China unveils its fastest supercomputer-Tianhe

Koji

Regular Member
Joined
May 24, 2009
Messages
758
Likes
1
Don't knock the $hit G, when you don't know it. India's R&D may be behind China's at the moment, but it's certainly far from "absoloutely pathetic". Check this out for a change:

TOP SUPERCOMPUTERS-INDIA


Supercomputers are developing, and fast- year on year. Indian supercomputers have regularly made the list in the last few years. A ranking in late 2007 saw 9 Indian systems on the top 500 supercomputers list, with 10 systems for China, 11 for Taiwan and 20 for Japan:

India supercomputer makes top-10 'most powerful' list


This may seem to represent a 'quantum' leap for the moment (literally eh?) but back in 2004, there was news of Japan working on a ten- petaflop supercomputer that could process 10 quadrillion calculations (10 x 10^15 flops/sec) per second, that is due in the 2012 timeframe and will take supercomputers to a whole new level.


Prior to China's one-quadrillion teraflops per second (1 x 10^15 flops/sec) supercomputer, India's 'eka', with an Rpeak of 172.6 teraflops (172.60 x 10^12 flops/sec) was the "fastest in Asia and the 4th fastest in the world". Hit this up for size:

Welcome to Computational Research Laboratories
Current rankings indicate that no Indian computer comes close to the top.
 

Rage

DFI TEAM
Senior Member
Joined
Feb 23, 2009
Messages
5,419
Likes
993
Current rankings indicate that no Indian computer comes close to the top.

Yeh, yeh we know sino-phile.

Do you get the gist of what I wrote?

Supercomputer rankings are changing all the time - and very fast: every six months or so, to be approximate. The rankings in November 2007 indicated that no supercomputer from China came anywhere close to the top.

Which is why "rankings" are always transient and temporary.

Get that, G?
 

sklow84

New Member
Joined
Nov 14, 2009
Messages
1
Likes
0
I was studying computer architecture when you were learning your ABC. This post will circulate amongst my friends as a hoarse frothing rant by another CCP drone.

There is no such thing as stability in instruction set design. Your post identifies you as a sheep wearing a wolf's skin (in other words you dont know the basics of computer science and are a google retard).

More complex instructions require complex ALU unit. MIPS (ultimately RISC) is based on a basic instruction set, with more complex instructions implemented in software and emulated (which is then broken down into the RISC instructions). RISC architecture provides a very fundamental advantage to server farms. The instructions are simple and hence the instructions take few cycles to execute. This means that more instructions can be pipelined easily and this can achieve more parallelism per core. Transmeta is a good example. They were in fact trying to take it further by using VLIW instruction computing (morphing multiple instructions into one) by dynamic compilation and heuristics.Now that is what I would call invention. Its not about software compatibility, neither is it about stability. Its about achieving parallelism by pipelining more instructions together in a single core.

In other words, think of RISC as little Chinamen lifting little loads but moving faster in a pipe compared to fat americans lifting heavy load but moving slowly in a pipe. Both need compilers (software typically written by Indians) to make the chinese move faster in the pipeline so that they can give a better efficiency compared to americans moving in the pipeline.



Another retarded argument. Anyone can invent an architecture? Then why is Von-Neuman model the only turing machine ever built? Anyone can invent an instruction set. And intel's instruction set got complicated by the day, which is the reason for this alternative RISC design proposed.

So, coming back to my question. What did China invent here? Zilch, nada, nothing.
I think you are confusing instruction set architecture (ISA) with micro-architecture and the basic computer architecture a.k.a the von neumann architecture. An ISA includes a specification of the set of opcodes (machine language), the native commands implemented by a particular processor. Instruction set architecture is distinguished from the microarchitecture, which is the set of processor design techniques used to implement the instruction set.
The von Neumann architecture is a design model for a stored-program digital computer that uses a processing unit and a single separate storage structure to hold both instructions and data.


For example, you can have an Athlon and Pentium implementing the same x86 ISA but have totally different micro-architecture. So in this case, China did invent its own micro-architecture free from licensing fees.

And yes there are many different types of ISAs around and China choose to implement an almost free and widely used one. They could have made a new way of executing codes but that would be commercially useless.

So what is RISC? Think of it as lil skinny malnourished yindoos carrying pots of cow dung on their head. To make it go faster, their big fat British masters need to whip them hard.

Compilers written by Indians? You do know that the Itanium compiler was created by this same Chinese institute right?

Basically what our dear friend said was not wrong. He was putting it in simple terms. Btw, what does explaining how RISC work got to do with him being wrong? Why doesn't software compatibility and architecture stability count? Please don't change the topic. China did invent their own microprocessor. Just as Athlon and Pentium are different. If it was so easy, why doesn't India have one? WHY? :twizt:

Don't dismiss China's effort easily, it took China 3 generations of chip design to only reach a pentium 4 1.5 gig level. Imagine how long it will take India? Same goes for manned space. You are planning to launch it 12 years after China and that too with a redesigned Soyuz. If China were using a Soyuz clone, we wouldn't need to test it 5 times for 5 years before using it for humans. India can just happily use it on the first try. Hmmm suspicious.

First read on the Godson architecture
HC20.26.620.Micro-Architecture of Godson-3 Multi-Core Processor
 

RAM

The southern Man
Senior Member
Joined
Jul 15, 2009
Messages
2,282
Likes
414
Country flag
China's "Milky Way One" supercomputer ranked fifth in world



According to the latest list of the world's 500 fastest supercomputers released Nov. 18, 2009, the "Milky Way One" ranked fifth in the world and first in Asia. "Milky Way One" is China's first teraflop supercomputer cluster system, developed by the Computer Science Department of National Defense University.

A Chinese supercomputer in the top ranks indicates that China already has high-performance calculation technology and becomes the second country with the ability to develop a teraflop supercomputer system.

The International Top500 organization releases world's supercomputers list twice annually based on Linpack testing. This organization aims to promote exchange and cooperation in supercomputing fields and spread its application.

"Milky Way One" is China's first teraflop supercomputer system, with a peak speed of 1,206 trillion functions per second. Linpack testing reaches 563.1 trillion functions per second.

Advanced supercomputing technology is a symbol of technological competitiveness especially in the information age. Among the top ten listed, there are 9 American produced supercomputers and the "Milky Way One". Experts believe that "Milky Way One" is a way of showing the uprising status of China's advanced computer technology and it will provide solutions for critical economic and technological issues.

As the core processor of National Supercomputing Tianjin Center, "Milky Way One" is expected to provide technological support for petroleum exploration data processing, bio-medical research, aviation equipment development, resources exploration, satellite remote sensing data processing, financial engineering, weather forecasting, climate prediction, marine data analysis, earthquake prevention, new material development and other important areas.

http://english.people.com.cn/mediafile/200911/18/P200911181642052280083276.jpg
 

arya

Senior Member
Joined
Sep 14, 2009
Messages
3,006
Likes
1,527
Country flag
hi


good work by china but we can do much better

jai hind
 

proud_hindustani

Regular Member
Joined
Nov 8, 2009
Messages
247
Likes
4
Country flag
Well Done.It's unbeleiveable that they can develope such kind of machine.
Now It's pakis' turn to develop such computers. they will beg from china for technology.

they will show to the world with claims " ghaas khayi hai magar yeh technology bhi banane me kaamyab ho gaye hai"
 

ajtr

Tihar Jail
Banned
Joined
Oct 2, 2009
Messages
12,038
Likes
721
China Wrests Supercomputer Title From U.S.


A Chinese scientific research center has built the fastest supercomputer ever made, replacing the United States as maker of the swiftest machine, and giving China bragging rights as a technology superpower.The computer, known as Tianhe-1A, has 1.4 times the horsepower of the current top computer, which is at a national laboratory in Tennessee, as measured by the standard test used to gauge how well the systems handle mathematical calculations, said Jack Dongarra, a University of Tennessee computer scientist who maintains the official supercomputer rankings.

Although the official list of the top 500 fastest machines, which comes out every six months, is not due to be completed by Mr. Dongarra until next week, he said the Chinese computer "blows away the existing No. 1 machine." He added, "We don't close the books until Nov. 1, but I would say it is unlikely we will see a system that is faster."

Officials from the Chinese research center, the National University of Defense Technology, are expected to reveal the computer's performance on Thursday at a conference in Beijing. The center says it is "under the dual supervision of the Ministry of National Defense and the Ministry of Education."

The race to build the fastest supercomputer has become a source of national pride as these machines are valued for their ability to solve problems critical to national interests in areas like defense, energy, finance and science. Supercomputing technology also finds its way into mainstream business; oil and gas companies use it to find reservoirs and Wall Street traders use it for superquick automated trades. Procter & Gamble even uses supercomputers to make sure that Pringles go into cans without breaking.

And typically, research centers with large supercomputers are magnets for top scientific talent, adding significance to the presence of the machines well beyond just cranking through calculations.

Over the last decade, the Chinese have steadily inched up in the rankings of supercomputers. Tianhe-1A stands as the culmination of billions of dollars in investment and scientific development, as China has gone from a computing afterthought to a world technology superpower.

"What is scary about this is that the U.S. dominance in high-performance computing is at risk," said Wu-chun Feng, a supercomputing expert and professor at Virginia Polytechnic Institute and State University. "One could argue that this hits the foundation of our economic future."

Modern supercomputers are built by combining thousands of small computer servers and using software to turn them into a single entity. In that sense, any organization with enough money and expertise can buy what amount to off-the-shelf components and create a fast machine.

The Chinese system follows that model by linking thousands upon thousands of chips made by the American companies Intel and Nvidia. But the secret sauce behind the system — and the technological achievement — is the interconnect, or networking technology, developed by Chinese researchers that shuttles data back and forth across the smaller computers at breakneck rates, Mr. Dongarra said.

"That technology was built by them," Mr. Dongarra said. "They are taking supercomputing very seriously and making a deep commitment."

The Chinese interconnect can handle data at about twice the speed of a common interconnect called InfiniBand used in many supercomputers.

For decades, the United States has developed most of the underlying technology that goes into the massive supercomputers and has built the largest, fastest machines at research laboratories and universities. Some of the top systems simulate the effects of nuclear weapons, while others predict the weather and aid in energy research.

In 2002, the United States lost its crown as supercomputing kingpin for the first time in stunning fashion when Japan unveiled a machine with more horsepower than the top 20 American computers combined. The United States government responded in kind, forming groups to plot a comeback and pouring money into supercomputing projects. The United States regained its leadership status in 2004, and has kept it, until now.

At the computing conference on Thursday in China, the researchers will discuss how they are using the new system for scientific research in fields like astrophysics and bio-molecular modeling. Tianhe-1A, which is housed in a building at the National Supercomputing Center in Tianjin, can perform mathematical operations about 29 million times faster than one of the earliest supercomputers, built in 1976.

For the record, it performs 2.5 times 10 to the 15th power mathematical operations per second.

Mr. Dongarra said a long-running Chinese project to build chips to rival those from Intel and others remained under way and looked promising. "It's not quite there yet, but it will be in a year or two," he said.

He also said that in November, when the list comes out, he expected a second Chinese computer to be in the top five, culminating years of investment.

"The Japanese came out of nowhere and really caught people off guard," Mr. Feng said. "With China, you could see this one coming."

Steven J. Wallach, a well-known computer designer, played down the importance of taking the top spot on the supercomputer rankings.

"It's interesting, but it's like getting to the four-minute mile," Mr. Wallach said. "The world didn't stop. This is just a snapshot in time."

The research labs often spend weeks tuning their systems to perform well on the standard horsepower test. But just because a system can hammer through trillions of calculations per second does not mean it will do well on the specialized jobs that researchers want to use it for, Mr. Wallach added.

The United States has plans in place to make much faster machines out of proprietary components and to advance the software used by these systems so that they are easy for researchers to use. But those computers remain years away, and for now, China is king.

"They want to show they are No. 1 in the world, no matter what it is," Mr. Wallach said. "I don't blame them."
 

Typhoonq

New Member
Joined
Oct 29, 2010
Messages
1
Likes
0
:emot154::emot154:
I was studying computer architecture when you were learning your ABC. This post will circulate amongst my friends as a hoarse frothing rant by another CCP drone.

There is no such thing as stability in instruction set design. Your post identifies you as a sheep wearing a wolf's skin (in other words you dont know the basics of computer science and are a google retard).

More complex instructions require complex ALU unit. MIPS (ultimately RISC) is based on a basic instruction set, with more complex instructions implemented in software and emulated (which is then broken down into the RISC instructions). RISC architecture provides a very fundamental advantage to server farms. The instructions are simple and hence the instructions take few cycles to execute. This means that more instructions can be pipelined easily and this can achieve more parallelism per core. Transmeta is a good example. They were in fact trying to take it further by using VLIW instruction computing (morphing multiple instructions into one) by dynamic compilation and heuristics.Now that is what I would call invention. Its not about software compatibility, neither is it about stability. Its about achieving parallelism by pipelining more instructions together in a single core.

In other words, think of RISC as little Chinamen lifting little loads but moving faster in a pipe compared to fat americans lifting heavy load but moving slowly in a pipe. Both need compilers (software typically written by Indians) to make the chinese move faster in the pipeline so that they can give a better efficiency compared to americans moving in the pipeline.



Another retarded argument. Anyone can invent an architecture? Then why is Von-Neuman model the only turing machine ever built? Anyone can invent an instruction set. And intel's instruction set got complicated by the day, which is the reason for this alternative RISC design proposed.

So, coming back to my question. What did China invent here? Zilch, nada, nothing.
You are simply jealous of China achievement. Inspite of China being a late comer in making Super Computer she has caught up with the best. What surprises me is that inspite of India claimed to be the super power in IT engineering and Software, she can't even reach World top 30 fastest Super Computer.
Actually this involves great amount of software input and the Chinese could have dominated the World if not of their weaknesses in the English Language.

For example, for the last 10 years the Chinese dominated in most of the Programming and Coding competition organised by International Olympia Informatics and International Collegiate Programming Contest.
On the other hand, Not a single Indian participant or University has achieved any championship.

The most recent one is the biggest embarrassement of all. Inspite of sending in the 2nd largest group of contestants sponsored by US National Security Agency for the Contest in Nov 2009, not a single Indian entered the Grand Final.
Guess what, the Chinese again dominated. Of the 70 finalists, 20 were Chinese, 10 Russian, 2 American.

The champion was a 18 year old undergraduate from Shanghai Jiatong University.
This university has the honour of Winning the most in ICPC contest.
 

Minghegy

Regular Member
Joined
Jan 13, 2010
Messages
387
Likes
9
I believe China and U.S. and other counties have much faster supercomputers than it, but public don't know them.
 

Dark Sorrow

Respected Member
Senior Member
Joined
Mar 24, 2009
Messages
2,475
Likes
2,936
Country flag
I am no programming guru so I would appreciate if you could explain to me how these supercomputers are tested for the global rankings.
These supercomputers have several testing methods
(1) Small Integer, Integer and Big Integer numbers are addes, subtracted, multiplied, divided, etc.(several such mathematical operations are performed) and exection rate and time is studied. This procedure is also carried out on float, double and double double(a very very large double) number. A new data type called vairent is also used which doesnot put restriction on size of number.
(2) The architecture of the GPU and CPU is studied. Core with RISC acrchitecture are prefered.
(3) Main memory refresh rate and transfer rate is also studied. Their are prcedure for ramdom copy, add, transfer, delete, etc. data from memory, this procedure is studied.
(4) Architecture of the computer as a whole is studied as well as with respect to the application for which it was built.
(5) The OS is studied and ts efficiency w.r.t. the system is studied.
(6) The methodlogy used to achieve parallelism is studied and its efficiency is observed.
There are many such procedures but it is not poassible to state all.
 

cir

Senior Member
Joined
Dec 28, 2010
Messages
1,996
Likes
269
And with these sprouting up around the nation,more and more advanced designs and cutting edge technologies are a surety:rofl::

Supercomputer in full operation
Updated: 2011-11-17 10:33By Han Ximin (Shenzhen Daily)

Supercomputer Dawn 6000, which performs 1.27 quadrillion (1.27 petaflops) calculations per second at peak speed, was in full operation at the National Supercomputing Center in Shenzhen yesterday.

The system was developed jointly by Dawning Information Industry Co Ltd and the Calculation Institute of the Chinese Academy of Sciences. It is the fourth-fastest computing system in the world.

Li Jun, president of Dawning Information Industry Co Ltd, said Dawn 6000 was the most advanced technology in the Dawning family. At a launching ceremony yesterday, he said it would be mainly used for high-speed calculations, cloud computing, and key national projects and research.

"The supercomputer can help reduce operational costs and improve the efficiency of local companies," said Zhang Haizhong, leading engineer at the center.

The National Supercomputing Center, which was established in 2009 and cost 800 million yuan ($127 million), was the key science and technological facility in Guangdong and the largest investment in science and technology in Shenzhen.

Shenzhen was one of the five cities in China selected by the National Development and Reform Commission to pilot cloud computing, according to a plaque presented at the Shenzhen Convention and Exhibition Center yesterday.

Other four cities are Beijing, Shanghai, Hangzhou and Wuxi.

A total of 21 cities including Beijing, Tianjin, Shanghai and Chongqing were also selected to pilot e-commerce.

Shenzhen Mayor Xu Qin said the production volume of the new-generation technological industry would be 1 trillion yuan by the end of 2015.

Background

There are four supercomputing centers in China in Tianjin, Shenzhen, Changsha and Ji'nan.

Tianhe-1A, the first petaflop computer in China which was installed in Tianjin, has been providing cloud computing for exploration, aircraft simulation, biomedicine, cartoon design and financial risk analysis.

High performance computers were also adopted in Changsha designed to operate at 1 petaflop per second.

The National Supercomputing Center in Jinan was inaugurated last month. It is the first petaflop computing center in China to adopt domestic CPU and systems software.



China to Build More Supercomputers With Homegrown Chips

China to Build More Supercomputers With Homegrown Chips | PCWorld Business Center
 

Global Defence

New threads

Articles

Top