?

Every Tesla P4 figure on this page traces to NVIDIA's own datasheet or specification table. How Flopper verifies specifications →

Specifications

Tesla K80
Tesla P4
Architecture
KeplerPascal
Launch Year
20142016
Form Factor
PCIePCIe
Memory
24 GB8 GB
Memory Bandwidth
480 GB/s192 GB/s
TDP
300 W75 W
Process Node
28nm16nm

Performance (TFLOPS)

Tesla K80
Tesla P4
1.9 TFLOPS No verified data available
5.6 TFLOPS 5.5 TFLOPS
No verified data available No verified data available
No verified data available No verified data available
No verified data available No verified data available
No verified data available No verified data available
FP6
No verified data available No verified data available
No verified data available No verified data available
No verified data available22 TOPS

Figures are dense. Where a vendor publishes a figure it sits under the dense one and never decides a winner. Hover any precision for its definition, or read the GPU glossary.

FLOPS by Precision

What actually differs

The Tesla P4 is the newer part: Pascal (2016) against the Tesla K80's Kepler (2014).

The Tesla K80 carries 16 GB more memory (24 GB against 8 GB), so larger models fit on one GPU before you shard. Bandwidth is 480 GB/s against 192 GB/s, which sets the ceiling on decode speed.

Power is 300 W against 75 W, and at FP32 the Tesla P4 does more work per watt. Weight the FP8 and FP16 rows for LLM work and the FP64 row for HPC.

New to the terms? The glossary explains FP64, FP32, INT8, Sparsity and TFLOPS. For how the generations relate, see NVIDIA architectures explained.

Frequently asked questions

What is better, the Tesla K80 or the Tesla P4?

They are close on paper: both deliver about 5.6 TFLOPS of dense FP32 throughput. Memory, bandwidth, and power are the deciding differences.

Which should you choose, the Tesla K80 or the Tesla P4?

On raw FP32 throughput the Tesla K80 and the Tesla P4 are level, so the decision rests on memory, power and price. Choose the Tesla K80 when the model has to fit on one GPU: it carries 16 GB more memory. The Tesla P4 draws 225 W less power, which matters in an air-cooled or power-capped rack.

Which has more memory, the Tesla K80 or the Tesla P4?

The Tesla K80 carries 24 GB of memory versus 8 GB for the Tesla P4. Memory bandwidth is 480 GB/s for the Tesla K80 and 192 GB/s for the Tesla P4.

How much power do the Tesla K80 and the Tesla P4 draw?

The Tesla K80 is rated at 300 W TDP and the Tesla P4 at 75 W. On FP32 throughput per watt, the Tesla P4 is the more efficient part.

Get Comparison Updates

New GPUs added weekly. Be the first to see how they compare.

Open in Full Compare Tool → Tesla K80 full specs Tesla P4 full specs Estimate cluster cost in the calculator →

© 2026 Flopper.io - Compare the GPUs Powering AI