JUPITER Booster
Forschungszentrum Jülich runs JUPITER for the EuroHPC Joint Undertaking, Europe's first exascale system, used for large-scale AI model training and simulation in climate, medicine, energy systems and European languages.
What it runs and how to get time
From the operator- Used for What the operator says the system is used for: science domains, missions or workloads. Taken from the operator, its funder or the system vendor, not from TOP500.
- AI model training including European language models, plus simulation for climate and weather, medicine, energy systems, chemistry, physics and engineering 4
- How to get time How researchers or customers get time on the system, naming the allocation programme where the operator states one.
- EuroHPC Access Calls (50% of time) and Gauss Centre for Supercomputing calls for German researchers (50%); JUPITER AI Factory serves startups and SMEs 4
- Cooling Cooling design stated by the operator or supplier.
- Warm-water direct liquid cooling in BullSequana XH3000 racks 3
- Power design Facility power, energy supply or heat reuse as stated by the operator. Not the benchmark power in the Power column.
- About 11 MW average, up to 20 MW for heavy AI workloads; waste heat to be fed into the Jülich campus heating network 4
- Facility Building, data centre or campus that houses the system, as described by the operator.
- Modular Data Centre of about 50 container modules over 2,300 m2, Forschungszentrum Jülich campus, Jülich, Germany 3
- Upgrade or successor Announced upgrade, expansion, successor or retirement plan, with dates where the operator gives them.
- JUPITER Cluster module being built; first partition (582 nodes) due end of 2026, second partition (162 nodes) to follow 7
Benchmark and hardware
Source per value- Rank Position in the TOP500 edition named in the Edition column. Historic systems keep the rank of the edition where they last appeared.
- 5 1
- Rmax PFLOP/s Measured double-precision LINPACK (HPL) performance of the listed configuration. Older TFLOP/s and GFLOP/s figures are converted to PFLOP/s.
- 1,000 1
- Rpeak PFLOP/s Theoretical peak double-precision performance of the listed configuration. This is not AI or mixed-precision throughput.
- 1,226.28 1
- HPL efficiency % Rmax divided by Rpeak: the share of theoretical peak the system delivered on LINPACK. Computed from the two sourced figures.
- 81.5 1
- Power kW Power reported for the benchmark run. Measurement boundaries vary by site; this is not facility power, TDP or cooling capacity. Empty when the system does not report it.
- 15,794 1
- Efficiency GFLOP/s per W Rmax per watt of reported benchmark power, computed from the two sourced figures (GFLOP/s per W). Empty when power is not reported.
- 63.32 1
- Site Host or operator named by TOP500 for the installation.
- EuroHPC/FZJ 1
- Country Country or region reported by TOP500 for the site.
- Germany 1
- Region World region of the installation country (grouping applied by this site).
- Europe 1
- Vendor Manufacturer or system supplier exactly as submitted to TOP500.
- Bull 1
- Vendor group Supplier grouped for filtering. Bull is the Eviden/Atos line; Other covers single-system suppliers.
- Bull 1
- System model Hardware platform in the cited TOP500 edition.
- BullSequana XH3000 1
- Processor CPU description in the cited TOP500 edition; clock and core count refer to the processor model.
- GH Superchip 72C 3GHz 1
- Accelerator Accelerator model listed for the system. Empty when TOP500 lists none.
- NVIDIA GH200 Superchip 1
- Accelerator vendor Maker of the accelerator. None listed means TOP500 names no accelerator for the system.
- NVIDIA 1
- Interconnect Network fabric exactly as listed in the cited TOP500 edition.
- Quad-Rail NVIDIA InfiniBand NDR200 1
- Fabric Interconnect family derived from the listed interconnect (grouping applied by this site).
- InfiniBand 1
- Cores cores Total cores submitted to TOP500, including accelerator cores where TOP500 counts them. Not a CPU-only count and not a CUDA-core count.
- 4,801,344 1
- Installed Installation year reported in the TOP500 system record. Not necessarily the year of production acceptance.
- 2025 2
- Edition TOP500 edition (YYYY-MM) that supplies the ranking and benchmark data. A historic edition does not assert current operating status.
- 2026-06 1
- OS Operating system from the TOP500 system record.
- Red Hat Enterprise Linux 2
- TOP500 record Link to the TOP500 database record for this exact submission.
- https://www.top500.org/system/180357/ 2
- Nodes nodes Compute node count as described by the operator. It can exceed the nodes used in the benchmark run.
- 5,884 3
- Accelerators per node per node Accelerator packages or superchips per node, as the operator describes them.
- 4 3
Sources
- TOP500 list, June 2026 accessed 2026-11-11
- TOP500 system record accessed 2026-11-11
- fz-juelich.de (operator page) accessed 2026-11-11
- JUPITER: Launching Europe's Exascale Era accessed 2026-11-11
- JUPITER - Exascale for Europe accessed 2026-11-11
- Next Milestone for JUPITER accessed 2026-11-11
- JUPITER (JSC systems page) accessed 2026-11-11