John A. Chandy

University of Connecticut, Storrs, Connecticut, United States

Are you John A. Chandy?

Claim your profile

Publications (65)19.92 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: Computation within a network switch accelerates data processing.•FPGA implementations of modules to process data on the fly through the network.•Data processing modules include data search, sort, k-min/k-max, and k-means clustering.
    Future Generation Computer Systems 11/2014; · 2.64 Impact Factor
  • 06/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we propose a new architecture for analog-to-digital converters (ADCs) using multistate spatial wavefunction-switched field-effect transistors (SWSFETs). SWSFETs are multiple quantum coupled well devices, where the wells are stacked vertically and the electron wavefunction switches from one well to another with the change in gate voltage. Quantum mechanical simulations of 3-well InGaAs-AlInAs SWSFET structures are presented. The designs and simulations of 2-bit and 3-bit ADCs using SWSFETs result in low power consumption and reduced device count which improves the speed of the data conversion.
    International Journal of High Speed Electronics and Systems 05/2014; 23(01n02).
  • [Show abstract] [Hide abstract]
    ABSTRACT: Spatial wave-function switched field effect transistor (SWSFET) switches the current flow between different channels inside the FET based on the applied voltage in its gate terminal. SWSFET can be used to implement multi-valued logic circuit with less number of circuit elements. Recently we presented unipolar inverter circuit using SWSFET. In this paper we develop a circuit model of SWSFET based on BSIM 3.2.0 and BSIM 3.2.4 and implement membership function using that circuit model of SWSFET. The spatial wave-function switched field effect transistor (SWSFET) has two or three low band-gap quantum well channels inside the substrate of the semiconductor. Applied voltage at the gate region of the SWSFET, switches the charge carrier concentration in different channels from source to drain region. A circuit model of SWSFET is developed in BSIM 3.2.0. Membership function is implemented using the circuit model of the SWSFET. Membership function implementation using less number of SWSFET will reduce the device count in future analog-to-digital converter (ADC) and digital-to-analog converter (DAC) circuits.
    International Journal of High Speed Electronics and Systems 05/2014; 23(01n02).
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we are presenting fast digital to analog convertor designs using Spatial Waveform Switched FETs (SWSFET). SWSFET was introduced by Jain et.al. These FETs have multiple channels stacked vertically. The Carrier wavefunction switches from one channel to another with the application of different gate voltages. Designs of multi-bit SRAM, logic and sequential cells using SWSFETs have been demonstrated. Here we are introducing the use of SWSFET in mixed signal architectures. Single cycle architectures for two-bit, four-bit and eight-bit analog to digital converters are presented. Four bit architecture has been simulated and results are discussed. SWSFET presents the opportunity with its multiple stacked channel features to extend the Moors law using next generation of devices.
    International Journal of High Speed Electronics and Systems 05/2014; 23(01n02).
  • [Show abstract] [Hide abstract]
    ABSTRACT: This paper describes novel multibit static random-access memories (SRAMs) implemented using four-channel spatial wavefunction switched field-effect transistors (SWS FETs) with Ge quantum wells and ZnSSe barriers. A two-bit SRAM cell consists of two back-to-back connected four-channel SWS FETs, where each SWS FET serves as a quaternary inverter. This architecture results in a reduction of the field-effect transistor (FET) count by 75% and data interconnect density by 50%. The designed two-bit SRAM cell is simulated using Berkeley short-channel insulated-gate field-effect transistor equivalent-channel models (for 25-nm FETs). In addition, the binary interface logic and conversion circuitry are designed to integrate the SWS SRAM technology. Our motivation is to stack up multiple bits on a single SRAM cell without multiplying the transistor count. The concept of spatial wavefunction switching (SWS) in the FET structure has been verified experimentally for two- and four-well structures. Quantum simulations exhibiting SWS in four-well Ge SWS FET structures, using the ZnSe/ZnS/ZnMgS/ZnSe gate insulator, are presented. These structures offer higher contrast than Si-SiGe SWS FETs.
    Journal of Electronic Materials 11/2013; 42(11). · 1.68 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Multistate behavior has been achieved in quantum dot gate field-effect transistor (QDGFET) configurations using either SiOx -cladded Si or GeOx -cladded Ge quantum dots (QDs) with asymmetric dot sizes. An alternative method is to use both SiOx -cladded Si and GeOx -cladded Ge QDs in QDGFETs. In this paper, we present experimental verification of four-state behavior observed in a QDGFET with cladded Si and Ge dots site-specifically self-assembled in the gate region over a thin SiO2 tunnel layer on a Si substrate. This paper also investigates the use of lattice-matched high-κ ZnS-ZnMgS-ZnS layers as a gate insulator in mixed-dot Si QDGFETs. Quantum-mechanical simulation of the transfer characteristic (I D–V G) shows four-state behavior with two intermediate states between the conventional ON and OFF states.
    Journal of Electronic Materials 11/2013; 42(11). · 1.68 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Three-state behavior has been demonstrated in Si and InGaAs field-effect transistors (FETs) when two layers of cladded quantum dots (QDs), such as SiOx -cladded Si or GeOx -cladded Ge, are assembled on the thin tunnel gate insulator. This paper describes FET structures that have the potential to exhibit four states. These structures include: (1) quantum dot gate (QDG) FETs with dissimilar dot layers, (2) quantum dot channel (QDC) with and without QDG layers, (3) spatial wavefunction switched (SWS) FETs with multiple coupled quantum well channels, and (4) hybrid SWS–QDC structures having multiple drains/sources. Four-state FETs enable compact low-power novel multivalued logic and two-bit memory architectures. Furthermore, we show that the performance of these FETs can be enhanced by the incorporation of II–VI nearly lattice-matched layers in place of gate oxides and quantum well/dot barriers or claddings. Lattice-matched high-energy gap layers cause reduction in interface state density and control of threshold voltage variability, while providing a higher dielectric constant than SiO2. Simulations involving self-consistent solutions of the Poisson and Schrödinger equations, and transfer probability rate from channel (well or dot layer) to gate (QD layer) are used to design sub-12-nm FETs, which will aid the design of multibit logic and memory cells.
    Journal of Electronic Materials 11/2013; 42(11). · 1.68 Impact Factor
  • Supriya Karmakar, J.A. Chandy, F.C. Jain
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we discuss logic circuit designs using the circuit model of three-state quantum dot gate field effect transistors (QDGFETs). QDGFETs produce one intermediate state between the two normal stable ON and OFF states due to a change in the threshold voltage over this range. We have developed a simplified circuit model that accounts for this intermediate state. Interesting logic can be implemented using QDGFETs. In this paper, we discuss the designs of various two-input three-state QDGFET gates, including NAND- and NOR-like operations and their application in different combinational circuits like decoder, multiplier, adder, and so on. Increased number of states in three-state QDGFETs will increase the number of bit-handling capability of this device and will help us to handle more number of bits at a time with less circuit elements.
    IEEE Transactions on Very Large Scale Integration (VLSI) Systems 05/2013; 21(5):793-806. · 1.14 Impact Factor
  • Orko Momin, John A. Chandy
    [Show abstract] [Hide abstract]
    ABSTRACT: In order to sustain scalability, clustered file systems distribute files across multiple I/O nodes in the cluster. The basis of many of these file systems is an object storage architecture where data is represented by variable sized containers called objects. The I/O nodes, however, use a POSIX file and directory interface which do not map well to the object interface. As a result, object storage nodes still have to resolve file paths and perform expensive lookup operations to find inodes and open objects stored as files. In this paper, we propose an alternative object based data interface where I/O node data can be accessed directly using object IDs instead of character strings. We modified an existing Linux file system to provide the object based interface which allows for faster data access times compared to the traditional directory-based interface. In addition, we modified a Object Storage Device (OSD) to use our object based interface to gain improvements in object creation and access. We believe this object based interface provides a useful alternative to the existing interface to provide better performance for cluster storage systems.
    Cluster Computing (CLUSTER), 2013 IEEE International Conference on; 01/2013
  • Cengiz Karakoyunlu, John A. Chandy
    [Show abstract] [Hide abstract]
    ABSTRACT: Object-Based Storage Devices (OSDs) offer an object-based data layout instead of using the traditional block-based design. Object storage ideas are increasingly being used in parallel file systems and cloud storage. Previous work has implemented the Parallel Virtual File System (PVFS) on OSDs by moving the metadata and data operations to the OSDs. In this paper, we present two new methods to further improve the performance of the existing PVFS-OSD implementation. These methods are using object collections to represent traditional directories on OSDs and delaying the creation of an OSD object until it is accessed by a write operation. The optimization methods are not particular to PVFS-OSD implementation and could be used with other parallel file systems-object storage implementations too. Experimental evaluations show that using collections yields up to 29 times of improvement in throughput; whereas delaying the creation of OSD objects provides 25% throughput improvement.
    Cluster Computing (CLUSTER), 2013 IEEE International Conference on; 01/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: Spatial Wavefunction-Switched (SWS) Field-Effect Transistors (FETs) consist of inversion layers comprising two or more coupled quantum wells (QWs). Carriers can be localized in any of the wells and vertically transferred between them by changing the gate voltage. In addition, carriers can also be laterally transferred between adjacent SWSFET devices by the manipulation of the gate voltages (Vg). This enables processing of two more bits simultaneously by changing the spatial location of the carrier ensemble wavefunction, which in turn determines the state of the device [e.g., electrons in well W2 (01), in W1 (10), in both (11), in neither (00)]. Experimentally, the capacitance-voltage data, having a distinct peak, has been presented in InGaAs-AlInAs two-quantum well structures. The peak(s) are attributed to the appearance of carriers, first in the lower well and subsequently their transfer to the upper well. Use of multiple channels allows for CMOS-like configuration with both transistors having n-channel mobilities. Simulation of an InGaAs SWS inverter computes a gate delay of 0.24ps. A cut-off frequency in excess of 8THz is computed for 12nm channel length InGaAs SWSFETs. Examples, including logic gates and a 3-bit full-adder, are presented to show the reduction of device count when SWS-FETs are employed.
    International Journal of High Speed Electronics and Systems 04/2012; 20(03).
  • [Show abstract] [Hide abstract]
    ABSTRACT: Three-state behavior has been demonstrated in Si and InGaAs quantum dot gate (QDG) field-effect transistorsootnotetextS. Karmakar, et al., J. Electronic Materials, 40, 1746, 2011.^,ootnotetextF Jain, J. Electronic Materials, 40, 1717, 2011. (FETs). Recently, spatial wavefunction switchedootnotetextIbid. (SWS) and quantum dot channelootnotetextF. Jain et al., Proc. II-VI Workshop, Oct.2011. (QDC) FETs have been reported to exhibit four-state operation. This paper presents simulations of versatile combinations of SWS features in QDC channels to optimally design multi-state transport in FETs that have the potential of scaling to sub-12nm regime. A QDC-FET channel is modeled as having superlattice-like mini-energy bands where the carrier wavefunctions are transferred across the channel as drain voltage is changed, producing step-like multi-state electrical characteristics. This behavior is analogous to that of single electron transistors.ootnotetextS. J. Shin, et al., Appl. Phys. Lett. 97, 103101, 2010. The difference is that QDC devices use more than a few electrons and operate at room temperature. The SWS feature additionally provides carrier transfer from lower to upper dot layer(s) in a QDC having more than one layer of quantum dots.
    02/2012;
  • [Show abstract] [Hide abstract]
    ABSTRACT: High-performance computing (HPC) storage systems rely on access coordination to ensure that concurrent updates do not produce incoherent results. HPC storage systems typically employ pessimistic distributed locking to provide this functionality in cases where applications cannot perform their own coordination. This approach, however, introduces significant performance overhead and complicates fault handling. In this work we evaluate the viability of optimistic conditional storage operations as an alternative to distributed locking in HPC storage systems. We investigate design strategies and compare the two approaches in a prototype object storage system using a parallel read/modify/write benchmark. Our prototype illustrates that conditional operations can be easily integrated into distributed object storage systems and can outperform standard coordination primitives for simple update workloads. Our experiments show that conditional updates can achieve over two orders of magnitude higher performance than pessimistic locking for some parallel read/modify/write workloads.
    High Performance Computing, Networking, Storage and Analysis (SCC), 2012 SC Companion:; 01/2012
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we present the design and implementation of an active storage framework for object storage devices. The framework is based on the use of virtual machines/execution engines to execute function code downloaded from client applications. We investigate the issues involved in supporting multiple execution engines. Allowing user-downloadable code fragments introduces potential safety and security considerations, and we study the effect of these considerations on these engines. In particular, we look at various remote procedure execution mechanisms and the efficiency and safety of these mechanisms. Finally, we present performance results of the active storage framework on a variety of applications.
    Mass Storage Systems and Technologies (MSST), 2012 IEEE 28th Symposium on; 01/2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents the multiple quantum well channel spatial wavefunction switched FETs (SWS-FETs) configured to implement multi-bit static random access memory (SRAM) cells. A 2-bit SRAM cell consists of two back-to-back connected 4-channel SWS-FETs, where each SWS-FET serves as quaternary inverter. This architecture results in reduction of FET count by 75% and data interconnect density by 50%. The designed 2-bit SRAM cell is simulated using BSIM equivalent channel models (for 25nm FETs). In addition, the binary interface logic and conversion circuitry are designed to integrate the SWS-SRAM technology. Quantum simulations for Si/Ge and InGaAs-based SWS-FETs are also presented.
    Lester Eastman Conference on High Performance Devices (LEC), 2012; 01/2012
  • C. Karakoyunlu, J.A. Chandy
    [Show abstract] [Hide abstract]
    ABSTRACT: The decreasing cost of disk drives has made it possible to maintain large-scale storage systems and to achieve high performance for I/O demanding applications. One possible optimization that can be applied to these types of systems is to move the I/O load to a subset of the disks; so that the commonly accessed data resides in a set of 'popular' disks, whereas the 'unpopular' disks can stay in low energy modes. There have been several approaches to this kind of optimization both at the hardware and the software level. In this paper, we propose techniques for an Energy-Aware Parallel File System and compare different server assignment configurations to understand what kind of improvements can be achieved at the file system level to save energy while maintaining acceptable performance.
    Green Computing Conference (IGCC), 2012 International; 01/2012
  • S. Karmakar, J.A. Chandy, F.C. Jain
    [Show abstract] [Hide abstract]
    ABSTRACT: The spatial wave-function switched field effect transistor (SWSFET) has two or three low band-gap quantum well channels that can conduct carrier flow from source to drain of the SWSFET. Because of this property, SWSFETs are useful to implement different multi-valued logic with reduced device count. In this work, we introduce the circuit model of a SWSFET and the design of a unipolar inverter where only one kind of charge carrier contributes to the current flow.
    Lester Eastman Conference on High Performance Devices (LEC), 2012; 01/2012
  • Source
    Ajithkumar Thamarakuzhi, John A. Chandy
    [Show abstract] [Hide abstract]
    ABSTRACT: High-performance computing is highly dependent on the communication network connecting the nodes. In this paper, we propose a 2-Dilated flattened butterfly (2DFB) network which provides non-blocking performance for relatively low cost overhead. We study the topological properties of the proposed 2DFB network and compare it with different nonblocking switching topologies. We observe that a dilation factor of two is sufficient to obtain nonblocking property for a flattened butterfly structure irrespective of its size or dimension. Dilating each link in a flattened butterfly causes an increase in cost. Therefore, we modeled the implementation cost of a 2DFB network and compared it with other popular nonblocking networks. We observe that the cost of a 2DFB is less than other nonblocking networks, while at the same time providing reduced latency because of its reduced diameter and hop count. We also propose a procedure to develop a conflict-free static routing schedule as well as an adaptive load balanced routing scheme (ALDFB) for 2DFB networks. Finally, we also describe the hardware implementation of a 2DFB network using the NetFPGA as the switching element and verify the nonblocking behavior of a 2DFB. We also show that the 2DFB topology can be used to build high speed switching systems with reduced cost.
    Computer Communications 09/2011; 34(15):1822-1835. · 1.35 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: High-performance computing is highly dependent on the communication network connecting the nodes. In this paper, we present a design and implementation of the 2-dilated flattened butterfly (2DFB) network topology, which provides nonblocking performance for relatively low cost overhead. The flattened butterfly is known to be a cost efficient topology for high-radix networks. Because of its inherent path diversity, it is able to provide comparable cost/performance with the folded-Clos network in adversarial traffic conditions. However, the flattened butterfly is a blocking network and this blocking behavior can degrade the performance of the network when all compute nodes are transmitting and receiving data at full bandwidth. 2DFB network exhibits nonblocking behavior which is a critical factor in deciding the quality of any high performance computing system network. We have implemented the network using the NetFPGA as the switching element and verified the nonblocking behavior of 2DFB. We also compare the performance of a 2DFB with other switching networks. We also show that the 2DFB topology can be used to build high speed switching systems with reduced cost.
    IEEE Latin America Transactions 08/2011; · 0.19 Impact Factor