What drives federated learning communication cost?

Cost = rounds × clients × model_size / compression. A 100MB model with 1000 clients over 100 rounds = 10TB without compression.

How does gradient compression help?

Sparsification or quantization can reduce gradient size by 10-100x at the cost of slightly slower convergence.

Federated Learning Communication Cost

Estimate total communication rounds and bandwidth needed for federated model training given model size, client count, and compression ratio.

Networking

Algorithms

Binary & Number

Systems

Dev Metrics

IP Subnet Calculator

IP Address

CIDR Prefix

Network

192.168.1.0

Broadcast

192.168.1.255

Subnet Mask

255.255.255.0

First Host

192.168.1.1

Last Host

192.168.1.254

Usable Hosts

254

Binary breakdown:

IP: 11000000.10101000.00000001.00000000

Mask: 11111111.11111111.11111111.00000000

Net: 11000000.10101000.00000001.00000000

How to Use

Enter the values for the patient or scenario you are assessing. Estimate total communication rounds and bandwidth needed for federated model training given model size, client count, and compression ratio. Use the federated comm cost result to inform your clinical assessment.

Related Calculators

Model Watermarking

Cybersecurity & Machine Learning