STATEMENT OF

KENNETH E. BAKER
ACTING DIRECTOR
OFFICE OF NONPROLIFERATION AND NATIONAL SECURITY
U.S. DEPARTMENT OF ENERGY

Mr. Chairman and members of the Subcommittee, I would like to thank you for the opportunity to address today=s hearing on the impact on U.S. national security interests of the sale or transfer of high performance computers to foreign entities or governments engaged in nuclear weapons research. I will focus my remarks on the potential implications of Russia=s acquisition of such high performance computers.

Russia has been designing and developing nuclear weapons for over 50 years. Over that period, Russia has developed and deployed a modern nuclear warhead stockpileCalthough the design philosophy underlying this stockpile differs from the U.S. approach.

We believe that the Russians have great confidence in their nuclear warhead stockpile at this time. The Russians ceased underground nuclear testing in 1990, and we believe they may have frozen their nuclear warhead designs somewhat earlier. The Russians have sought to cope with typical stockpile maintenance issues, such as material aging, through a program of regular remanufacturing of nuclear warheads. Developments within Russia of a socio-economic originCpay disruptions, falling worker morale, and possible supplier disruptionsCcould jeopardize Russia=s traditional approach to stockpile maintenance. If true, the ability of Russian nuclear weapons designers to continue to certify stockpile reliability and safety in a zero-yield environment could erode over time. Moreover, the Russians may not be willing to bear the continued expense of maintaining the enormous production complex required to sustain this approach.

Traditionally, if the Russians were concerned about a flaw or some other problem in the performance of a particular nuclear design, modifications or changes to the design could be tested prior to recertification of the nuclear warhead. Under a CTBT, however, this recourse is denied to the Russians. Many of the details of Russia=s approach to this dilemma remain classified; however, Russian nuclear weapons designers have identified high performance computers as one of the key requirements of their future approach to stockpile maintenance.

For example, the Russians have an extensive nuclear testing history and, consequently, a rich data base of nuclear test information. The Russians could use high performance computers in conjunction with these data to enhance their ability to perform more precise nuclear weapons simulations. Russian nuclear weapons designers should be able to adjust their simulation models through reference to such data and then more accurately predict nuclear weapons performance. Today=s high performance computer systems can also be used for some multidimensional safety simulations, but would not provide a three-dimensional calculational capability sufficient to address stockpile aging issues in a Comprehensive Test Ban Treaty environment.

The Russians may also have increasing concerns about the safety of their nuclear stockpile. Although our reasons for believing this are classified, our own experience is that safety calculations are much more dependent on high performance computers than the calculations associated with estimates of warhead design yields. Consequently, this capability, if integrated with other activities we believe the Russians have underway, could be intended to address their nuclear warhead safety concerns.

On January 13, 1997, Minister for Atomic Energy Victor Mikhaylov told Itar Tass that the Russian nuclear weapon facilities, Arzamas 16 and Chelyabinsk 70, purchased several high performance computers. He stated, Athe new equipment will simulate nuclear explosions needed to increase the safety of Russian nuclear arsenals following Russia=s accession to the Comprehensive Test Ban Treaty.@ The Department of Commerce and the Department of Justice are conducting an investigation into this case, and I would defer to my colleagues for details. A final determination of the impact on U.S. national security must also await a more precise definition of actual Russian acquisitions.

At this time, the Department of Energy judges that Russia=s existing computing capability is such that the primary benefit of this acquisition would be an increase in the speed of operations and a reduction in the time required to perform such calculations. Over time, if the Russians are successful in implementing their new approach to stockpile maintenance there may also be some considerable cost savingsCalthough the savings in human or financial resources are impossible to calculate. Most importantly, the Russians could have the ability to recertify their stockpileCwithout resort to underground nuclear testing.

There are a number of nuclear design problems for which high performance computers make no contribution at all. First, we believe that the Russians are generally satisfied with the current nuclear warhead designs in their existing stockpile. Even the most modern of these designs was performed on computers far less capable than these. Nor do these high performance computers represent an advance in the Russian=s ability to achieve advanced nuclear weapons concepts. These represent the so-called Afourth generation of nuclear weapons@ and generally include such concepts as X-ray lasers and nuclear-pumped optical lasers. While the Russians are known to be interested in such concepts, their existing computing capabilityCprior to the acquisition of these high performance computersCis sufficient for these designs; the gaps in the Russian program require a more extensive underground testing program for resolution. Finally, in the absence of CTBT obligations, the systems reportedly acquired could provide a capability for advanced weapon conceptsCbut only in parallel with a robust testing program. The primary contribution in this regard would be in the rapid interpretation and comprehension of nuclear test data.

Let me be clear. In no way should the assessment here suggest a lack of concern. The alleged Russian purchases of supercomputers are contrary to U.S. policy, which precludes direct assistance to the maintenance of reliability of Russia=s nuclear weapons.

I would like to thank the committee for its time, and I would be willing to submit a more detailed, classified assessment upon request from the committee.