March 15, 2017

US needs a lot of supercomputing spending or risk falling broadly behind China by 2020

In June of 2016, China newest supercomputer, the Taihu Light system, nearly triple the previous record-setting numbers of another Chinese supercomputer, Tianhe -2. These results were achieved using a Chinese designed and fabricated 28nm System -on -Chip (SOC) architecture. This announcement was coupled with impressive advancements in indigenously developed system software and high performance computing ( HPC) applications , as shown by having three Chinese research teams as finalists for the 2016 Gordon Bell prize. These results indicate that China has attained a near -peer status with the U.S. in HPC. The U.S. asserted its intention to maintain a leadership position in HPC in the July 2015 Executive Order establishing the National Strategic Computing Initiative (NSCI).

It is now clear that future U.S. leadership will be challenged by the Chinese. The 2012 Net Assessment of Foreign HPC noted the aggressive development of Chinese HPC capabilities, and , in particular , the accelerated rate of investment that China was making in these areas . In September 2016, experts and leaders from the U.S. HPC community convened to update that assessment in light of the recent Chinese results. Meeting participants expressed significant concern that – absent aggressive action by the U.S. – the U.S. will lose leadership and not control its own future in HPC. There was broad support for maintaining HPC leadership, and a common belief in the importance of that leadership, as well as wide agreement on key steps to maintain this leadership position.

National security requires the best computing available, and loss of leadership in HPC will severely compromise our national security. HPC plays a vital role in the design, development or analysis of many – perhaps almost all – modern weapon systems and national security systems: e.g., nuclear weapons, cyber, ships, aircraft, encryption, missile defense, precision - strike capability, and hypersonics. Loss of leadership in HPC could significantly reduce the U.S. nuclear deterrence and the sophistication of our future weapons systems. Conversely, if China fields a weapons system with new capabilities based on superior HPC, and the U.S. cannot accurately estimate its true capabilities, there is a serious possibility of over - or under - estimating the threat. Either possibility leads to unwelcome situations such as distortions in the allocation of R and D resources and strategic planning for defense, uncertainty in national policy -making, and incorrect responses to world events. The U.S. enjoys relatively cost effective HPC due to our indigenous ecosystem and trusted access. Loss of leadership could easily require the USG to acquire HPC capabilities in much the same way it acquires aircraft carriers, at vastly increased cost.

HPC leadership has important economic benefits because of HPC’s role as an enabling technology. A loss of leadership will be felt beyond the HPC vendor community, which would be significantly impaired: HPC resources are required for the development of a variety of military, scientific , and industrial capabilities. Loss of a U.S. leading position would threaten our ability to compete internationally in all of these fields. The effects of losing leadership can be expected to be long -lived; HPC help s drive the development of talent in math, science, and engineering. The economic and social benefits of the “Post Moore’s Law ” era may very well be concentrated where HPC leadership exists. For industrial applications, reliance on foreign HPC resources could threaten the loss of intellectual property and competitive edge. Personal email and private information, social networks, and the emerging Internet of Things are all subject to even greater privacy risks if offshore entities have superior HPC analytics or control the data / information markets.

Leadership positions, once lost, are expensive to regain. To maintain U.S. leadership in HPC, a surge of USG investment and action is needed to address HPC priorities. Many of these priorities have been out lined by the NSCI and some are clearly under way, in particular , DOE efforts in accelerating delivery of a capable exascale computing system (Objective #1) , and some aspects of establishing “post -Moore's Law” computing (Objective #3), led by IARPA (Intelligence Advanced Research Projects Activity ) and other agencies.

Форма для связи


Email *

Message *