Hewlett Packard Enterprise Frontier, or OLCF-5, is the world's first exascale supercomputer. It is hosted at the Oak Ridge Leadership Computing Facility (OLCF) in Tennessee, United States and became operational in 2022. As of December 2023[update], Frontier is the world's fastest supercomputer. It is based on the Cray EX and is the successor to Summit (OLCF-4). Frontier achieved an Rmax of 1.102 exaFLOPS, which is 1.102 quintillion floating-point operations per second, using AMD CPUs and GPUs.[2][3][4][5][6]
 | |
| Active |
|
|---|---|
| Operators | Oak Ridge National Laboratory and U.S. Department of Energy |
| Location | Oak Ridge Leadership Computing Facility |
| Power | 22.7 MW[1] |
| Operating system | HPE Cray OS |
| Space | 680 m2 (7,300 sq ft) |
| Speed | 1.206 exaFLOPS (Rmax) / 1.71481 exaFLOPS (Rpeak)[1] |
| Cost | US$600 million (estimated cost) |
| Purpose | Scientific research and development |
| Website | www |
Measured at 62.86 gigaflops/watt, the smaller Frontier TDS (test and development system) topped the Green500 list for most efficient supercomputer[6] until it was dethroned in efficiency by the Flatiron Institute's Henri supercomputer in November 2022.[7]
Design
editFrontier uses 9,472 AMD Epyc 7713 "Trento" 64 core 2 GHz CPUs (606,208 cores) and 37,888 Instinct MI250X GPUs (8,335,360 cores). They can perform double-precision operations at the same speed as single precision.[8]
"Trento" is an optimized third-generation EPYC CPU[9] ("Milan"), which is based on the Zen 3 microarchitecture.
It occupies 74 19-inch (48 cm) rack cabinets.[10] Each cabinet hosts 64 blades, each consisting of 2 nodes.
Blades are interconnected by HPE Slingshot 64-port switches that provides 12.8 terabits/second of bandwidth. Groups of blades are linked in a dragonfly topology with at most three hops between any two nodes. Cabling is either optical or copper, customized to minimize cable length. Total cabling runs 145 km (90 mi). Frontier is liquid-cooled by 4 350-horsepower pumps, which flow around 6,000 gallons (22,712.47 Liters) of non-pre chilled water through the system each minute, allowing 5x the density of air-cooled architectures.[8][11]
Each node consists of one CPU, 4 GPUs and 4 terabytes of flash memory. Each GPU has 128 GB of RAM soldered onto it, and each CPU has 512GB of local DDR4 memory.[8] [12]
Frontier has coherent interconnects between CPUs and GPUs, allowing GPU memory to be accessed coherently by code running on the Epyc CPUs.[13]
Frontier uses an internal 75 TB/s read / 35 TB/s write / 15 billion IOPS flash storage system, along with the 700 PB Orion site-wide Lustre filesystem.[14]
Frontier consumes around 21 megawatts (MW) (which is equivalent to the power needed for 15,000 single-family homes), compared to its predecessor Summit's 13 MW. [11]
History
editOne of the largest challenges during development was power consumption. Existing information pointed to hundreds of thousands of GPUs being necessary to achieve 1 exaFLOP, with a total power consumption of 150-500 MW. Thus, high efficiency was a primary target of the project.[8]
Oak Ridge partnered with HPE Cray and AMD to build the system at a cost of US$600 million. It began deployment in 2021[15] and reached full capability in 2022.[16] It clocked 1.1 exaflops Rmax in May 2022, making it the world's fastest supercomputer as measured in the June 2022 edition of the TOP500 list, replacing Fugaku.[1][17]
Upon its release, the supercomputer topped the Green500 list for most efficient supercomputer, measured at 62.68 gigaflops/watt.[6] ORNL Director Thomas Zacharia said: "Frontier is ushering in a new era of exascale computing to solve the world’s biggest scientific challenges." He added: "This milestone offers just a preview of Frontier’s unmatched capability as a tool for scientific discovery. It is the result of more than a decade of collaboration among the national laboratories, academia and private industry, including DOE's Exascale Computing Project, which is deploying the applications, software technologies, hardware and integration necessary to ensure impact at the exascale."[14]
References
edit- ^ a b c "TOP500 June 2024". May 14, 2024. Retrieved May 16, 2024.
- ^ Wells, Jack (March 19, 2018). "Powering the Road to National HPC Leadership". OpenPOWER Summit 2018. Archived from the original on August 4, 2020. Retrieved March 25, 2018.
- ^ Bethea, Katie (February 13, 2018). "Frontier: OLCF'S Exascale Future – Oak Ridge Leadership Computing Facility". Oak Ridge National Laboratory - Leadership Computing Facility. Archived from the original on March 10, 2018.
- ^ "DOE Under Secretary for Science Dabbar's Exascale Update". insideHPC. October 9, 2020. Archived from the original on October 28, 2020.
- ^ Don Clark (May 30, 2022). "U.S. Retakes Top Spot in Supercomputer Race". The New York Times. Archived from the original on June 1, 2022. Retrieved June 1, 2022.
- ^ a b c Larabel, Michael (May 30, 2022). "AMD-Powered Frontier Supercomputer Tops Top500 At 1.1 Exaflops, Tops Green500 Too". Phoronix. Archived from the original on June 6, 2022. Retrieved June 1, 2022.
- ^ Anton Shilov (November 15, 2022). "Nvidia Steals AMD's Supercomputer Efficiency World Record". Tom's Hardware. Archived from the original on February 6, 2023. Retrieved December 10, 2022.
- ^ a b c d Choi, Charles Q. (June 24, 2022). "The Beating Heart of the World's First Exascale Supercomputer". IEEE Spectrum. Archived from the original on August 14, 2022. Retrieved August 14, 2022.
- ^ "Crusher Quick-Start Guide — OLCF User Documentation". docs.olcf.ornl.gov. Archived from the original on November 8, 2022. Retrieved November 8, 2022.
- ^ "FRONTIER Spec Sheet". Archived from the original on May 31, 2022. Retrieved May 31, 2022.
- ^ a b Service, Robert F. (November 17, 2023). "World's fastest supercomputers are helping to sharpen climate forecasts and design new materials". Science. Retrieved September 4, 2024.
- ^ "Frontier User Guide". Oak Ridge National Laboratory. Retrieved September 4, 2024.
- ^ "AMD Preparing More Linux Code For The Frontier Supercomputer". Archived from the original on May 28, 2021.
- ^ a b "Frontier supercomputer debuts as world's fastest, breaking exascale barrier". Oak Ridge National Laboratory. May 30, 2022. Archived from the original on June 1, 2022. Retrieved June 17, 2022.
- ^ "US Closes in on Exascale: Frontier Installation Is Underway". HPC Wire. September 29, 2021. Archived from the original on January 3, 2022. Retrieved January 3, 2022.
- ^ "First Look At Oak Ridge's "Frontier" Exascaler, Contrasted To Argonne's "Aurora"". Next Platform. October 4, 2021. Archived from the original on January 3, 2022. Retrieved January 3, 2022.
- ^ "US Takes Supercomputer Top Spot with First True Exascale Machine". Archived from the original on May 31, 2022. Retrieved May 31, 2022.