1: \begin{abstract}

2:     Asynchronous Federated Learning with Buffered Aggregation (FedBuff) is a state-of-the-art algorithm known for its efficiency and high scalability.

3:     However, it has a high communication cost, which has not been examined with quantized communications.

4:     To tackle this problem, we present a new algorithm (\algname), with a quantization scheme that establishes a shared ``hidden'' state between the server and clients to avoid the error propagation caused by direct quantization.

5:     This approach allows for high precision while significantly reducing the data transmitted during client-server interactions.

6:     We provide theoretical convergence guarantees for \algname and corroborate our analysis with experiments on a standard benchmark.

7: \end{abstract}

8: