Posted on 06/24/2020 3:33:30 AM PDT by ShadowAce
For years, x86 processors and Linux have ruled supercomputing. Linux still runs 500 out of the TOP500 supercomputers in the world. For just about as long, x86 CPUs have dominated supercomputers -- until now. On June 22, Japan's Fugaku supercomputer, powered by Fujitsu's 48-core A64FX SoC and running Red Hat Enterprise Linux (RHEL), became the first ARM-powered supercomputer to be dubbed the fastest computer in the world.
In winning over the others, it wasn't even close. Fugaku turned in a High-Performance Linpack (HPL) result of 415.5 petaflops, besting the second-place IBM Summit system by a factor of 2.8x.
Fujitsu has been working on creating ultra-high-speed silicon since it turned from the obsolete SPARC architecture to the ARM-based A64FX. This was the first CPU to adopt ARMv8-A's Scalable Vector Extension (SVE). This is an ARM extension to its instruction set specifically for supercomputers. Fujitsu worked with ARM to develop the A64FX.
As fast as it is, in single or further reduced precision, which is used in machine learning and AI applications, Fugaku's peak performance made it the first supercomputer to break the 1,000 petaflops -- or 1 exaflop -- barrier. Fugaku is installed at the RIKEN Center for Computational Science (R-CCS) in Kobe, Japan.
AMD and Cray are working with the US Department of Energy along with Oak Ridge National Laboratory to build a supercomputer, Frontier that uses EPYC CPUs and Radeon Instinct GPUs to officially bust the exaflop wall in 2021. At this point, it appears it will be a race between Fugaku and Frontier to run faster than an exaflop.
For now, Summit -- with its Power9 CPUs and NVIDIA Tesla V100 GPUs -- is the second-fastest supercomputer. It delivers 148.8 petaflops on HPL.
At No. 3, we find Lawrence Livermore National Laboratory (LLNL)'s Sierra with 94.6 petaflops on HPL. Like Summit, it uses Power9 CPUs and NVIDIA Tesla V100 GPUs. China's Sunway TaihuLight is No. 4, with an HPL mark of 93 petaflops. It's powered entirely by Sunway 260-core SW26010 processors. At No. 5, we find China's Tianhe-2A (Milky Way-2A), with HPL performance of 61.4 petaflops coming from a hybrid architecture combining Intel Xeon CPUs and custom-built Matrix-2000 coprocessors.
Intel's x86 architecture is the heart of only one of the top five supercomputers. But it continues to be the top processor architecture. You'll find it in 481 of the TOP500 systems. Of these, Intel claims 469, with AMD installed in 11, and Hygon in the remaining one. ARM processors are present in just four TOP500 systems, three of which use the new Fujitsu A64FX processor, with the other powered by Marvell's ThunderX2 processor. But being No. 1 counts for a lot.
China continues to dominate the TOP500 in systems, as 226 Chinese supercomputers are on the list. The US is No. 2 with 114 systems; Japan is third with 30; France has 18; and Germany claims 16. In terms of raw computing power, the US edges out China in aggregate list performance, with 644 petaflops to 565 petaflops. Small-but-powerful Japan, despite its much smaller system count, delivers an aggregate 530 petaflops.
As you'd expect from that breakdown, Chinese manufacturers dominate the list: Lenovo (180), Sugon (68), and Inspur (64) account for 312 of the 500 installed systems. Of US vendors, HPE claims 37 systems, while Cray/HPE has 35 systems. Fujitsu is represented by just 13 systems, but thanks to Fugaku win, the company leads the list in aggregate performance with 478 petaflops. Lenovo took second in performance with 355 petaflops.
Finally, as has been true since 2017, all 500 of the world's fastest supercomputers run Linux. When it comes to high-performance, it's a Linux world.
Hadn't thought of that, but you have a point there.
Looked it up to refresh my own memory.
The beast of a system weighed ~ 57,089 pounds.
Took up ~ 3,752 ft2 of floor space, with another 2,450 ft2 room for power, refrigeration, & auxiliary equipment.
It used 21 types of vacuum tubes, but not for memory, after all.
It had [at most] 12,288 words of 36-bit magnetic core memory.
Also [at most] 32,768 words of magnetic drum memory.
Up to 24 tape drives could be connected.
Took an hour to boot up: 15 minutes at quarter voltage, another 15 minutes at half-voltage, etc.
Two trucks arrived every week -- one to haul away the burnt-out tubes; another with new tubes.
We wrote programs in Basic Assembly Language [BAL]. Way before [Dartmouth] BASIC.
12 years later, we [Kemosabe] were booting up 16-bit minicomputers with fanfold paper tape.
The fantastic notion that everyone would one day be walking around with his nose glued to a handheld pocket-size touch-sensitive screen -- with embedded high-res cameras -- sending full-color video clips to strangers around the globe via satellite, and storing terabytes of data in a faraway distributed "cloud" of servers, all at nearly zero customer cost, was outrageous, laughable science fiction.
How far we have come.
The internet & worldwide web [combined] is one of the greatest human achievements in history.
You -- of all FReepers -- certainly appreciate the historical & cultural perspectives.
Hard to believe anyone got any work done without the ubiquitous internet.
Still, much of today's chaos is the result of placing significant computing power in the hands of mere civilians.
——
I learned LISP from John McCarthy himself. He was our Professor in every LISP class offered.
Got the LISP 1.5 Programmer's Manual [1965] right here on my reference shelf.
John discovered / designed LISP, publishing his paper "Recursive Functions of Symbolic Expressions and Their Computation by Machine" in Communications of the ACM, April 1960.
He worked at M.I.T., but was eager to start his own research Lab.
He negotiated a deal with Stanford to move west, and create & run the Stanford Artificial Intelligence Laboratory (SAIL), in exchange for agreeing to teach LISP part-time to graduate students.
I was a Research Assistant at SAIL [1970 -- 71], while pursuing a Master's Degree in Computer Science.
My Professors in other CS courses were Don Knuth & Alan Kay.
Alan had a sketchbook, and explained to us his ideas & diagrams for a portable, folding computer that evolved into the MacBook.
We often worked late into the night, programming [a DEC PDP 10] in LISP & Stanford Artificial Intelligence Language [also SAIL]. It had an ALGOL-like structure, with extensions like Pieces of Glass that were essentially translucent windows.
Inventing the future.
John McCarthy, Don Knuth, Alan Kay -- holy ****! Of the Mac, Kay once said, it was the first computer worthy of criticism. :^) I'm not sure what the first computer I used was, I only know that, if I drove to, say, [favorite ugly city, fill it in yourself] and on the way there I saw an old pole building where a fire had taken out about the rear third of it and the wall on that side was a huge piece of faded translucent plastic, I'd be looking at exactly the right facility to put a museum that featured that computer system. And I'm in the ballpark of ten years younger than you (estimate), by contrast, I had it *easy*. :^) One of my sisters studied FORTRAN (I'm still not sure why, although she was a math major, that's probably it) and the only assignment I can recall from that dim past was a Snoopy in profile over a calendar, all done in text characters, line printer, wide paper.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.