Breaking the Supercomputing Barrier
(July 5, 2011)
Dear Subscribers and Readers,
I hope our American subscribers had a great July 4th weekend and holiday! I understand some of our British subscribers have mixed feelings about America's Independence Day; but America's alliance with Great Britain has proved very fruitful for both countries, especially since the outbreak of World War I. For all of the lost ideals in the post-financial-crisis and post-9/11 world, I am optimistic on what this country represents. After all, we have already made great strides in areas of civil rights and on an international basis, human rights—starting with the fall of the Berlin Wall in 1989. As the information age and the biotech age progress, not only would global material wealth reach unprecedented heights; but physical and spiritual health as well. It's not always “morning in America,” but it is always darkest before the dawn. For all the “problems” that we covered—including the overstretched budget deficits and the >US$60 trillion in unfunded Social Security and Medicare liabilities—I sincerely believe that subsequent generations would lead more fulfilling lives in both the material and spiritual realms.
Let us now begin our commentary with a review of our 13 most recent signals in our DJIA Timing System:
1st signal entered: 50% short position on October 4, 2007 at 13,956;
2nd signal entered: 50% short position COVERED on January 9, 2008 at 12,630, giving us a gain of 1,326 points.
3rd signal entered: 50% long position on January 9, 2008 at 12,630;
4th signal entered: Additional 50% long position on January 22, 2008 at 11,715;
5th signal entered: 100% long position SOLD on May 22, 2008 at 12,640, giving us gains of 925 and 10 points, respectively;
6th signal entered: 50% long position on June 12, 2008 at 12,172;
7th signal entered: Additional 50% long position on June 25, 2008 at 11,863;
8th signal entered: Additional 25% long position on February 24, 2009 at 7,250;
9th signal entered: 25% long position SOLD on June 8, 2009 at 8,667, giving us a gain of 1,417 points;
10th signal entered: 50% long position SOLD on March 29, 2010 at 10,888, giving us a loss of 1,284 points.
11th signal entered: 50% long position SOLD on April 27, 2010 at 11,044, giving us a loss of 819 points;
12th signal entered: 50% long position initiated on May 21, 2010 at 10,145;
13th signal entered: 50% long position SOLD on December 15, 2010 at 11,487, giving us a gain of 1,342 points; the DJIA Timing system is currently in a neutral position.
Wall Street analysts have gotten increasingly concerned about the possibility of a Chinese “hard landing” sometime this year or next year. Specifically, Moody's claims that debt incurred by Chinese local governments had been understated by about US$540 billion, or about 10% of Chinese GDP. The principal problems of gauging local government indebtedness in China are two-fold: 1) The opaqueness of the system; as exemplified by three different estimates provided by the CBRC (China's banking regulator), the People's Bank of China, and the Chinese National Audit Office; and 2) the fact that local governments are technically prohibited from bank borrowing, but have done so anyway through other means such as setting up different investment arms. Because of this, the ultimate size of local government indebtedness may never be known.
More important—according to the Chinese National Audit Office—nearly 25% of this debt is due this year, and another 17% next year. Moreover, local government indebtedness varies. An estimated 25% of local governments have (known) debts equal to more than 100% of revenues. As the People's Bank of China continue to tighten liquidity, some of these “marginal” local governments could experience solvency problems—beginning a “domino effect” unless the central government bails them out. Of course, the latter depends on the political will of the Community Party—but given the vast resources of the central government, as well as the blistering growth of the Chinese economy, bailing out local governments would have little to no immediate costs. Indeed, the Chinese government was indirectly responsible for bailing out the U.S. banking system/economy through its Treasuries purchases. To sit back and let its local governments fail would make no sense from a political and economic standpoint. While a Chinese slowdown could occur due to tightening liquidity, I do not expect a Chinese “hard landing” in the immediate future.
Last month, the 37th semi-annual edition of the top 500 list of the world's most powerful supercomputers was published at the 2011 International Supercomputing Conference in Hamburg, Germany. Most notably, while the top 500 list is still dominated by Intel-based systems (although many of the top-performing systems utilize IBM's “Blue Gene” processors; with AMD a distant second, and IBM third), the new number one supercomputer, the “K Computer” is built by Fujitsu, using its own SPARC64 Vlllfx CPUs. Powered by 68,544 CPUs, each with eight cores (for a total 548,352 cores), the “K Computer” has a peak performance of 8.16 petaflops and replaces the Tianhe-1A supercomputer at the National Supercomputing Center in Tianjin, with a performance of “just” 2.6 petaflops.
From a geopolitical standpoint, Japan now occupies the top spot in terms of the fastest-performing supercomputer for the first time since 2004 (when its “Earth Simulator,” with a performance of 36 teraflops, occupied the top spot). The K Computer is about 225 times faster than the Earth Simulator in terms of peak performance. On a country basis, 42.9% of the top 500 supercomputers (by supercomputing power; note that the NSA – which houses some of the most powerful systems in the world – stopped reporting in 1998) are located in the US. Japan is second, with 19.0% of the world's supercomputing power (a jump from 6.7% just six months ago). Rounding out the top five are China (12.1%), Germany (5.5%), and France (5.4%). The UK, which ranked third two years ago (with 5.5% of the world's supercomputing power), is now in 6th place, housing just 3.2% of the world's supercomputing power.
Aside from providing the most up-to-date supercomputing statistics, the semi-annual list also publishes the historical progress of global supercomputing power – as well as a reasonably accurate projection of what lies ahead. Following is a log chart summarizing the progression of the top 500 list since its inception in 1993, along with a ten-year projection:
Today, a typical desktop with an Intel or AMD processor operates at about 75 gigaflops (note that we are ignoring the GPU in our graphics processor from our calculations) – or the equivalent of an “entry-level” supercomputer on the top 500 list in 2000. By early next year, the power of a typical desktop should reach 100 gigaflops, or the equivalent of the most powerful supercomputer in the world in 1993. On the highest end, the power of Japan's K Supercomputer is equivalent to the combined performance of the world's top 500 supercomputers just four years ago. Moreover, IBM is on track to construct a Blue Gene/Q supercomputer with a targeted performance of 20 petaflops (or 20 million gigaflops) for the National Nuclear Security Administration early next year. Code-named “Sequoia,” this supercomputer will possess more than one-third of the combined performance of all the supercomputers in the top 500 list as of today (see above chart) once it is online.
Simulations that would have taken 10 years of computing hours for the most powerful supercomputer six months ago will only take just 2 ½ years on the K Supercomputer, and just one year once Sequoia comes online (roughly, since Linpack—the benchmark used to measure supercomputing performance—is not exactly representative of real-world supercomputing performance). Tasks that take an immense amount of computing time today – such as weather forecasts, gene sequencing, airplane and automobile design, protein folding, etc. – will continue to be streamlined as newer and more efficient processors/software are designed. By 2018, the top supercomputer should be able to reach a sustained performance of an exaflop (i.e. 1,000 petaflops)—this is both SGI's and Intel's goal. IBM believes that such a system is needed to support the “Square Kilometre Array”—a radio telescope in development that will be able to survey the sky 10,000 times faster than ever before, and 50 times more sensitive than any current radio instrument—and will provide better answers to the origin and evolution of the universe. The ongoing :democratization” of the supercomputing industry would also result in improvements in solar panel designs, better conductors, more effective drugs, etc. As long as global technology innovation isn't stifled, the outlook for global productivity growth – and by extension, global economic growth and standard of living improvements – will remain bright for years to come. Should the quantum computer be commercialized soon (note that Lockheed Martin has already bought one such computer from D-Wave), subscribers should get ready for the next major technological revolution (and secular bull market) by 2014 to 2020. Make no mistake: The impact of the next technological revolution will dwarf that of the first and second industrial revolutions.
Let us now discuss the most recent action in the U.S. stock market using the Dow Theory. Following is the most recent action of the Dow Industrials vs. the Dow Transports, as shown in the following chart from January 2008 to the present:
For the week ending July 1, 2011, the Dow Industrials rose 539.21 points, while the Dow Transports rose 289.33 points, or a whopping 5.5%. While the Dow Transports made an all-time high, this was not confirmed by the Dow Industrials. Moreover, the market is now highly overbought—thus, I urge subscribers to remain cautious. Given the combination of weak technical and liquidity conditions, the market action should remain tough this summer. We remain neutral in our DJIA Timing System as we take a wait-and-see approach.
I will now continue our commentary with a quick discussion of our popular sentiment indicators – those being the bulls-bears percentages of the American Association of Individual Investors (AAII), the Investors Intelligence, and the Market Vane's Bullish Consensus Surveys. The four-week moving average of these sentiment indicators decreased from a reading of 4.4% to 3.5% for the week ending July 1, 2011. Following is a weekly chart showing the four-week moving average of the Market Vane, AAII, and the Investors Intelligence Survey Bulls-Bears% Differentials from January 1999 to the present:
After peaking at 30.8% (its highest reading since late February 2007) in late January, the four-week MA declined to 3.5% last week—near its most oversold level since mid-September last year. Despite this oversold condition, it still isn't as oversold as it was during last summer's correction. We will retain our neutral position in our DJIA Timing System, as we believe the market should remain tough this summer.
I will now close out our commentary by discussing the latest readings of the ISE Sentiment Index. For newer subscribers, I want to provide an explanation of ISE Sentiment Index and why it has turned out to be (and should continue to be) a useful sentiment indicator. Quoting the International Securities Exchange website: The ISE Sentiment Index (ISEE) is designed to show how investors view stock prices. The ISEE only measures opening long customer transactions on ISE. Transactions made by market makers and firms are not included in ISEE because they are not considered representative of market sentiment due to the often specialized nature of those transactions. Customer transactions, meanwhile, are often thought to best represent market sentiment because customers, which include individual investors, often buy call and put options to express their sentiment toward a particular stock.
When the daily reading is above 100, it means that more customers have been buying call options than put options, while a reading below 100 means more customers have been buying puts than calls. As noted in the above paragraph, the ISEE only measures transactions initiated by retail investors – and not transactions initiated by market makers or firms. This makes the indicator a perfect contrarian indicator. Since the inception of this index during early 2002, its track record has been one of the best relative to that of other sentiment indicators. Following is the 20-day and 50-day moving average of the ISE Sentiment Index vs. the daily S&P 500 from May 1, 2002 to the present:
The 20 DMA decreased from 105.1 to 101.8 last week—remaining close to its most oversold level since late July 2010. Similarly, the 50 DMA is also highly oversold—and is in fact at its most oversold level since September last year. While the oversold conditions in this indicator suggests the market could rally further, the fact that our other sentiment indicators is not as oversold is troubling—and suggests caution is warranted when purchasing stocks, especially given the weak technical and liquidity conditions. We will remain neutral in our DJIA Timing System and will take a wait-and-see approach.
Conclusion: Despite the structural challenges in the developed world (unprecedented government debt levels, aging of the baby boomers, etc.), the technological revolution steams ahead. At some point in the next five to ten years, the world will reach a “tipping point” as the next major breakthrough materializes, whether it is the commercialization of the quantum computer, carbon nanotubes, or second-generation algae-based biofuels. Whichever it is, that breakthrough will accelerate the commercialization of other technologies—ushering in the next technological revolution. This will also be accompanied by the beginning of the next secular bull market. In the meantime, I believe the U.S. stock market remains in a consolidation period this summer. I also believe that both the Euro and the Yen are unattractive relative to the U.S. Dollar. In the meantime, I continue to expect a tough summer for U.S. stocks. We remain neutral in our DJIA Timing System, for now. Subscribers please stay tuned.
Henry To, CFA, CAIA