Embedded HPC

Aurora HPC 20-23

Super Computer Capabilities/Record Breaking Energy Efficiency



Click to Enlarge


Features

• Extreme performance and high density
• Best in the market for energy efficiency
• Based on Intel® Xeon™
• GPU accelerators

• Optional on-board FPGA

• Intel Phi accelerators
• Air and water cooling

• Ruggedized enclosures

   

The AURORA HPC 20-23 is available as a board-only solution, and also within a rugged enclosure, with stand-alone or external cooling, providing a ready-to-go solution for embedded high performance applications. It fits into the Aurora HPC

Supercomputing capabilities – Aurora HPC 20-23 boards are used in the Eurotech Aurora Tigon supercomputers, the greenest and most dense machines in the high performance computing market.

GPU acceleration – Aurora HPC 20-23 is able to perform at 2.8 Tflop/s (1.7 sustained). It has very fast I/O with Infiniband or Gigabit ETH.

Energy efficient – Aurora HPC 20-23 marks a record in energy efficiency with > 3.15 GFlop/s per Watt.

Reliable – No moving parts eliminate vibrations. Direct cooling avoids hot spots while the soldered memory provides speed and robustness.

FPGA – Aurora HPC 20-23 is provided with an Altera Stratix FPGA connected with 2 links PCIe3 x8 to the CPUs.

Optionally water cooled – Optionally cooled with direct hot liquid cooling that doesn’t require air conditioning and ventilation. Pervasive cooling is applied to every component of the board to maximize the cooling effectiveness.

Applications

The Aurora HPC 20-23 is ideal for both data center and embedded usage.

  • High performance computing (HPC)Heavy computational loads
    Simulation
    Signal processing
    Rugged HPC


Cooling

The Aurora HPC 20-23 board is water cooled to maximize performance, energy efficiency and deployability in demanding environments.
The Aurora HPC 20-23 can be easily connected to an heat exchanger through Eurotech or third party plumbing.


 


Aurora 20-23 Block Diagram


Cooling Options

Integrated Water Cooling


Technical Specifications
COMPUTING POWER

2.8 TFlop/s per board (peak)

PROCESSOR

2 x Intel Xeon E5 series per node

ACCELERATORS

2 x Nvidia Kepler K20 GPU per node OR
2 x Intel Xeon Phi 5115D per node

ARCHITECTURE

2 CPUs connected via 8.0GT/s QPI link
2 accelerators connected via PCIe2 to the board
2 x PCIe sockets

MEMORY
Up to 128 GB soldered RAM per node. ECC DDR3 SDRAM 1866 MT/s
FPGA

Stratix V SGXB6

CHIPSET

Intel I/O hub (Patsburg)

INTERFACES

1 x 40 Gbps QDR Infiniband
1 x 1Gbs Ethernet (Intel PXE)
2 x USB host (internal)
1 x standard VGA - Matrox MGA G200 with 8Mb RAM integrated in the BMC
Optional: 1+1 3D Torus or 3D mesh BW:up to 240+240Gbps, Latency: ~1us

PCI EXPRESS

2 x PCI Express x 16

SERIAL ATA

1 x 3 Gbps standard connector

BIOS

Insyde H2O

LOCAL STORAGE Up to 2 TB GB 2,5” Sata Disk or up to 512 GB 1,8” microSATA SSD
HW MONITORING AND CONTROL

IPMI - temperature, current (hall effect current sensor) and voltage
CAN BUS

OPERATING SYSTEM
Linux, Windows 7, Windows Embedded
 
Environmental
DIMENSIONS
W 210mm, D 515 mm, H 16mm With cold plate: W 210mm, D 515 mm, H 36mm
COOLING Aurora Direct Hot Liquid Cooling (external or embedded heat exchanger configuration)
POWER CONSUMPTION
860 W peak
POWER SUPPLY
48 V DC input
1 x Local DC/DC 48v DC to 9.6 V DC nominal
2 x Local DC/DC 48v DC to 12 V DC nominal
High conversion efficiency: 98%
OPERATING TEMPERATURE 5°C to +60°C

 

back to top