Aurora Supercomputer With Intel Ponte Vecchio Fails To Beat All-AMD Frontier In Exascale Race

Photo of author
Written By Editor

Who keeps posting articles without emotional mental changes

Yesterday, Top500.org launched the list of the fastest supercomputers on earth and it appears like Intel's Aurora supercomputer stopped working to beat AMD's Frontier, the only exaflop maker that's totally functional.

Ignore Zettascale, Intel Is Yet To Achieve Exascale Performance With Aurora Supercomputer & Ponte Vecchio Chips, All-AMD Frontier The Only Exaflop+ System On The Planet

Intel's Ponte Vecchio accelerators powering the Aurora Supercomputer at the Argonne Leadership Computing Facility in Illinois, USA are now functional however the guaranteed calculate efficiency is far from attainable as the system is running at a quarter of its anticipated calculate output. Both in regards to raw efficiency and raw effectiveness, the AMD Frontier system beats Intel's Aurora system.

Associated Story AMD Readies More Ryzen 5000 3D V-Cache CPUs For AM4: Ryzen 7 5700X3D & Ryzen 5 5500X3D

Following are the leading 3 supercomputers on earth:

  • Frontier stays the No. 1 system in the TOP500. This HPE Cray EX system is the very first United States system with an efficiency surpassing one Exaflop/s. It is set up at the Oak Ridge National Laboratory (ORNL) in Tennessee, USA, where it is run by the Department of Energy (DOE). It presently has actually attained 1.194 Exaflop/s utilizing 8,699,904 cores. The HPE Cray EX architecture integrates 3rd Gen AMD EPYC CPUs enhanced for HPC and AI, with AMD Instinct 250X accelerators and a Slingshot-11 adjoin.
  • Aurora attained the No. 2 area by sending an HPL rating of 585 Pflop/s determined on half of the complete system. It is set up at the Argonne Leadership Computing Facility, Illinois, USA, where it is likewise run by the Department of Energy (DOE). This brand-new Intel system is based upon HPE Cray EX - Intel Exascale Compute Blades. It utilizes Intel Xeon CPU Max Series processors, Intel Data Center GPU Max Series accelerators, and a Slingshot-11 adjoin.
  • Eagle the brand-new No. 3 system is set up by Microsoft in its Azure cloud. This Microsoft NDv5 system is based upon Xeon Platinum 8480C processors and NVIDIA H100 accelerators and accomplished an HPL rating of 561 Pflop/s.

So beginning with the information, the Aurora Supercomputer utilizes a mix of Intel Xeon Max"4th Gen Sapphire Rapids"and Data Center Max GPUs"Ponte Vecchio". The system includes an overall of 4.742 Million cores which's method less than the last figures that were anticipated in the system. The system outputted a peak FP64 calculate efficiency of 585.34 PFLOPs which is less than half the peak efficiency(Rmax)of the AMD-based Frontier which is the only exaflop-capable device as it strikes 1,194 PFLOPs. The system has actually been functional because 2022. In regards to effectiveness, the Aurora supercomputer peaks out with a Rpeak of 1059.33 PFLOP/s while the Frontier supercomputer peaks out at 1,679.82 PFLOP/s. The Aurora supercomputer has a peak power draw of 24.6 MW however that is the figure for the whole system & not the present setup which is keeping up half of the specs/hardware. The Frontier supercomputer has a power draw of 22.7 MW. R peak worths are computed utilizing the marketed clock rate of the CPU. For the performance of the systems you ought to take into consideration the Turbo CPU clock rate where it uses.

by means of Top500 There's not a surprise that Aurora remains in the state in which it is right now. The system is lastly functional, it is no place close to the 2 Exaflops that Intel had actually assured. This existing circumstance has actually been partially due to the significant hold-ups with Intel's Sapphire Rapids and Ponte Vecchio chips which led AMD to take the very first area in the exascale race and maintain it for an entire year. Intel

Aurora was very first revealed all the method back in 2019 which was 4 years earlier. If you desire to go truly back, then the initial style for the Aurora was set out all the method back in 2015 with an arranged arrival in 2018. At that time, it was a 180 PFLOPs system and based upon the now stopped

Xeon Phi "Knights Hill "platform. It is not understood when we'll see Aurora lastly dispense its 2 Exaflops of calculate however AMD is currently preparing a 2 Exaflop +system referred to as El-Capitan which must likewise end up being functional in the coming year. That will be a more damage to Intel in the HPC and supercomputing area.2023-11-14_19-29-45

2023-11-14_19-29-48
2023-11-14_19-29-49
2023-11-14_19-29-51
2023-11-14_19-29-52
2023-11-14_19-29-53
2023-11-14_19-29-54
2023-11-14_19-29-55
Categories PC

Leave a Comment