Skip to content
cognition.com·

🔒Cognition Team Cracks RSA-260 With GPU-Optimized GNFS

GPU-powered factorization drops RSA-260 barrier

TL;DR

The Cognition team used GPU-optimized GNFS to factor RSA-260, a 260-digit number, at a cost of $400,000. This breakthrough lowers the barrier to entry for cryptanalytic work and large-scale scientific computing.

The Cognition team cracked RSA-260, a 260-digit number, using a GPU-optimized version of the general number field sieve (GNFS). This marks a significant milestone in computational mathematics and cryptography, as it demonstrates the power of GPU clusters in tackling complex factorization problems. For developers and researchers, this means that the cost of performing such computations is dropping, making it more accessible for smaller teams and projects. The factorization cost about $400,000, with 643 GPU-days in polynomial selection, 3,813 GPU-days sieving, and 467 GPU-days linear system solving. The team used spare compute resources, highlighting the potential for more efficient use of existing infrastructure.

Cognition Team Cracks RSA-260 With GPU-Optimized GNFS — cognition.com

Key Points

1

Cognition team factored RSA-260 using GPU-optimized GNFS, a 260-digit number.

2

Factorization cost about $400,000, with 643 GPU-days in polynomial selection.

3

3,813 GPU-days were spent sieving, and 467 GPU-days were used for linear system solving.

4

The team used spare compute resources, highlighting efficient infrastructure use.

5

Devin, a software engineer, autonomously handled measurements and cluster operations.

Why It Matters

If you're working on cryptanalytic projects or large-scale scientific computing, this breakthrough lowers the cost barrier. For example, factoring RSA-1024 would cost roughly $30 million, making it more accessible for smaller teams. However, RSA-2048 remains roughly a billion times harder, so the impact is limited to specific use cases.

RSAGNFSGPUcryptographicsecurity

Frequently Asked Questions

Why does this matter?

If you're working on cryptanalytic projects or large-scale scientific computing, this breakthrough lowers the cost barrier. For example, factoring RSA-1024 would cost roughly $30 million, making it more accessible for smaller teams. However, RSA-2048 remains roughly a billion times harder, so the impact is limited to specific use cases.

What happened?

The Cognition team used GPU-optimized GNFS to factor RSA-260, a 260-digit number, at a cost of $400,000. This breakthrough lowers the barrier to entry for cryptanalytic work and large-scale scientific computing.

Comments

Subscribe to join the conversation...

Be the first to comment

Enjoyed this article?

Get it daily. 7am. Free. Reads in 5 minutes.

Join 3,472 builders reading daily.

Also get