Data-Intensive, Compute-Intensive Apps System Design

Telusko · Intermediate ·🏗️ Systems Design & Architecture ·4mo ago

Key Takeaways

Designs data-intensive and compute-intensive applications using system design principles

Full Transcript

Instagram and weather simulation apps both run on the server but in order to scale them the approaches are pretty much different because when your app is slow because it is reading and writing too much data those kind of applications are called data inensive applications and when in your app the calculations are pretty heavy those are called compute intensive applications. In order to scale a data inensive application, you can do various operations like adding caches strategies, maybe add caching systems, add databases, sh the databases or maybe replication. These all helps in scaling a data inensive application. But in order to scale a compute inensive application, you need to add the power to CPU or GPU. Suppose in a game if your character is standing and the shadow is on the left side because the sun is on the right. It can't be handled through databases. You need computational powers only. That's why some of the high-end games performs very poorly when you have a degraded system. You can't do anything but to add more CPU powers in order to work with them. So yeah, these are the two applications on which we have to first think and then pick the components wisely. For more such system design content, please subscribe to Terescu.
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Related Reads

📰
Cache Stampede: The Failure Mode Hiding Behind a 99% Hit Rate
Learn how a 99% cache hit rate can still lead to a cache stampede failure mode and how to identify and mitigate it
Dev.to · speed engineer
📰
A Differential Test Harness for Native vs. Generic XDP: Methodology and Baseline
Learn to test native vs. generic XDP using a differential test harness and understand the methodology and baseline for comparison
Dev.to · Kazuru
📰
The Everyday Backend Engineer: Step 10 — The Observer Pattern
Learn the Observer pattern to decouple objects and improve backend system maintainability
Dev.to · Manoj Khatri
📰
ArborDb: a Rust document store where reading one field doesn't get slower as the record grows
Learn about ArborDb, a fast Rust document store that maintains read performance regardless of record size
Dev.to · Frédéric Meyer
Up next
9-Step Software Architect Roadmap 2026 | System Design | #shorts
SCALER
Watch →