What's new

A Multicore Processor Designed For PetaFLOPS Computation

My previous post shows that China possesses world-class chip design (e.g. only 30% slower than Intel's very best). The following article shows China's world-leading supercomputer integration software.

New York Times: "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."

Who are you going to believe? The New York Times (e.g. "China is a [computer] technology superpower") or ao333 (e.g. China is soooo far behind).

http://www.nytimes.com/2010/10/28/technolo...rss&emc=rss

"Chinese Supercomputer Wrests Title From U.S.
By ASHLEE VANCE
Published: October 28, 2010

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.

chinatianhe1a.jpg

The Tianhe-1A computer in Tianjin, China, links thousands upon thousands of chips.

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.'”
 
Last edited:
Get this into your head:

China's processor design is a decade behind the US
China's fabs are 3 decades behind the US.

Conclusion?

Chinese designed CPUs will be used in servers and portable devices.
No Chinese fabbed CPU will be commercially viable for another 20 years.

Let’s assume that there is a big gap in both design and fab of CPUs between China and the US, but I doubt that 20 year-gap you put. Furthermore I doubt that the Taiwanese chip makers is inferior to Korean ones in the pecking order.

OK, the following are the China's Surprises to catch up:

Surprise No 1: Sheer high IQ


The size of China’s population + High IQ ensures China possesses largest amount of people in the world with IQ>160. Yes, IQ matters ! :D

Surprise No 2: the sheer size of Chinese market

According to the PwC recent report on China’s chip market (evertiq.com :: Components - China's impact on the chip industry ), China’s chip market growth in recent year is just stunning, and the trend is still steeply upwards.

Quote the PwC research - “ It has become clear that market growth in China is far more significant to the worldwide semiconductor industry than the nation’s production volumes… China has become a dominating consumer of semiconductors. China’s consumption growth continuously outrun the rest of the world” .

We all know with that size, all top quality players are there which eventually will forge a world-class chip environment for domestic design & fab

Surprise No 3. China’s exponential IPO market for start-up chip makers


Quote the PwC research - “ Chinese domiciled companies the third largest group of semiconductor IPOs completed worldwide between 2005 and 2009… during the last four quarters, as the semiconductor industry recovered from the global recession, Chinese companies have accounted for more than half of all semiconductor IPOs completed worldwide. Moreover, Chinese financial markets have provided for more than 80% of all the semiconductor IPO funding.”

I see that most of the start-ups are insignificant to challenge world’s top players, yet some of the entrepreneurs will eventually make it to big time. It’s just a matter of maths.

Surprise No 4 China’s financial might

China’s financial might (4 of the top 10 biggest banks in the world) in funding of the industry, both domestic and worldwide, will eventually pay off , not only financially, but also in term of in terms of power in the semiconductor worldwide industry as Money = King.

Quote the PwC research - “During the past five years, China has emerged as a significant source of new semiconductor companies and, more recently, of financial funding for start-ups.”


Surprise No 5 China’s returnees.

A large amount of Chinese engineers in chip design and fab have been educated in the US and stayed there to work for top houses there. With the takeoff of China’s economy, some of them would be attracted back to China in the future; and some of them would likely to have world class expertise to help China’s indigenous chip makers.

Surprise No 6 Industrial Espionage

We all know that China excels at it. Taiwanese and Singaporean chip makers are all ethnic Chinese. That helps!

Surprise No 7 Proven empirical successful experiences

When needed desperately, China made her N bomb, hydrogen bomb, Nuclear sub, ICBMs, etc in a very short space of time at late 60s, making her one of the top players in the fields.

Now, correct me if I am wrong here, but chip design & fab as I know it are the core and tech pinnacle of semiconductor industry. The industry is likely to be the strategic core R&D target of China. With China's mass of high IQers coupled with abundant funding, I don’t see why that gap could not be bridged in , say, about 10 years time.


Now try to get those into your head! :toast_sign:
 
Last edited:
Let’s assume that there is a big gap in both design and fab of CPUs between China and the US, but I doubt that 20 year-gap you put. Furthermore I doubt that the Taiwanese chip makers is inferior to Korean ones in the pecking order.

Surprise No 1: Sheer high IQ

The size of China’s population + High IQ ensures China possesses largest amount of people in the world with IQ>160. Yes, IQ matters ! :D

I know Taiwan seems high tech in China. But it's insignificant compared to South Korea on the global stage.

That was a measure of urbaners in China, which are concentrated elites from the whole country, unlike the West which is not so. Don't extrapolate that to 1.3 billion ppl, lol...

Surprise No 2: the sheer size of Chinese market

We've been here before. Market for tech doesn't work.

Surprise No 3. China’s exponential IPO market for start-up chip makers

Surprise No 4 China’s financial might

I don't think the West is so cash-stripped to let China invest in strategic projects.

Surprise No 5 China’s returnees.

Surprise No 6 Industrial Espionage

The West has known this for a long time. As such Chinese rarely make it to the top these days.

Surprise No 7 Proven empirical successful experiences

That was due to 10 years of Soviet nuclear aid. Semiconductor is a completely knew industry to China.
 
I know Taiwan seems high tech in China. But it's insignificant compared to South Korea on the global stage.

This troll can't stop putting his foot in his mouth. He keeps spouting garbage and lives in his own fantasy world.

I have already posted articles from MIT's Technology Review and the New York Times on China's computer technology prowess. If you want to believe the troll instead, I can't help you.

The problem with the troll is that he keeps making claims without backing them up with newslinks. He just makes one wild claim after another. The latest claim by the troll is that "Taiwan is insignificant compared to South Korea on the global stage." This claim is laughable and I will prove it with facts.

Taiwan has 22,247 more patents granted by the U.S. Patent and Trademark Office than South Korea. Who's the technology giant again?

Taiwan's AU Optronics wins big patent dispute against Korean LG Display

Foolish Korean electronics giant LG Display falsely accused Taiwan's AU Optronics of infringing four key patents in the manufacture of liquid crystal displays (i.e. LCDs). An U.S. court found that the opposite was true. Korean LG Display had infringed the four key patents that belonged to Taiwan's AU Optronics.

Total U.S. patents granted by the U.S.P.T.O. (i.e. U.S. Patent and Trademark Office) to Taiwan is a cumulative 94,579 patents. For South Korea, it is only 72,332 patents (See Patents By Country, State, and Year - All Patent Types (December 2009)). South Korean giants should think twice before they make false accusations against Taiwanese companies.

http://www.tomsguide.com/us/Patent-Infring...,news-6673.html

"LG [Display] LCDs May Be Banned Worldwide
6:50 PM - May 3, 2010 - By Kevin Parrish - Source : Tom's Guide US

A patent infringement case may block the sale of certain LG LCD panels.

ZoomComputerworld reports that Taiwan-based AU Optronics (AUO) is trying to halt the import and sale of LG Display LCD panels across the globe. If an injunction is successful, this could ultimately hurt consumers and their choice of LCD options, as LG currently commands over a quarter of the LCD panel market.

Over the past three and a half years, LG and AUO have been in a legal scuffle in regards to patents covering material and processes used in making LCD panels. Friday marked the end of the long, multifaceted battle, with AUO emerging as the winner based on LG's inability to prove that the rival company infringed on its LCD patents.

But in February AUO filed a counter-suit and won. Judge Joseph J. Farnan Jr. said in a 77-page verdict that AUO provided enough evidence to show that LG literally infringes on patents asserted by AUO--LG was unable to prove otherwise. Now AUO is warning consumers not to purchase "unauthorized infringing products from LG for sale or use in the U.S. without the need for further court action."

Computerworld said that LG may file further appeals or motions in its case, or it may reach a settlement with AUO that may prevent the possible ban of LG-based LCD screens. As it stands now, over the last six months, LG has shipped more LCD screens across the globe than any other LCD manufacturer. A cease in that kind of distribution could however allow lesser-known brands to step forward and fill the void."
 
Last edited:
Are you an idiot?

...

I know it's your party's policy to not give in to foreign pressure. But you're overexerting yourself for 50 cents. Maybe it's 60 now given the inflation problem I've been hearing about there?

This troll post is reported! :tdown:
 
I know Taiwan seems high tech in China. But it's insignificant compared to South Korea on the global stage.

Please ignore the troll. He doesn't care about facts and says weird stuff. If you listen to him, you will share his distorted and false view of the world.

I do not have an endless amount of time to keep posting factual news articles to show you that the troll is nuts. Consider yourself forewarned. He's a lunatic with a careless disregard for the truth.

On the world economic stage, Taiwan can make South Korea feel small.

Who has more foreign exchange? Taiwan or South Korea? It's Taiwan.

List of countries by foreign exchange reserves - Wikipedia, the free encyclopedia

World #5 Taiwan: $380 billion U.S. dollars.

World #7 South Korea: $293 billion U.S. dollars.


Who makes more money each year? Taiwan or South Korea? It's Taiwan again.

List of sovereign states by current account balance - Wikipedia, the free encyclopedia

World #13 Taiwan: $33 billion U.S. dollars.

World #32 South Korea: $6 billion U.S. dollars.
 
Last edited:
I know Taiwan seems high tech in China. But it's insignificant compared to South Korea on the global stage.

This moronic claim of yours makes me almost speechless...

Taiwan's patent per cap is arguablely the highest in the world.

btw, with more US-registered patents, Taiwan only has total population half that of SK.

That was a measure of urbaners in China, which are concentrated elites from the whole country, unlike the West which is not so. Don't extrapolate that to 1.3 billion ppl, lol...

What "urbaners"? "elite"? China is a highly homogenous country where high IQ can be extrapolated into almost entire 1.3B :agree: It's more so when her mass education program start to bear fruits.


We've been here before. Market for tech doesn't work.

Been where before?

Any largest end market is ultimately crucial for the corresponding industry as a whole as it defines and leads the global trend in R&D and the profits ($$$$$$$...) thus logically attracts all the top global players .

Therefore the day when China becomes the top end market for high end chips consumption should be very near to the day when China's indigenous chip makers lead the world.

I don't think the West is so cash-stripped to let China invest in strategic projects.

you don't think... doesn't make it a reality. The one who holds the money bag will eventually call the shot.

The West has known this for a long time. As such Chinese rarely make it to the top these days.

China hijacked the entire US internet traffic for 18 min just weeks ago, hello? :lol:

That was due to 10 years of Soviet nuclear aid. Semiconductor is a completely knew industry to China.

Soviet helped more in the mass infrastructure and the basics, not the very top end. So no, it's mainly due to a group of world-class overseas Chinese scientists whom returned to China at 50s and 60s.

yeah, Semiconductor industry is relatively new to China, but hey so is High Speed Railway ... :rofl:

:toast_sign:
 
Ok, just add a little bit different flavor to the discussion.

Why China's Chip Industry Won't Catch America's
September 3, 2009

Chinese semiconductor companies have produced some design wins, but they are still struggling when it comes to any true silicon breakthroughs
By Vivek Wadhwa

In the 1980s, the U.S. was consumed with fear that Japan would become the preeminent power in manufacturing and technology. Those fears never came to pass. Today the same fears are focused on China. The Middle Kingdom appears to be an even more daunting foe, with its enormous foreign reserves, fast-growing economy, oceans of scientists and engineers, and enormous subsidies to high-tech companies. How real is the China threat?

There is no doubt that China is making rapid strides in both infrastructure and technology, but U.S. anxiety of being overtaken by China appears to be misplaced. It takes more than money and might to achieve innovation. This is what I learned when researching the inflated estimates of engineering graduation rates in China and by analyzing its pharmaceutical industry. And this is one of the key findings in a new book titled Chips and Change: How Crisis Reshapes the Semiconductor Industry (MIT Press). Written by professors Clair Brown and Greg Linden of the University of California at Berkeley, the book provides a wealth of information about semiconductor development cycles as well as a fresh and informed look at some of China's key technological capabilities in those realms.

A few years ago, China seemed to be on track to dominate the global semiconductor industry in the same way it currently dominates the electronics manufacturing sector. In 2004, China's most advanced chip manufacturer, SMIC (SMI), went public on the Hong Kong and New York stock exchanges. The next year, two Chinese chip design companies, Actions (ACTS) and Vimicro (VIMC), had successful Nasdaq IPOs. Boosters of China's chip industry said there were hundreds more semiconductor design firms waiting in the wings and many new Chinese chip manufacturers were also starting up.

Barriers to Progress
Five years later, most Chinese chip companies remain unprofitable. Why? A number of interlocking reasons that provide clues as to why training lots of engineers and spending money to subsidize companies and build facilities is not enough to create a successful industry.

Because of China's poor reputation for protecting intellectual property, multinationals have limited technology transfer to China. For instance, chip giant Intel (INTC) is now building a plant in northeastern China but has long delayed locating its most cutting-edge fabrication facilities in China, even though this increases the cost of logistics to supply China-based electronics factories, which are among the biggest consumers of Intel processors.

While Chinese semiconductor companies have produced some design wins, the general perception is that China's design shops are good with common reference designs but struggle to produce true silicon breakthroughs. This is likely a product of the variable quality of engineers produced by China's fledgling academic programs which, while ambitious and well funded, produce quantity at the cost of quality. Likewise, China's pool of MBAs does not have the depth of consistency of MBAs from countries with longer histories of business studies and research. And unlike young engineers and MBAs in Silicon Valley, young Chinese engineers and MBAs lack role models who have taken companies public or created truly innovative products earlier in their careers.

China also suffers from an unhealthy competition for marquee facilities among its local governments. This can lead to overinvestment in favored industries, such as semiconductors. SMIC has been hired by at least two city governments to build and manage chip manufacturing plants for which the municipalities, and not SMIC, will bear the risk. This is occurring at a time when China already has excess chip capacity and idle fabrication lines in a number of semiconductor plants. Such lack of market discipline is critical because semiconductor lines not in use have absorbed billions of dollars in investments that could have been used for far more productive endeavors.

Strong Stimulus
None of this is to say that China won't develop the skills to innovate in the future or won't emerge as a formidable competitor in industries like clean energy, where the U.S. has not yet established an effective beachhead. China has a lot of technological strengths and the advantage of large numbers, and it is making massive investments in this area. It will gain sizable cost and technology advantages and some of her entrepreneurs are bound to be both smart and lucky. The current global recession may even work to the benefit of Chinese companies because the government adopted a large stimulus program and consumer purchasing remains strong. And returnees to China from Silicon Valley have begun to add both seasoning and depth to the Chinese entrepreneurial bench. These returnees have worked for top American companies and know how to compete and innovate.

But will China actually overtake the U.S. and other centers of innovation in the next decade in the most complex areas of value creation such as new designs for semiconductors? To get an answer, it's worth revisiting the U.S. experience with Japan in the 1980s. With fears over the loss of competitiveness hanging like a pall over the U.S., the American government negotiated strict trade agreements with Japan and ponied up hundreds of millions of dollars over a decade to help fund a semiconductor R&D consortium. Oddly enough, neither of those moves proved to be hugely consequential; indeed, the trade agreements were somewhat regressive.

Rather, to compete and survive, U.S. companies changed what they were doing. Intel quit the low-profit, low-innovation field of making commodity memory chips and successfully bet the store on microprocessors. Startups, especially in Silicon Valley, exploited opportunities made possible by advances in design automation and by the contract chip manufacturing pioneered in Taiwan to usher in new forms of semiconductor companies that outsourced the relatively low-margin business of chip production. Today the top three design-only (or fabless) companies—Qualcomm (QCOM), Broadcom (BRCM), and Nvidia (NVDA)—are all based in the U.S., and one of them (Nvidia) was launched by a Taiwanese immigrant to the U.S. These multibillion-dollar companies were all started in the decade after 1985, when the fear of Japan was reaching its peak.

Centralized, Not Spontaneous
And what of Japan's monolithic national technology policies? These policies that favored huge business conglomerates like Hitachi and Toshiba stifled the market for chip startups. As a result, Japan to date has no fabless companies of any size, and fabless ethos never caught hold in a corporate culture where ownership of factories is still viewed as essential. Japanese firms have remained overly inward-looking and dependent on the mature and slow-growing domestic market. With a rapidly aging populace and a stagnant economy, this market has lagged behind the rest of the world. The result has been a distinct lack of competitiveness of Japanese companies in the two biggest boom areas of the past two decades, PCs and cell phones.

China is different from Japan in some key ways but not so different that the lessons of Japan Inc. don't apply. The takeaway from Brown and Linden's book is that our fears of Chinese technology dominance are overblown. Yes, China has produced some successful companies, such as search engine Baidu (BIDU) and telecom equipment maker Huawei. Chinese companies lead the world in the fast-growing sector of multiplayer online games. But a centralized effort to replicate the spontaneous upwelling of creativity and innovation, which has formed and informed the U.S. technology sector in general and the semiconductor sector in particular, is a difficult and unlikely route to global technology hegemony.

Wadhwa is senior research associate at the Labor & Worklife Program at Harvard Law School and executive in residence at Duke University. He is an entrepreneur who founded two technology companies. His research can be found at Vivek Wadhwa ? * Tech Entrepreneur, Academic, Researcher, and Writer. He can be followed on Twitter at vwadhwa.

Why China's Chip Industry Won't Catch America's - BusinessWeek
 
Ok, just add a little bit different flavor to the discussion.

Why China's Chip Industry Won't Catch America's
September 3, 2009

Chinese semiconductor companies have produced some design wins, but they are still struggling when it comes to any true silicon breakthroughs
By Vivek Wadhwa

In the 1980s, the U.S. was consumed with fear that Japan would become the preeminent power in manufacturing and technology. Those fears never came to pass. Today the same fears are focused on China. The Middle Kingdom appears to be an even more daunting foe, with its enormous foreign reserves, fast-growing economy, oceans of scientists and engineers, and enormous subsidies to high-tech companies. How real is the China threat?

There is no doubt that China is making rapid strides in both infrastructure and technology, but U.S. anxiety of being overtaken by China appears to be misplaced. It takes more than money and might to achieve innovation. This is what I learned when researching the inflated estimates of engineering graduation rates in China and by analyzing its pharmaceutical industry. And this is one of the key findings in a new book titled Chips and Change: How Crisis Reshapes the Semiconductor Industry (MIT Press). Written by professors Clair Brown and Greg Linden of the University of California at Berkeley, the book provides a wealth of information about semiconductor development cycles as well as a fresh and informed look at some of China's key technological capabilities in those realms.

A few years ago, China seemed to be on track to dominate the global semiconductor industry in the same way it currently dominates the electronics manufacturing sector. In 2004, China's most advanced chip manufacturer, SMIC (SMI), went public on the Hong Kong and New York stock exchanges. The next year, two Chinese chip design companies, Actions (ACTS) and Vimicro (VIMC), had successful Nasdaq IPOs. Boosters of China's chip industry said there were hundreds more semiconductor design firms waiting in the wings and many new Chinese chip manufacturers were also starting up.

Barriers to Progress

Five years later, most Chinese chip companies remain unprofitable. Why? A number of interlocking reasons that provide clues as to why training lots of engineers and spending money to subsidize companies and build facilities is not enough to create a successful industry.

Because of China's poor reputation for protecting intellectual property, multinationals have limited technology transfer to China. For instance, chip giant Intel (INTC) is now building a plant in northeastern China but has long delayed locating its most cutting-edge fabrication facilities in China, even though this increases the cost of logistics to supply China-based electronics factories, which are among the biggest consumers of Intel processors.

While Chinese semiconductor companies have produced some design wins, the general perception is that China's design shops are good with common reference designs but struggle to produce true silicon breakthroughs. This is likely a product of the variable quality of engineers produced by China's fledgling academic programs which, while ambitious and well funded, produce quantity at the cost of quality. Likewise, China's pool of MBAs does not have the depth of consistency of MBAs from countries with longer histories of business studies and research. And unlike young engineers and MBAs in Silicon Valley, young Chinese engineers and MBAs lack role models who have taken companies public or created truly innovative products earlier in their careers.

China also suffers from an unhealthy competition for marquee facilities among its local governments. This can lead to overinvestment in favored industries, such as semiconductors. SMIC has been hired by at least two city governments to build and manage chip manufacturing plants for which the municipalities, and not SMIC, will bear the risk. This is occurring at a time when China already has excess chip capacity and idle fabrication lines in a number of semiconductor plants. Such lack of market discipline is critical because semiconductor lines not in use have absorbed billions of dollars in investments that could have been used for far more productive endeavors.

Strong Stimulus
None of this is to say that China won't develop the skills to innovate in the future or won't emerge as a formidable competitor in industries like clean energy, where the U.S. has not yet established an effective beachhead. China has a lot of technological strengths and the advantage of large numbers, and it is making massive investments in this area. It will gain sizable cost and technology advantages and some of her entrepreneurs are bound to be both smart and lucky. The current global recession may even work to the benefit of Chinese companies because the government adopted a large stimulus program and consumer purchasing remains strong. And returnees to China from Silicon Valley have begun to add both seasoning and depth to the Chinese entrepreneurial bench. These returnees have worked for top American companies and know how to compete and innovate.

But will China actually overtake the U.S. and other centers of innovation in the next decade in the most complex areas of value creation such as new designs for semiconductors? To get an answer, it's worth revisiting the U.S. experience with Japan in the 1980s. With fears over the loss of competitiveness hanging like a pall over the U.S., the American government negotiated strict trade agreements with Japan and ponied up hundreds of millions of dollars over a decade to help fund a semiconductor R&D consortium. Oddly enough, neither of those moves proved to be hugely consequential; indeed, the trade agreements were somewhat regressive.

Rather, to compete and survive, U.S. companies changed what they were doing. Intel quit the low-profit, low-innovation field of making commodity memory chips and successfully bet the store on microprocessors. Startups, especially in Silicon Valley, exploited opportunities made possible by advances in design automation and by the contract chip manufacturing pioneered in Taiwan to usher in new forms of semiconductor companies that outsourced the relatively low-margin business of chip production. Today the top three design-only (or fabless) companies—Qualcomm (QCOM), Broadcom (BRCM), and Nvidia (NVDA)—are all based in the U.S., and one of them (Nvidia) was launched by a Taiwanese immigrant to the U.S. These multibillion-dollar companies were all started in the decade after 1985, when the fear of Japan was reaching its peak.

Centralized, Not Spontaneous
And what of Japan's monolithic national technology policies? These policies that favored huge business conglomerates like Hitachi and Toshiba stifled the market for chip startups. As a result, Japan to date has no fabless companies of any size, and fabless ethos never caught hold in a corporate culture where ownership of factories is still viewed as essential. Japanese firms have remained overly inward-looking and dependent on the mature and slow-growing domestic market. With a rapidly aging populace and a stagnant economy, this market has lagged behind the rest of the world. The result has been a distinct lack of competitiveness of Japanese companies in the two biggest boom areas of the past two decades, PCs and cell phones.

China is different from Japan in some key ways but not so different that the lessons of Japan Inc. don't apply. The takeaway from Brown and Linden's book is that our fears of Chinese technology dominance are overblown. Yes, China has produced some successful companies, such as search engine Baidu (BIDU) and telecom equipment maker Huawei. Chinese companies lead the world in the fast-growing sector of multiplayer online games. But a centralized effort to replicate the spontaneous upwelling of creativity and innovation, which has formed and informed the U.S. technology sector in general and the semiconductor sector in particular, is a difficult and unlikely route to global technology hegemony.

Wadhwa is senior research associate at the Labor & Worklife Program at Harvard Law School and executive in residence at Duke University. He is an entrepreneur who founded two technology companies. His research can be found at Vivek Wadhwa ? * Tech Entrepreneur, Academic, Researcher, and Writer. He can be followed on Twitter at vwadhwa.

Why China's Chip Industry Won't Catch America's - BusinessWeek

Thank you, gipt, for bringing in some other perspective.

That said, the info conveyed in this peice is nothing new. Long it is, only a few points (as I highlighted) the author tried to make, unsuccessfully. Let me debunk it:

point 1: " It takes more than money and might to achieve innovation. This is what I learned when researching the inflated estimates of engineering graduation rates in China and by analyzing its pharmaceutical industry ".

All it takes to innovate IMO is:

a concentrated centrain amount of talents - high IQer with appropriate education + urgent needs + sh!t load of funding + Time/hard work ( so called "luck" will be trivial)

The author claims it takes more, but what more, he can't explain.
Instead, he brings up China's inflated graduate growth # & his personal anecdotal experiences with pharma - so? those # and personal impression are not the key( see above), far from it in fact.


Point 2 "Because of China's poor reputation for protecting intellectual property"

Not entirely true, for 1) the very top IRs are rather protected no matter how China wants to get. e.g. intel's top secrects. Are they that dumb to let China know? and 2), as I recently knew from UK PM Cameroon's recent trip to China, such a question naturally arised as well regarding China's "bad reputaton on IR portection". A column of British legal experts who operate in China were interviewed by BBC for their views. They actually said that the claim of "China's bad repuatation" has been overblown by the media, because China domestic patents represent a huge chunk of the world in total, with Chinese firms occupy more than 90% of them they are actually quait repepect the IRs as their very own livinghoods are depending on them.

Plus, it's always my opinion that if it's so easy for someone to copy your patent, the patent itself is much less worthy, especially high tech patents. The point is that if China wants to steal it , she will find a way; otherwise, to put up this unsubtantial claim is rather stupid.


Point 3 " This is likely a product of the variable quality of engineers produced by China's fledgling academic programs which, while ambitious and well funded, produce quantity at the cost of quality.Likewise, China's pool of MBAs does not have the depth of consistency of MBAs from countries with longer histories of business studies and research. "

The keys words are "quantity instead of quality" and "consistancoes of MBAs". The mid and low end markets aside, if one thinks China's very top end chip design R&D , as it is the issue we are talking here, is somehow "quantity instead of quality", then the claim is just weird to say the best, because it is simplely ilogical as top top R&D guys from any country don't need to produce low-end toys in order ot make a living. All of them are well funded thus have no need to do such a thing but concentraing on high-end research work.

Secondly, what the fvck MBAs have anything to do with R&D, Project Management aside ( PM itself is rather easy for any high IQer with basic trainign to handle)? Here I see that the author starts to try bullshitting his way out of cluelessness. :partay:

Point 4 "China also suffers from an unhealthy competition for marquee facilities among its local governments" -- again , it has nothing to do with degree of inventiveness of R&D. What's the next point, is this author starting to throw in "because of Mao's little red book", I wonder. :woot:


Point 5 , seems to be the last one, "Japan style : Centralized, Not Spontaneous"

This claim is an absolute bonker!

Management styles, centralised or "spontaneous" ones ( what's the definition of "spontaneous" by the way? :woot: the guy must be thinking that he is dating, :lol:) , or " Moon Walk Fashion", or whatever, fundamentally have nothing to do with creativeness, which is basically a product of genius + funding + hard work ( see my view at the start). management style could stimulate a bit , but is non-essential !


Throughout th epiece, the author have demostrated his rather low interllectual capability, by mixing around a whole bunch of none-quantifiable, none-essential, none-proven hersays into a messy logic that if one doesn't have Western MBAs, cenbtrailised, did some marquee projects, "quantity instead of quality", and "poor IP protection", then one can't innovate, while completely ignores the much more fundamental factors (see my assertion above) that are driving innovations.

He repeatedly ask "will China take ove USA? Chinese are stuggling with no SC - style innovations", followed by offering no further decent analysis and new thoughts.

Therefore, this acticle ( what the author wrote) is of bullshit quality.


At the end, I noticed that this Wadhwa is an Indian. His founding of 2 start-up could mean not too much, depending on what function he operated in those - the tech guy , or some finance banker or MBA-like marketing "guru"... Acoording to his argument logic, I would say teh latter. I personally know a Dutch who is much more impressive by founding 6 high tech ( he is a former a chief tech guy of a Dutch telecom opeartor) - with 1 listed in Nasdaq and 1 listed in Euonext. So what? And Businessweek is a publication with known reputaion as good as Daily Mail. :lazy:
 
a0333's claim that China's CPUs have 'commercial instruction sets' is just plain wrong.

A CPUs instruction set is not classified as 'commercial' or ‘military’, the CPU's ISA ( Instruction Set Architecture ) is made up of different types of instructions like for MIPS they are of Type I , R or J. Each instruction is binary coded which is the machine code. Without going into lengthy details in short there is no such thing is a 'commercial instruction set'.

The MIPS architecture is perfectly good architecture that can be used in both mil and commercial application. I find it hard to believe that SMIC would not be able to fabricate a MIPS based System-on-Chip that would not be mil spec.

Also the new trend is to use ruggedized COTS ( Commercial Off The Shelf ) components even in mil applications to keep costs low. Following is one of the suppliers for COTS embedded computer used on the MQ-1 drones
Parvus Corp - Rugged COTS Embedded Computer and Network Subsystems - Army Technology
 
gpit said:
Ok, just add a little bit different flavor to the discussion.

Why China's Chip Industry Won't Catch America's
September 3, 2009

Chinese semiconductor companies have produced some design wins, but they are still struggling when it comes to any true silicon breakthroughs
By Vivek Wadhwa
...
It takes more than money and might to achieve innovation. This is what I learned when researching the inflated estimates of engineering graduation rates in China and by analyzing its pharmaceutical industry.

Anyone that claims that China is not proficient at science innovation has not been following current events.

Let's hear it from the "global expert" CAS on China's chemical patents.

China Leads All Nations in Publication of Chemical Patents According to CAS

"China Leads All Nations in Publication of Chemical Patents According to CAS, the World's Most Authoritative Publisher of Chemical Information

Columbus, Ohio (November 23, 2009) - Chemical Abstracts Service (CAS), the global expert on chemical information, reports that China's patent office is now the world's leading producer of patent invention applications in chemistry. China trailed Japan's patent office, the World Intellectual Property Organization (WIPO), and the United States Patent and Trademark Office (USPTO) for more than a decade, but passed the USPTO in 2005, WIPO in 2006, and exceeded Japan for the first time on a monthly basis in 2008. In 2009, China will record an entire year as the number one producer of chemical patents, and CAS projects that trend to continue.

"Chemistry is widely recognized as 'the central science,'" according to Dr. Matthew Toussant, senior vice president of editorial operations at CAS. "Chemical patents are a critical component to many industrial processes and scientific realms, including medicine and natural products," said Toussant. "In fact, on average, 35 percent of new patent invention applications involve chemical substances."

"CAS has been recording the phenomenal growth of patent documents in the last decade, with the number of chemistry-related patent publications by the USPTO and WIPO growing by more than 500 percent," said Christine McCue, vice president of marketing at CAS. "Meanwhile, Chinese invention applications increased by nearly 1,400 percent, with much of that growth taking place in the pharmaceutical sector. More than half of the Chinese patent applications during this period were from inventors within China, which surely indicates that Chinese scientists now also recognize the importance of monetizing research discoveries."

Hundreds of CAS scientists, aided by state-of-the-art technology, identify and record the chemistry obscured in patents that standard search engines cannot locate. Proprietary technology systems developed by CAS enable scientists working around the world to analyze patents from 60 global patent authorities. Patent documents meeting CAS selection criteria from nine major patent offices are available in CAS databases within two days of the patents' issuance, and are fully indexed in less than 27 days. CAS scientists add value to the data they collect, entering chemical names, a unique CAS Registry Number, literature references, property data, commercial availability, preparation details, spectra, and regulatory information from international sources into CAS databases.

Media Contact

Crystal Poole Bradley
614-447-3611
cas-pr@cas.org"
 
Last edited:
Military processors are not "produced" by neither Russia, China nor India. They are high peformance processors smuggled from the United States or Japan which are then reprogramed with a different instruction set to ensure compatibility and security to carry out specific military functions. No one can fab military-grade processors other than the US and Japan.



so how does one "reprogram" a different instruction set to a processor? cause its not possible as the instructions are based on hardware, as in, its not possible to turn a x86 Pentium into a MIPS processor(unless its emulation but thats not "reprogramming")

and believe or not military grade processor generally means increased security and/or environment tolerance and/or increased up time. rarely is military goods significantly better than civilian goods in terms of performance strictly speaking, they are generally just more rugged and costs a lot more.
 
a0333's claim that China's CPUs have 'commercial instruction sets' is just plain wrong.

A CPUs instruction set is not classified as 'commercial' or ‘military’, the CPU's ISA ( Instruction Set Architecture ) is made up of different types of instructions like for MIPS they are of Type I , R or J. Each instruction is binary coded which is the machine code. Without going into lengthy details in short there is no such thing is a 'commercial instruction set'.

The MIPS architecture is perfectly good architecture that can be used in both mil and commercial application. I find it hard to believe that SMIC would not be able to fabricate a MIPS based System-on-Chip that would not be mil spec.

Also the new trend is to use ruggedized COTS ( Commercial Off The Shelf ) components even in mil applications to keep costs low. Following is one of the suppliers for COTS embedded computer used on the MQ-1 drones
Parvus Corp - Rugged COTS Embedded Computer and Network Subsystems - Army Technology

Do you even understand the difference between commercial and non-commercial instruction sets? One is open and the other is closed source, you idiot. How is China going to execute tasks on a closed-source military processor? You wipe it out and implement another instruction set you know the source codes to.

WE CAN'T UPDATE OUR FIRMWARE ANYMORE?!

so how does one "reprogram" a different instruction set to a processor? cause its not possible as the instructions are based on hardware, as in, its not possible to turn a x86 Pentium into a MIPS processor(unless its emulation but thats not "reprogramming")

and believe or not military grade processor generally means increased security and/or environment tolerance and/or increased up time. rarely is military goods significantly better than civilian goods in terms of performance strictly speaking, they are generally just more rugged and costs a lot more.

They have 2 completely different hardware designs... You Chinese ppl still need more common tech sense, seriously...

Military microprocessor technology expands, but requires less silicon space - Military & Aerospace Electronics
 
Last edited:
Do you even understand the difference between commercial and non-commercial instruction sets? One is open and the other is closed source, you idiot. How is China going to execute tasks on a closed-source military processor? You wipe it out and implement another instruction set you know the source codes to.

WE CAN'T UPDATE OUR FIRMWARE ANYMORE?!



They have 2 completely different hardware designs... You Chinese ppl still need more common tech sense, seriously...

Military microprocessor technology expands, but requires less silicon space - Military & Aerospace Electronics


Please don’t make yourself look foolish , ‘firmware’ is generally the lower level coding done using assembly language instructions , typically this looks like this

mov r1,#00fff
mov r2,#00fff
add r3,r1,r2

etc etc

The above is taken through assembler->linker->binary file creator ( .elf or intel hex or some proprietry format ) and thenprogrammed into a non-volatile memory device like an EEPROM/Flash/HDD.

Whether or not to lock down the firmware after programming the flash or EEPROM is not really related to processor architecture.

There are different classes of processors like RISC ( Reduced Instruction set computer ) and CISC (Complex Instruction set Computer ) and processors have different feature sets which include things like types of caches , hardware multipliers , barrel shifters , support for Memory Management Unit etc etc depending on application requirements these features are selected.

The x86 family is CISC class architectures for these there is a further lower level which is called the ‘micro programmed code’ that’s because in a CISC architecture a single assembly level instruction will get converted into multiple micro programmed code instructions inside the hardware.

MIPS is a RISC class architectures, these don’t require ‘micro programmed code’ and single instruction gets executed as its new trend is to use RISC anyway, the ARM processor is also RISC.

In short I don’t think you know a lot about processor architectures, you might know a thing or two about manufacturing but saying that instruction sets are ‘commercial and military’ and that ‘FIRMWARE CANNOT BE UPDATED’ just show ignorance on your part and other memebrs are right to point out that you are just trolling in this topic.
 
I don't want to beat a dead horse and this will be my last post on the subject. Many people know that Taiwan has a 80-90% share of the worldwide OEM (i.e. original equipment manufacturer) notebook computer market. However, there are many other billion-dollar high-tech fields that are affected by Taiwanese patents.

I will provide an example (see last paragraph from articles below):

Epistar, for instance, is the world`s largest red-light and the third largest blue-light LED epitaxy supplier, boasting over 50% global market share for LED TV back-lit device. Among the world`s top five LED firms, it is the only one specializing in upstream epitaxy production, with 1,100 patents including those still pending, firmly consolidating its global status.

"Taiwan is currently the world's No. 1 LED supplier by volume and No. 2 by revenue"

taiwanledtvbacklightfea.jpg

Rising sales of LED backlights for LCD TVs are driving strong growth of Taiwan’s LED industry. Firms involved in the supply chain include: LED TV panel makers AUO, CMO, CPT; LED TV brand names Amstran, BenQ, CMO; LED chip makers Epistar, Formosa Epitaxy, Tekcore, Huga, Optotech, Lextar, Chi Mei Lighting; LED packagers Unity Opto, Everlight, Harvatek, LiteOn, LHTC, Wellypower; and LED lead frame producer I-Chiun (Photo Credit: Samsung)

Semiconductor Today

"Taiwan’s LED industry to grow 18% in 2011
13 October 2010
Semiconductor Today

Taiwan’s LED industry revenue is likely to grow 18% next year from this year’s estimated NT$86.4bn (US$2.7bn at US$1:NT$32) according to the government-backed Photonics Industry & Technology Development Association (PIDA), reports the Taiwan Economic News (CENS).

According to the PIDA, Taiwan is currently the world's No. 1 LED supplier by volume and No. 2 by revenue. Of its 2010 revenue of NT$86.4bn, an estimated NT$54.4bn (US$1.7bn) will be generated by the packaging segment and NT$32bn (US$1bn) by the chip-making sector.

LED backlights for mobile phones remain the major revenue earner, accounting for 37% of the total, followed by 32% contributed by LED devices used in electronic equipment.

However, road sign, outdoor billboard and lighting applications will serve as the revenue growth engine. PIDA’s statistics show that LED road signs and billboards accounted for 20% of Taiwan`s LED industry revenue in 2009, up from 2008's 5%, while lighting application comprised 7% of the revenue in 2009, also up from 5% in 2008. The PIDA also points out that this year LED TV and other consumer electronics applications will begin driving demand for LEDs.

In conjunction with its plan to also build Taiwan into the world’s center of LED lighting modules and light sources, the government has developed a plan to boost the island's LED industry revenue more than six-fold from 2010 to NT$540bn (US$16.8bn) in 2015, creating 54,000 jobs for the industry.

Industry executives says that the strength of Taiwan’s LED industry lies in its complete processing chain manufacturing everything from epitaxial wafers and chips to packaging and modules. However, the industry’s weakness lies in its 80% dependence on imported manufacturing equipment and materials, it is noted.

The executives add that cooperation with mainland China is crucial to Taiwan’s LED industry in the light of the mainland's huge market for LED lighting projects. In addition, they urge the Taiwan government to offer lucrative incentives to users of LED lighting, in addition to coming up with industry standards for LED lighting."

ledstreetlightnewslead5.jpg

Millions of LED street lights are being deployed in Taiwan. The supply chain involves: LED chips from BridgeLux, Cree, Epistar, Formosa Epitaxy, Nichia, Osram, SemiLEDs; LED packages from Everlight, LiteOn, AOT, Bright, Harvatek, Lustrous; LED thermal modules from TTIC, CCI, AVC, Neng Tyi, Lustrous, NeoPac Opto, Advanced Thermal Devices, AuguX; LED lamp poles from Toalux, Everready Precision; and LED street lamp systems from FITI, NeoPac Opto, Bright LED, TTIC, Advanced Thermal Devices, Harvatek, LEOTEK, Delta, TGI, Unity Opto, Neo-Neon, Tatung, Genius, Topco, Anteya, Yeong Li, Alliance Optotek, AuguX, Everlight

Taiwanese LED Makers Aggressively Expanding in Light of Bright Future | CENS.com - The Taiwan Economic News

"Taiwanese LED Makers Aggressively Expanding in Light of Bright Future
2010/04/13
...
The market potential of LED has also attracted the attention of major electronics firms on the island, many of which have bought into LED firms, including Hon Hai, AU Optronics, Inventec, Lite-On, United Microelectronics Corp. (UMC), and Taiwan Semiconductor Manufacturing Corp. (TSMC). TSMC, for instance, has invested US$40 million in BridgeLux, a epitaxy-wafer maker in the U.S., thereby gaining a seat on the company`s board of directors. Some have set up their LED subsidiaries, such as AU Optronics.

Some local electronics firms have even set their sight on the market across Taiwan Strait.
Epistar, for instance, has teamed up with Lite-On and a Chinese home-appliances maker in setting up an LED firm in Changzhou of Jiangsu Province at cost of US$120 million, in addition to joining hands with UMC for investing US$16 million in an LED plant in Shandong Province. Powerchip, a memory-chip maker, has also resolved to invest US$15 million in establishing an LED firm in Xuzhou of Jiangsu Province.
...
As a result, market players foresees tremendous growth potential for the global LED market, now reaching only US$8 billion annually, in the coming years. Taiwan stands in a good stead to tap that potential, thanks to its well-established LED component industry, which ranks first place worldwide in output volume and second place in output value, trailing only Japan. Output value of Taiwan`s LED industry hit US$1.5 billion in 2008, for 20% of global market share.

Epistar, for instance, is the world`s largest red-light and the third largest blue-light LED epitaxy supplier, boasting over 50% global market share for LED TV back-lit device. Among the world`s top five LED firms, it is the only one specializing in upstream epitaxy production, with 1,100 patents including those still pending, firmly consolidating its global status.

(by Philip Liu)"
 
Last edited:
Back
Top Bottom