
Alps
Swiss national supercomputer using Nvidia GH200 chips for AI and scientific research.
Alps is a high-performance computer operated by the Swiss National Supercomputing Centre in Lugano. It uses Nvidia Grace Hopper chips to support artificial intelligence and scientific workloads for researchers.
- Uses Nvidia GH200 chips with over one million cores.
- Powered entirely by hydropower in Switzerland.
- Costs CHF 100 million to build.
- Ranks in the global top ten for speed.
- Cools itself using deep lake water.
What it is
The Alps supercomputer is a powerful machine located in Lugano, Switzerland. It is run by the Swiss National Supercomputing Centre, known as CSCS. The system was officially opened in September 2024. It uses Nvidia Grace Hopper chips to deliver high performance. These chips combine ARM processors with graphics cards. The machine helps scientists train large language models and analyse complex data. It ranks among the top ten most powerful computers globally.
How it came about
The Swiss National Supercomputing Centre started in 1991. It has provided computing services for decades. Past projects included analysing particle physics data and weather simulations. The name Alps began in 2020 with earlier systems. The current version arrived in late 2024. It replaced older hardware to meet growing demand. Researchers previously noted it could train early AI models quickly. The project received funding from the Swiss government through the ETH Domain.
The compute
The system relies on Nvidia GH200 integrated circuits. Each chip holds seventy-two ARM CPUs and one hundred thirty-two GPUs. There are over one million processor cores in total. Nodes connect using the Slingshot-11 network at high speeds. Memory includes fast LPDDR and HBM types. The software stack supports various scientific applications. Hewlett Packard Enterprise supplied the hardware infrastructure. The design allows flexible installation of computing units.
Where the GPUs are
The computer sits in a dedicated building in Lugano-Cornaredo. Lake water cools the system from a depth of forty-five metres. Pipes carry cold water to heat exchangers inside the facility. This method reduces energy use significantly. The data centre spans three floors. Infrastructure sits on the lowest level. Computing racks occupy the top floor. The setup supports continuous operation with emergency power backups.
Customers and deals
Access is granted by a panel of scientific experts. Users include universities and research institutes. EPFL and Yale have already secured time on the machine. They use it for health data analysis and AI training. The system supports the Swiss AI Initiative. ETH Zurich also benefits from this infrastructure. Services are reserved for selected academic and scientific groups.
Money and ownership
The initial cost reached approximately one hundred million Swiss francs. Annual running costs stand at thirty-seven million francs. Electricity bills alone add fifteen million francs yearly. The system runs on hydropower to lower emissions. Funding comes from the Swiss Confederation. No private investors or stock listings are involved. The state covers all operational expenses.
Its place in AI
Alps plays a key role in Swiss science. It enables advanced climate and medical studies. The machine supports local AI development efforts. It competes with systems in the US and Japan. Environmental groups note its low carbon footprint. However, concerns remain about grid pressure. The project demonstrates national commitment to high-tech research.
Questions and criticism
Critics point to high electricity consumption. Ten megawatts equals the usage of thousands of homes. Water extraction from the lake raises ecological questions. Grid stability faces strain during peak loads. Some argue the environmental cost outweighs benefits. Proponents claim the scientific gains justify resource use. CO2 emissions stay minimal due to hydropower.



Questions people ask about Alps
What chips does Alps use?
It uses Nvidia Grace Hopper superchips. These combine ARM CPUs with Hopper GPUs on one chip.
Who can use the supercomputer?
A panel of experts selects users. Access goes to scientific and university researchers.
How much does it cost to run?
Annual operating costs are around thirty-seven million Swiss francs.
Where is it located?
It sits in Lugano-Cornaredo, Switzerland, near Lake Lugano.
Good reading and sources
- Swiss National Supercomputing Centre, Hewlett Packard Enterprise and NVIDIA Announce World鈥檚 Most Powerful AI-Capable SupercomputerNvidia, Article
- Understanding Data Movement in Tightly Coupled Heterogeneous Systems: A Case Study with the Grace Hopper SuperchiparXiv, Article
The facts on this page come from the Wikipedia article "Alps (supercomputer)", written by its contributors and used under the CC BY-SA 4.0 licence, and from its Wikidata record (CC0). EveryNeocloud rewrote the text in its own words, with AI assistance, and compares every page with the article to make sure no passage is copied. It may be reused under the same CC BY-SA 4.0 licence with a link back.
It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-17.
Like Alps
All AI supercomputers
馃嚭馃嚫 United StatesNCAR-Wyoming Supercomputing Center
A high-performance computing and data archival facility in Cheyenne, Wyoming.Supercomputer 路 Other Nvidia GPUs 路 2003
馃嚬馃嚰 TaiwanTaiwania 3
A CPU-based supercomputer for scientific research and biomedical analysis in Taiwan.Supercomputer 路 Other Nvidia GPUs 路 2020
馃嚪馃嚭 RussiaChristofari
Sberbank's Nvidia-based supercomputers in Moscow for neural network trainingSupercomputer 路 Other Nvidia GPUs 路 2019
More from Nvidia
All of them
馃嚥馃嚲 Malaysia and 4 moreYTL AI Cloud
Malaysian AI cloud and supercomputer run from a solar-backed Johor data centre parkNeocloud 路 Nvidia Blackwell 路 2024
馃嚞馃嚙 United KingdomNscale
UK-based GPU cloud provider and data centre operator founded in 2024.Neocloud 路 Other Nvidia GPUs 路 2024
馃嚭馃嚫 United StatesColossus
An AI supercomputer cluster in Tennessee and Mississippi built by xAI.Supercomputer 路 Nvidia Hopper 路 2024
馃嚛馃嚜 GermanyPolarise
German sovereign AI cloud running its own data centres and open-weight model APIsNeocloud 路 Nvidia Blackwell 路 2024