Google DeepMind sets new theoretical record in matrix multiplication
Google DeepMind has lowered the theoretical upper bound for matrix multiplication to below 2.371177 using the AI agent AlphaEvolve. The advancement concerns one of the oldest open mathematical questions in computer science.
What happened?
Together with academic researchers, Google DeepMind has pushed down the theoretical upper bound for matrix multiplication, known as the exponent omega (ω). By utilising its AI-based coding agent AlphaEvolve, the team successfully lowered the limit from the previous record of 2.371339 to below 2.371177. Matrix multiplication is one of computer science's oldest open problems and serves as the foundation for modern computational capacity.
Key facts
| Ny teoretisk gräns (ω) | < 2,371177 |
|---|---|
| Tidigare rekord (ω) | 2,371339 |
| Använd AI-agent | AlphaEvolve |
”matrismultiplikation är den grundläggande beräkningsoperation som driver modern computing, inklusive AI”
Why it matters
The breakthrough demonstrates how AI agents can be employed to solve long-standing and complex mathematical problems. Matrix multiplication is the fundamental computational operation driving modern AI and supercomputers, meaning any theoretical or practical improvement could have significant implications for future computing infrastructure.
Who is affected?
The result is primarily of theoretical interest to researchers in algorithms and computer science, but it ultimately affects all developers and companies building compute-intensive software. More efficient algorithm designs may eventually lead to faster training and execution of AI models for developers globally.
Impact on the EU
As the breakthrough pertains to theoretical computer science and fundamental algorithms, it is unaffected by EU regulations such as the AI Act or GDPR. The research is available globally to both academia and industry.
What else you should know
Matrix multiplication is a central type of calculation in everything from 3D graphics and physics simulations to the training of large language models. The researchers used the coding agent AlphaEvolve to search through and optimise the complex tensor structures required to lower the theoretical bound.
Quick answers about this story
Vad har hänt?
När hände det?
Varför spelar det roll?
Vilka berörs av genombrottet?
The link opens in a new window and leads to the publisher's own site.
Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.
AI-verktyg i artikeln
Topics
Get similar news straight to your inbox
The reader's room
Send in a question or an addition. The newsroom reads everything before it's published and replies when relevant. No AI-generated text – just people.
Sign in to submit a comment or question.
Read the article through your role
- Decide whether this affects strategy over 6–12 months or is just noise.
- Discuss with leadership: do we own the right question or does ownership need to move?
- Ask: what risk are we taking by NOT acting on this this quarter?
Generated angle — not editorial analysis of "Google DeepMind sets new theoretical record in matrix multip"