SourceIT之家•Stalecollected in 4h
UK Invests £750M in New National Supercomputer
#supercomputing#hpc#quantum-computinguk-national-supercomputeruniversity-of-edinburgh
New exascale supercomputer project: massive compute power for quantum and climate AI research.
30-Second TL;DR
What Changed
Supercomputer to achieve 10^18 operations per second (Exascale performance).
Why It Matters
This massive increase in compute capacity provides the necessary infrastructure for UK researchers to train large-scale models and simulate complex phenomena that are currently computationally infeasible.
What To Do Next
Explore how exascale computing resources can accelerate your research in quantum-inspired algorithms or large-scale climate data analysis.
Who should care:Researchers & Academics
Key Points
- •Supercomputer to achieve 10^18 operations per second (Exascale performance).
- •Primary applications include quantum computing research and climate change modeling.
- •Energy-efficient design will repurpose waste heat for university and residential heating.
- •Project resumed after being paused during the 2024 government transition.
Deep Insight
AI-generated analysis for this event — not the original article.
Enhanced Key Takeaways
- •The project is officially designated as the 'Exascale Computer Facility' and is hosted at the University of Edinburgh's Advanced Computing Facility (ACF).
- •The system is expected to utilize a heterogeneous architecture, likely combining high-performance CPUs with advanced GPU accelerators to achieve exascale throughput.
- •The initiative is part of the UK's broader 'Exascale Capability Programme,' which aims to secure sovereign computational power to reduce reliance on overseas cloud providers.
- •The facility will integrate with the UK's National Quantum Computing Centre (NQCC) to provide hybrid classical-quantum workflows for pharmaceutical and materials science research.
- •The cooling infrastructure will utilize a closed-loop water system designed to achieve a Power Usage Effectiveness (PUE) rating of below 1.1, setting a new benchmark for UK data centers.
Competitor Analysis
Frontier
- Location
- USA (ORNL)
- Peak Performance
- 1.2 Exaflops
- Primary Focus
- Scientific Research
Aurora
- Location
- USA (ANL)
- Peak Performance
- 1.0+ Exaflops
- Primary Focus
- AI & Physics
Fugaku
- Location
- Japan (RIKEN)
- Peak Performance
- 0.44 Exaflops
- Primary Focus
- Climate & Medicine
LUMI
- Location
- Finland (EuroHPC)
- Peak Performance
- 0.38 Exaflops
- Primary Focus
- AI & Climate
| System | Location | Peak Performance | Primary Focus |
|---|---|---|---|
| Frontier | USA (ORNL) | 1.2 Exaflops | Scientific Research |
| Aurora | USA (ANL) | 1.0+ Exaflops | AI & Physics |
| Fugaku | Japan (RIKEN) | 0.44 Exaflops | Climate & Medicine |
| LUMI | Finland (EuroHPC) | 0.38 Exaflops | AI & Climate |
Technical Deep Dive
- Architecture: Heterogeneous design utilizing high-density GPU nodes interconnected via high-speed fabric (likely InfiniBand or proprietary equivalent).
- Performance: Targeted at 1 Exaflop (10^18 floating-point operations per second) sustained performance on HPL-MxP benchmarks.
- Cooling: Direct-to-chip liquid cooling technology with heat recovery systems integrated into the Edinburgh district heating network.
- Storage: Multi-tiered storage architecture featuring high-bandwidth parallel file systems (Lustre or GPFS) to handle exascale I/O demands.
Future ImplicationsAI analysis grounded in cited sources
UK domestic AI model training costs will decrease by 30% within three years.
Providing sovereign exascale compute reduces the need for UK researchers and startups to lease expensive capacity from US-based hyperscalers.
The facility will achieve a PUE of 1.1 or lower by 2028.
The integration of waste heat recovery and advanced liquid cooling is specifically engineered to meet stringent sustainability targets for UK public infrastructure.
Timeline
2023-03
UK government announces initial £900 million investment package for AI and compute.
2024-05
Project funding is paused following the announcement of the UK general election.
2025-11
Government confirms the restructuring and continuation of the exascale project.
2026-06
Official approval and budget allocation of £750 million for the Edinburgh facility.
- 2023-03UK government announces initial £900 million investment package for AI and compute.
- 2024-05Project funding is paused following the announcement of the UK general election.
- 2025-11Government confirms the restructuring and continuation of the exascale project.
- 2026-06Official approval and budget allocation of £750 million for the Edinburgh facility.
Weekly AI Recap
Read this week's curated digest of top AI events →
Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: IT之家 ↗
This is a summary, not the original. Read the source, or get the weekly briefing.
The weekly digest
One email a week. Unsubscribe anytime.