Standalone · Frontier

Beyond the buzz: when disaggregation actually pays

Splitting prefill and decode is the hot trend. But the 2025 paper that named the trend is honest about when it helps and when it doesn't. Reading it changed how I think about the architecture.

Every inference blog in 2025 had a post about disaggregation: split prefill and decode onto separate machines, and watch your throughput soar. The reality, from the paper Beyond the Buzz: A Pragmatic Take on Inference Disaggregation (arXiv 2506.05508), is more interesting and more honest.

The promise

Prefill is compute-bound, decode is memory-bound. They have different bottlenecks, so the argument goes, put them on different machines. Prefill machines can be compute-heavy, decode machines can be memory-heavy, and each runs at its own optimum. No more one GPU doing both badly.

The promise is real, and it's why NVIDIA Dynamo, Mooncake, and every serious serving vendor are building it. But the paper's contribution is the when.

The honest tradeoffs

Disaggregation costs you something real: network transfer. The KV cache has to move from the prefill machine to the decode machine, over the network, for every request. That's gigabytes per long request, and it's not free.

So the paper's finding, in my words: disaggregation pays when the workload is skewed. When prefill and decode are wildly asymmetric (long prompts, short outputs, or the reverse), separating them lets each side specialize. When the workload is balanced, the network overhead eats the gains.

Mental model

Disaggregation is like having a prep chef and a line cook instead of one cook doing everything. It wins in a busy restaurant with a clear split of tasks. It loses in a quiet one where the transfer between stations costs more than it saves.

What I took from it

The takeaway

The best architecture articles don't sell you a solution, they tell you when the solution works. Disaggregation is a tool, not a trend, and the tool fits a specific workload shape.

Back to the blog