You are here

Hybrid Technologies Amplifier Chain for>30 Gbps Per Data Link Energy Efficient Digital Output from 4K


OBJECTIVE: The objective is to demonstrate means of transporting high speed, digital data from 4K to 300K via a well integrated set of technologies that will minimize the heat loading on the low temperature stages. DESCRIPTION: After the inherent inefficiency of 4K coolers is considered, the consumption of wall power by Nb superconducting digital logic in performing its calculations is 100x smaller than SOA 22 nm Si. However, for systems with high levels of processing to realize a net energy benefit, parasitic heat loads associated with signal and data communications between room and low temperature need to be severely limited since all that energy must be removed from the low temperature end. Optical data transport from 300 to 4K is well advanced and should be assumed to occur. However, current optical modulators do not exist with V(pi) of less than 25mV for speeds above 50 Gbps per line. Thus the heat load associated with transitioning to optical data transmission at 4K is unacceptable. Therefore some of the temperature gradient must be spanned by electrical signals and an amount of amplification sufficient for the signals to remain visible above the upper end noise floor is essential. Superconducting digital logic operates on the basis of magnetic flux quanta, each of which has a voltage /time area of 2 mVps and current devices produce pulses of ~1 ps duration. The conversion temperature must be selected to minimize the total energy dissipation, including issues such as inherent noise of amplifiers, stage temperature of coolers, and energy efficiency of different stage temperatures of coolers. Moreover, the output data format must be converted from the RZ logic of superconductivity to the NRZ norm of room temperature logic somewhere on its way up the cryopackaging. This situation requires proper co-design of the first stage amplifiers which combine the superconducting output drivers and another digital technology capable of gain, ideally on a single 4K substrate, plus other, higher temperature amplification/signal conditioning stages sufficient to deliver the data without induced errors to 300K. The goal is for the entire chain to minimize total energy consumption for realistic values of cooler efficiency at each heat sink temperature and lead thermal transport. PHASE I: Develop and demonstrate by simulation of energy consumption and bit error rate a complete amplifier chain that deposits via dissipation and thermal conduction less than 1 mW at 4K for each 30 Gbps data link exporting data from 4K to 300K with BER<1 in a trillion. The likely manufacturing cost and simplicity of production and assembly into packaging serving>100 simultaneous data streams will be secondary criteria in evaluating the goodness of designs. The phase 1 proposal should include at least a notional definition of the sort of amplifier chain to be worked. PHASE II: Convert the circuit simulations into real demonstration articles and perform multiple fab/test/redesign cycles to produce the desired demonstration. Demonstrate these links operating in a real superconducting dsp processor during the phase 2 second option and confirm the magnitude of the system energy efficiency benefit. PHASE III: Transition the new technology into advanced digital/mixed signal superconducting systems, such as full spectrum RF awareness receivers, beam formers, or high performance computing. PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATIONS: Successful development of this amplifier chain will enable the full speed advantage (>5x fastest semiconductor digital logic demonstration) of superconducting logic to be used in high speed circuits. These are needed in cell phone ground stations (spectrum reuse by multiple subscribers), server farms serving applications such as 4G wireless, Google, and cloud computing, and for computer animation and simulation such as for the entertainment and weather forecasting communities. REFERENCES: 1. A. Kadin et al., IEEE Transactions on Applied Superconductivity, Vol 17, pp. 975-978 (2007). 2. R.J. Webber et al, IEEE Transactions on Applied Superconductivity, Vol 19, pp. 999-1002 (2009). 3. J D Cressler, JOURNAL DE PHYSIQUE IV, Colloque C6 SupplCment au Journal de Physique III, Volume 4, juin 1994 4. Niklas Wadefalk, IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 51, NO. 6, JUNE 2003
US Flag An Official Website of the United States Government