MENTAL MODEL · M4532

Von Neumann Bottleneck

DecideHigh supportDecision Science

Included

Updated 2026-08-08

Loading revision record…

INTRODUCTION

English translation pending.

CORE DEFINITION

Named after John von Neumann's stored-program architecture, the bottleneck describes how a single bus carrying both instructions and data forces the processor to idle while waiting for memory. Because computation has grown far faster than memory bandwidth, the gap, often called the memory wall, limits throughput regardless of how capable the arithmetic units are. Caching, prefetching, and near-memory computing are the standard responses.

SCAFFOLDING EFFECT

psychology

Reduce cognitive load

- Find the bus: ask who is waiting on whose data rather than who is smartest in the room - Move the work: push computation to where the data already sits instead of shipping data to the decider - Cut the traffic: cache, batch, and prefetch to reduce how often the slow channel is used

anchor

Anchor fast decisions

Instructions and data share one channel, so every operation requires at least one trip to memory. As processors accelerated, memory latency did not keep pace, and the transfer time now dominates execution. Adding faster arithmetic therefore yields little, while reducing data movement or relocating computation near the data attacks the actual constraint.

MINIMUM ACTION

In progress 0/1

Practice this model in one real situation:

Identify the transfer bottleneck by asking who waits on whose data. 2. Push computation toward the data instead of moving data to the decider. 3. Use caching and prefetching to cut the number of slow round trips. 4. Parallelize independent work to break serial dependence on the shared channel.

Check to track your progress (stored locally)

Learning progress0%

account_treeGenealogyexpand_more

menu_bookReferencesexpand_more

Source support: Explicit

link
en.wikipedia.orghttps://en.wikipedia.org/wiki/Von_Neumann_architectureZH · Explicit
verified

RELATED MODELS