November 24, 2010
China led the TOP500 pack this year with its Tianhe-1A, a supercomputer laden with over 7000 NVIDIA Tesla GPUs, which has nudged Oak Ridge National Labs’ Jaguar out of the prime slot.
As Michael Feldman at HPCwire reported, this marks “the first time a non-US supercomputer has held the number one spot in six years” and also signals China’s heavy re-emergence into the supercomputing space, although it has only had a presence in the top 20 three times.
This distinction has sparked a number of conversations laced with varying forms of the verb “compete” as the rest of the world looks to China’s potential ability to lead. The mode of consuming hardware, software and services from the United States might not be the case forever—and recent discussions about this fact are more pervasive.
Supercomputing might is a symbol of technological dominance, innovation, and an outward-facing sort of development for the sake of itself. This is not to say that there is no inherent value in supercomputers, of course—this is far from the case since they contribute more than simple displays of national pride or power—there is broad practicality that is overlooked.
And speaking of what’s practical in large-scale computing, what about clouds? And what about Chinese efforts to build and make use of them?
Not only does its position at the pinnacle of the Top500 list, and announcements about a coming host of petascale datacenters, show that China is committed to developing high-end machines to rival those of the usual Western contenders, it has also been making some key plays in the cloud computing space. While the pace might not be as quick and the progress not as publicized, China is on the road to building a vast cloud infrastructure to supports its diverse national research and commercial efforts.
While all the investment from internal government and external vendors goes a long way to support a vast cloud paradigm shift, without the core infrastructure to support the software and services, this investment could possibly be doomed to fall flat—or at the very least take much longer to realize than 2015.
So if China’s got big money coming in to feed even bigger cloud computing ambitions, who’s to say that this infrastructure problem (arguably the biggest barrier preventing a U.S.-like cloud success story for China) is where the heated Top500 conversations about competitiveness should really be rooted?
The Issue of Infrastructure
While there is significant investment rolling in to fund China’s cloud boom, there have been a number of questions posed about the feasibility since there remain some profound infrastructure issues. Even with massive investment in cloud computing development, without a solid foundation, China’s cloud vision will not thrive.
Earlier this month, Christopher Mims at MIT’s Technology Review asked whether or not it makes sense for Beijing to become a world-class cloud leader when it lacks the basic broadband infrastructure to do so.
He cites a recent article in China.org.cn that states, “Internet Data Centers (IDC) are the basic platforms for cloud computing, but China’s IDCs are mostly small-scale, energy-thirsty facilities that provide poor service. China’s IDC market amounted to 6.39 million yuan in 2009, just 5 percent of the Asian market. More than 90 percent of China’s IDCs occupy less than 400 square meters.”
Mims also notes that the United States “paved the way for the cheap cloud services that we now take for granted through large-scale federal action, namely the Telecommunications Act of 1996, which allowed for more competition between the carriers of various services.” To complement this, the internet bubble that followed left plenty of “dark infrastructure “behind for companies like Google and others to acquire at greatly reduced investment—and do something useful.
China will continue to build the cloud vision with new capital outlay in the form of billion dollar government grants and investment from U.S.-based companies…even if the ground upon which its all constructed hasn’t yet been leveled.
Earlier this summer, Shanghai unveiled its “Sea of Clouds” vision to the tune of 3.12 billion yuan with the goal of cementing the city as the foremost leader in the cloud revolution.
Almost immediately, Beijing announced its Propitious Cloud Project (wonder if that’s just awkward titling or bad translation) which injected 50 billion yuan into the industry. This plan is focused on refinement of all things cloud-related, including the infrastructure, software and delivery. It has been touted as the beginning of the 2015 goal of driving 200 billion yuan into the pipeline. This is what Beijing calls “the largest cloud computing center in China and even the world” but news on the progress of that is still mounting—and questionable--since defining “largest” is no easy task given multiple variables.
This week the Chinese newspaper, People’s Daily, reported that “China will build a complete industrial chain of cloud computing technology and create the ‘China Cloud Valley’ in Harbin, capital of northeastern China’s Heilongjing Province over the next three years.”
Harbin is already a data center haven due to its plentiful supplies of inexpensive power and water resources. This new Cloud Valley “will include a cloud computing center base; an application, innovation and research development base; and a business incubator base” and will focus on the so-called Internet of Things—specifically software and service outsourcing as well as the media and animation market.
Information on the availability and quality of network services is still scant on the “cloud valley” announcement. When you think about it, the play on the term “Silicon Valley” was apropos at the time because it referred to an abundance of something tangible; something that formed the very ground everything stood upon.
With that in mind, what can the concept of a “Cloud Valley” –something that’s being constructed on that which is not concrete, isn’t clearly defined, and is not part of something that already existed? Where is the infrastructure-related solidity for China that it hopes to build such a vast network of pathways on?
Overseas Delivery for China’s Cloud Goals
Investment in China’s cloud computing future has been underway in earnest since 2008, beginning with IBM’s announcement of its cloud center in Jiangsu Province.
Microsoft has also been making some in strengthening relationships in China. For example, way back in 2008, they helped to establish a cloud development and training platform in Hangzhou to help Chinese enterprises improve their software R&D capabilities. Last year they signed a deal with Taiwan telecom giant Chunghwa to allow cooperation in delivering on-site cloud services.
Back in September, Microsoft China announced the opening of their Cloud Innovation Center in Shanghai which, as Bob Muglia, President of the Server and Tools Business division said, would strengthen worldwide partnerships and deliver a “healthy IT ecosystem” for China and the West.
Muglia noted that “Today, about 70 percent of Microsoft engineers are doing work related to the cloud, including people in Microsoft’s Asia-Pacific Research and Development Group” and that this number would do nothing but grow.”
Grow as it might, it seems that if there is any investment that is critical to prop up this kind of widespread initiative, it should be in the foundations at this point rather than the pieces that come after the concrete has been laid.
Despite a great deal of funding from the Chinese government and from Western companies, including, among others, IBM, Microsoft and Mellanox, (all of whom had very recent announcements geared toward growing a presence in China’s emerging clouds) there is a critical lack of the appropriate infrastructure to support lofty goals.
If China is to become a serious player in the emerging industries that will revolve around cloud computing—which now includes everything from hardware to services to mobile technology developments—this infrastructure shortage could be the great Achilles’ heel.
Companies like Microsoft, Mellanox (with this announcement of the public cloud computing center) and others, if truly invested in China’s future and their stake in it must somehow see that there is a piece missing from the puzzle. If this is the case, why pony up significant resources to help propel the clouds forward when there are only small patches of earth underneath them?
May 23, 2013 |
The study of climate change is one of those scientific problems where it is almost essential to model the entire Earth to attain accurate results and make worthwhile predictions. In an attempt to make climate science more accessible to smaller research facilities, NASA introduced what they call ‘Climate in a Box,’ a system they note acts as a desktop supercomputer.
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/02/2012 | AMD | Developers today are just beginning to explore the potential of heterogeneous computing, but the potential for this new paradigm is huge. This brief article reviews how the technology might impact a range of application development areas, including client experiences and cloud-based data management. As platforms like OpenCL continue to evolve, the benefits of heterogeneous computing will become even more accessible. Use this quick article to jump-start your own thinking on heterogeneous computing.