ترغب بنشر مسار تعليمي؟ اضغط هنا

SFQ bias for SFQ digital circuits

60   0   0.0 ( 0 )
 نشر من قبل Sergey Tolpygo
 تاريخ النشر 2020
  مجال البحث فيزياء
والبحث باللغة English




اسأل ChatGPT حول البحث

Superconductor electronics fabrication technology developed at MIT Lincoln Laboratory enables the development of VLSI digital circuits with millions of Josephson junctions per square centimeter. However, conventional DC and multi-phase AC biasing techniques already encounter serious challenges for scaling circuits above several hundred thousand junctions. In this work, we propose a novel AC-based biasing scheme for RSFQ-type logic families requiring DC bias. The major step toward this scheme is a superconducting AC/DC rectifier which we introduced at ASC 2014. Initially, we proposed to connect the rectifiers to payload cells via superconducting inductors with large inductance in order to reduce parasitic effects of flux quantization. Recently, we discovered that this powering scheme works even better at a much lower value of the inductance, when it is just sufficient to hold only one or two flux quanta in the inductive loop between the converter and the payload. In this case, flux quantization in the loop becomes beneficial because the value of current fed into the payload is defined by the value of the coupling inductance. Therefore, our AC/SFQ converter powers the payload cell by a single flux quantum rather than by DC current. Such mode of operation is extremely energy efficient because the energy is used only to recover flux quantum consumed by the cell during the logic operation. We present designs of AC/SFQ converters comprising an AC/DC rectifier and a current conditioning circuit which we termed an SFQ filter. We also present test results and demonstrate AC/SFQ powering a payload circuit using circuits fabricated in a new, 150-nm node of Lincoln Laboratory fabrication technology using self-shunted Nb/AlOx-Al/Nb Josephson junctions with 600 $mu$A/$mu$$m^2$ critical current density and 200 nm minimum linewidth of inductors.

قيم البحث

اقرأ أيضاً

In this work, we briefly overview various options for Josephson junctions which should be scalable down to nanometer range for utilization in nanoscale digital superconducting technology. Such junctions should possess high values of critical current, $I_c$, and normal state resistance, $R_n$. Another requirement is the high reproducibility of the junction parameters across a wafer in a fabrication process. We argue that Superconductor - Normal metal - Superconductor (SN-N-NS) Josephson junction of variable thickness bridge geometry is a promising choice to meet these requirements. Theoretical analysis of SN-N-NS junction is performed in the case where the distance between the S-electrodes is comparable to the coherence length of the N-material. The restriction on the junction geometrical parameters providing the existence of superconductivity in the S-electrodes is derived for the current flowing through the junction of an order of $I_c$. The junction heating, as well as available mechanisms for the heat removal, is analyzed. The obtained results show that an SN-N-NS junction with a high (sub-millivolt) value of $I_cR_n$ product can be fabricated from a broadly utilized combination of materials like Nb/Cu using well-established technological processes. The junction area can be scaled down to that of semiconductor transistors fabricated in the frame of a 40-nm process.
Ultra-fast & low-power superconductor single-flux-quantum (SFQ)-based CNN systolic accelerators are built to enhance the CNN inference throughput. However, shift-register (SHIFT)-based scratchpad memory (SPM) arrays prevent a SFQ CNN accelerator from exceeding 40% of its peak throughput, due to the lack of random access capability. This paper first documents our study of a variety of cryogenic memory technologies, including Vortex Transition Memory (VTM), Josephson-CMOS SRAM, MRAM, and Superconducting Nanowire Memory, during which we found that none of the aforementioned technologies made a SFQ CNN accelerator achieve high throughput, small area, and low power simultaneously. Second, we present a heterogeneous SPM architecture, SMART, composed of SHIFT arrays and a random access array to improve the inference throughput of a SFQ CNN systolic accelerator. Third, we propose a fast, low-power and dense pipelined random access CMOS-SFQ array by building SFQ passive-transmission-line-based H-Trees that connect CMOS sub-banks. Finally, we create an ILP-based compiler to deploy CNN models on SMART. Experimental results show that, with the same chip area overhead, compared to the latest SHIFT-based SFQ CNN accelerator, SMART improves the inference throughput by $3.9times$ ($2.2times$), and reduces the inference energy by $86%$ ($71%$) when inferring a single image (a batch of images).
We demonstrate the successful operation of a multi-element superconducting nanowire single-photon detector (SSPD) array integrated with a single-flux-quantum (SFQ) readout circuit in a compact 0.1 W Gifford-McMahon cryocooler. A time-resolved readout technique, where output signals from each element enter the SFQ readout circuit with finite time intervals, revealed crosstalk-free operation of the four-element SSPD array connected with the SFQ readout circuit. The timing jitter and the system detection efficiency were measured to be 50 ps and 11.4%, respectively, which were comparable to the performance of practical single-pixel SSPD systems.
In radio interferometry, the quantization process introduces a bias in the magnitude and phase of the measured correlations which translates into errors in the measurement of source brightness and position in the sky, affecting both the system calibr ation and image reconstruction. In this paper we investigate the biasing effect of quantization in the measured correlation between complex-valued inputs with a circularly symmetric Gaussian probability density function (PDF), which is the typical case for radio astronomy applications. We start by calculating the correlation between the input and quantization error and its effect on the quantized variance, first in the case of a real-valued quantizer with a zero mean Gaussian input and then in the case of a complex-valued quantizer with a circularly symmetric Gaussian input. We demonstrate that this input-error correlation is always negative for a quantizer with an odd number of levels, while for an even number of levels this correlation is positive in the low signal level regime. In both cases there is an optimal interval for the input signal level for which this input-error correlation is very weak and the model of additive uncorrelated quantization noise provides a very accurate approximation. We determine the conditions under which the magnitude and phase of the measured correlation have negligible bias with respect to the unquantized values: we demonstrate that the magnitude bias is negligible only if both unquantized inputs are optimally quantized (i.e., when the uncorrelated quantization error model is valid), while the phase bias is negligible when 1) at least one of the inputs is optimally quantized, or when 2) the correlation coefficient between the unquantized inputs is small. Finally, we determine the implications of these results for radio interferometry.
We present a fabrication scheme and testing results for epitaxial sub-micrometer Josephson junctions. The junctions are made using a high-temperature (1170 K) via process yielding junctions as small as 0.8 mu m in diameter by use of optical lithograp hy. Sapphire (Al2O3) tunnel-barriers are grown on an epitaxial Re/Ti multilayer base-electrode. We have fabricated devices with both Re and Al top electrodes. While room-temperature (295 K) resistance versus area data are favorable for both types of top electrodes, the low-temperature (50 mK) data show that junctions with the Al top electrode have a much higher subgap resistance. The microwave loss properties of the junctions have been measured by use of superconducting Josephson junction qubits. The results show that high subgap resistance correlates to improved qubit performance.
التعليقات
جاري جلب التعليقات جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
mircosoft-partner

هل ترغب بارسال اشعارات عن اخر التحديثات في شمرا-اكاديميا