Our website uses cookies to give you the most optimal experience online by: measuring our audience, understanding how our webpages are viewed and improving consequently the way our website works, providing you with relevant and personalized marketing content.
You have full control over what you want to activate. You can accept the cookies by clicking on the “Accept all cookies” button or customize your choices by selecting the cookies you want to activate. You can also decline all non-necessary cookies by clicking on the “Decline all cookies” button. Please find more information on our use of cookies and how to withdraw at any time your consent on our privacy policy.

Managing your cookies

Our website uses cookies. You have full control over what you want to activate. You can accept the cookies by clicking on the “Accept all cookies” button or customize your choices by selecting the cookies you want to activate. You can also decline all non-necessary cookies by clicking on the “Decline all cookies” button.

Necessary cookies

These are essential for the user navigation and allow to give access to certain functionalities such as secured zones accesses. Without these cookies, it won’t be possible to provide the service.
Matomo on premise

Marketing cookies

These cookies are used to deliver advertisements more relevant for you, limit the number of times you see an advertisement; help measure the effectiveness of the advertising campaign; and understand people’s behavior after they view an advertisement.
Adobe Privacy policy | Marketo Privacy Policy | MRP Privacy Policy | AccountInsight Privacy Policy | Triblio Privacy Policy

Social media cookies

These cookies are used to measure the effectiveness of social media campaigns.
LinkedIn Policy

Our website uses cookies to give you the most optimal experience online by: measuring our audience, understanding how our webpages are viewed and improving consequently the way our website works, providing you with relevant and personalized marketing content. You can also decline all non-necessary cookies by clicking on the “Decline all cookies” button. Please find more information on our use of cookies and how to withdraw at any time your consent on our privacy policy.

Skip to main content

BullSequana X1000 supercomputers

Paving the way towards Exascale challenges

Open supercomputers

Open for future technologies. BullSequana X1000 proposes a large choice of processors (Intel CPUs, accelerators, ARM processors…) and different interconnect technologies (BXI, InfiniBand…), thus offering maximum investment protection. It is designed to integrate the most advanced technologies, ready to support the future technologies that will make it possible to reach the exaflops level.
Limit energy consumption. Controlling energy consumption is the main roadblock on the path to exascale. BullSequana X1000 is ultra energy-efficient: it targets a PUE very close to 1. 100% of the components of sequana – both compute nodes and switches – are cooled with warm water up to 40°C, using an enhanced version of the Bull Direct Liquid Cooling (DLC) technology.
Accelerate application performance. Exascale application performance requires massive parallelism. BullSequana X1000 features the Bull Exascale Interconnect (BXI) developed by Atos specifically for exascale. BXI introduces a revolution with a hardware acceleration technology that frees up processors from all communication tasks.

Download BullSequana X1000 technical factsheet

Bull Sequana X1000 cell

BullSequana X1000 compute cabinet

There are two types of compute cabinets:
The base compute cabinet provides liquid cooling to the compute blades it houses and also to the switch cabinet.
The extension compute cabinet is optional and has its own cooling sub-system.

Three types of interfaces connect the compute and switch cabinets together:
Liquid interface with two flexible pipes from Base compute cabinet hydraulic chassis to switch cabinet manifolds.
Interconnect interface: each node is connected through one or two interconnect links, one per NIC.
Management interface for all management controllers in the compute cabinet (PMC, BMC, HMC): one 1Gb/s Ethernet link per node plus 4 sidebands signals for platform management.

The compute nodes are connected to the switch cabinet (L1 switches) via 12 copper cable octopuses named:
Sideplane Octopus Double for 2 NICs per node,
Sideplane Octopus Single for 1 NIC per node.

Each BullSequana X1000 compute cabinet houses 48 horizontal compute blades, with the associated power modules at the top of the cabinet and the redundant hydraulic modules for cooling at the bottom of the cabinet.

24 blades are mounted on the front side of the cabinet, while the 24 other blades are mounted on the rear side.
Each 1U blade contains three compute nodes, except for GPU-based blades which contain one node.

Each cell can therefore contain up to:

  • 288 dual-socket Intel® Xeon® nodes,
    or
  • 96 dual-socket Intel® Xeon® nodes with 4 Nvidia® Volta GPUs,
    or
  • 288 dual-socket Cavium® ThunderX2™ nodes.

Bull Sequana X1000 cell

BullSequana switch cabinet

The central Bull Sequana X1000 switch cabinet houses:

two levels of interconnect switches in full fat-tree topology,
their associated DC power generator (SWPG: SWitch Power Group) and
management components such as Ethernet switches and one redundant server for Island management and administration (ISMA).

The switch cabinet provides internal connections (within the cell) through copper cables and external connections through optical cables.

288 external ports are available to interconnect each cell with the rest of the system to form different topologies, such as Fat-tree, All-to-all or other directly connected configurations.

External nodes (such as for example input/ output gateways, services nodes…) can be plugged into the L1 switches replacing compute nodes.


Bull Sequana X1120

Bull Sequana X1120 compute blade

It integrates 3 compute nodes, each powered by 2 Intel® Xeon® Processor Scalable Family (Cascade Lake). In each 1U blade, a cold plate with active liquid flow cools all hot components by direct contact – the sequana compute blades contain no fan.

Design
1U blade comprising 3 compute nodes side-by-side
Processors
3 x 2 future Intel® Xeon® Processor Scalable Family (Cascade Lake)
Architecture
3 x 1 Intel® C620 chipset
Memory
3 x 12 DDR4 memory slots (max 384 GB with 32 GB DIMMs)
+ 4 optional NVRAM DIMMS (NVRAM availability TBC)
I/O slots
InfiniBand EDR 1 port mezzanine board
OR BXI 1 or 2 ports mezzanine board
Storage
3 x 1 optional SATA drive
3 x 1 optional NVMePCIe SSD drive via PCIe switch
Power supply
in cabinet
Cooling
Cooling by direct contact with DLC coldplate or through heat spreaders for DIMMS
Physical specifications
44.45 x 600 x 540 mm (H x W x D) Weight: 25 kg max
OS and software
Red Hat Enterprise Linux & Bull Supercomputer Suite support
Warranty
Standard warranty: 1 year – Extended warranty: consult us
Regulatory compliance
Safety (EC, IEC, UL, CSA certifications); Electromagnetic Compatibility (EC, FCC, ICES-03, VCCI certifications), Environment (RoHS II & WEEE directives, REACH regulation)

Bull Sequana X1110

Bull Sequana X1110 compute blade

It integrates 3 compute nodes side-by-side, each powered by 2 Intel® Xeon® E5-2600 V4 processors (code named Broadwell-EP). In each 1U blade, a cold plate with active liquid flow cools all hot components by direct contact – the Sequana compute blades contain no fan.

Design
1U blade comprising 3 compute nodes side-by-side
Processors
3 x 2 Intel® Xeon® E5-2600 v4 processors
Architecture
3 x 1 Intel® C610 chipset
Memory
3 x 8 DDR4 memory slots (max 256 GB with 32 GB DIMMs)
I/O slots
InfiniBand EDR 1 port mezzanine board
OR BXI 1 or 2 ports mezzanine board
Storage
3 x 1 optional SATA SSD drive
Power supply
in cabinet
Cooling
Cooling by direct contact with DLC coldplate or through heat spreaders for DIMMS
Physical specifications
44.45 x 600 x 540 mm (H x W x D) Weight: 25 kg max
OS and software
Red Hat Enterprise Linux & Bull Supercomputer Suite support
Warranty
Standard warranty: 1 year – Extended warranty: consult us
Regulatory compliance
Safety (EC, IEC, UL, CSA certifications); Electromagnetic Compatibility (EC, FCC, ICES-03, VCCI certifications), Environment (RoHS II & WEEE directives, REACH regulation)

Bull Sequana X1125

BullSequana X1125 GPU-accelerated blade

It includes a 2 Intel® Xeon® Processor Scalable Family (Cascade Lake) compute node equipped with 4 NVIDIA® Tesla Volta V100 GPUs.

In each 1U blade, a cold plate with active liquid flow cools all hot components by direct contact – the BullSequana compute blades contain no fan.

Design
1U blade with 1 accelerated compute node
Processors
2 Intel® Xeon® Processor Scalable Family
4 Nvidia® Tesla Volta V100 GPUs
Architecture
1 Intel® C620 chipset
Memory
12 DDR4 memory slots (max 384 GB with 32 GB DIMMs)
I/O slots
InfiniBand EDR 1 port mezzanine board
OR BXI 1 port mezzanine board
Storage
1 optional SATA SSD drive
Power supply
in cabinet
Cooling
Cooling by direct contact with DLC coldplate or through heat spreaders for DIMMS
Physical specifications
44.45 x 600 x 540 mm (H x W x D) Weight: 25 kg max
OS and software
Red Hat Enterprise Linux & Bull Supercomputer Suite support
Warranty
Standard warranty: 1 year – Extended warranty: consult us
Regulatory compliance
Safety (EC, IEC, UL, CSA certifications); Electromagnetic Compatibility (EC, FCC, ICES-03, VCCI certifications), Environment (RoHS II & WEEE directives, REACH regulation)

BullSequana X1310 with ARM processors

It integrates 3 compute nodes, each powered by 2 Cavium® ThunderX2™ Armv8 processors. In each 1U blade, a cold plate with active liquid flow cools all hot components by direct contact – the Sequana compute blades contain no fan.

Design
1U blade comprising 3 compute nodes side-by-side
Processors
3 x 2 Cavium® ThunderX2™ Armv8 processors with up to 32 cores
Architecture
3 x 1 motherboard compatible with Cavium Borg reference platform
Memory
3 x 16 DDR4 memory slots (max 1024 GB with 64 GB DIMMs)
I/O slots
InfiniBand EDR 1 port mezzanine board
Storage
3 x 1 optional SATA drive
Power supply
in cabinet
Cooling
Cooling by direct contact with DLC coldplate or through heat spreaders for DIMMS
Physical specifications
44.45 x 600 x 540 mm (H x W x D) Weight: 25 kg max
OS and software
Red Hat Enterprise Linux & Bull Supercomputer Suite support
Warranty
Standard warranty: 1 year – Extended warranty: consult us
Regulatory compliance
Safety (EC, IEC, UL, CSA certifications); Electromagnetic Compatibility (EC, FCC, ICES-03, VCCI certifications), Environment (RoHS II & WEEE directives, REACH regulation)

BullSequana X1000 is powered by

Intel Xeon

Intel® Xeon® Processor Scalable Family

Related resources

Brochure

BullSequana X1000 Brochure

Discover the range of supercomputers developed by Atos’s R&D, the systems needed to meet the major challenges of the 21st century.

Factsheet

BullSequana X1000 in detail

Technical description of the BullSequana X1000 cabinets and compute blades.

Press Release

Atos delivers the most powerful supercomputer in Germany to Forschungszentrum Jülich

The BullSequana X1000 named JUWELS (short for Jülich Wizard for European Leadership Science) is operational at the Jülich Supercomputing Centre (JSC). The supercomputer has a peak performance of 12 petaflop/s, making it Germany’s fastest supercomputer.

Interested in our Bull Sequana X1000?