A DC-DC boost converter is widely used for power management applications. However, the right-half-plane zero effect still remains a bottleneck for fast transient recovery. This paper is intended to propose an embedded...
详细信息
ISBN:
(纸本)9780769548890
A DC-DC boost converter is widely used for power management applications. However, the right-half-plane zero effect still remains a bottleneck for fast transient recovery. This paper is intended to propose an embedded reconfigurable boost converter solution with ultra-fast transient recovery through near-lossless augmentation. This augmentation branch requires an extra synchronous phase in parallel, comprising a small inductor and a half-bridge switch cell. This branch is activated only during large-signal transient recovery. The original converter topology is restored for steady state operations, thereby improving the overall efficiency. A prototype augmented boost converter is tested, and the proposed controller is implemented using an fpga device. The extra branch closely nullifies current and voltage overshoot/undershoots. It is also expected to have almost no impact on efficiency.
The use of Soft Physical Hash (SPH) functions has been recently introduced as a flexible and efficient way to detect Intellectual Property (IP) cores in microelectronic systems. Previous works have mainly investigated...
详细信息
ISBN:
(纸本)9781479906017;9781479905591
The use of Soft Physical Hash (SPH) functions has been recently introduced as a flexible and efficient way to detect Intellectual Property (IP) cores in microelectronic systems. Previous works have mainly investigated software IP to validate this approach. In this paper, we extend it towards the practically important case of fpga designs. Based on experiments, we put forward that SPH functions-based detection is a promising and low-cost solution for preventing anti-counterfeiting, as it does not require any a-priori modification of the design flow. In particular, we illustrate its performances with stand-alone fpga designs, re-synthetized fpga designs, and in the context of parasitic IPs running in parallel.
This paper presents a high speed configurable fpga architecture for k-means clustering. The proposed architecture is highly pipelined, parallel and fully configurable. It can achieve an operating frequency of 400 MHz,...
详细信息
ISBN:
(纸本)9781467357623;9781467357609
This paper presents a high speed configurable fpga architecture for k-means clustering. The proposed architecture is highly pipelined, parallel and fully configurable. It can achieve an operating frequency of 400 MHz, which is at least three times faster than prior works. The proposed architecture addresses the high speed and throughput requirements of machine vision, multi-media and data mining applications.
Data retention time distribution of a dynamic random access memory (DRAM) has a heavy impact on its yield, power, and performance. Accurate and detailed information of data retention time distribution thus is very imp...
详细信息
ISBN:
(纸本)9781467344364
Data retention time distribution of a dynamic random access memory (DRAM) has a heavy impact on its yield, power, and performance. Accurate and detailed information of data retention time distribution thus is very important for the DRAM designer and user. This paper proposes an fpga-based test platform for analyzing the data retention time distribution of a DRAM. Based on the test platform, a test flow is also proposed to classify the DRAM cells with different data retention times with respect to different supply voltage and temperature. We have demonstrated the test platform and test flow using a Micron 2Gb DRAM.
Apart from traditional test and measurement systems where clock synchronization is required, new emerging application areas like SmartGrids and 4G cellular mobile backhaul networks present strong timing constraints in...
详细信息
ISBN:
(纸本)9781467351942;9781467351928
Apart from traditional test and measurement systems where clock synchronization is required, new emerging application areas like SmartGrids and 4G cellular mobile backhaul networks present strong timing constraints in terms of precise time synchronization. Precision Time Protocol (PTP), as defined in IEEE 1588 standard, offers sub-microsecond synchronization using conventional Ethernet networks. Thus, its acceptance is heavily increasing. However, the protocol performance was reduced in large cascaded networks with varying latencies. This drawback was later softened by the second version of the standard with the introduction of the Transparent Clock (TC) device. In this paper, a general overview of PTPv2 and the utilization of TCs is outlined. The main contribution is a new TC architecture for a fpga-based network device that benefits from reconfigurable devices flexibility.
Time-to-digital converter (TDC) implemented in field-programmable-gatearrays (fpgas) with the use of tapped delay line (TDL) methods using dedicated carry chain structures has been utilized to obtain precise timing i...
详细信息
ISBN:
(纸本)9781479905348
Time-to-digital converter (TDC) implemented in field-programmable-gatearrays (fpgas) with the use of tapped delay line (TDL) methods using dedicated carry chain structures has been utilized to obtain precise timing information for time-of-flight (TOF) PET. However, such chain-structural fpga-based TDCs suffer from the uneven tap delay, which is one of the main limitations for enhancing the timing performance of TDC. To subdivide the non-uniform tap delay, we propose a TDC with dual-TDL having different starting points. In a fpga (Spartan-6 LX45, Xilinx USA), two delay lines for a channel were implemented close to the fpga input pad and one of two delay lines had starting point in further slice from the input pad. In calibration process, the size of wide time bins from the uneven tap delay was reduced using different characteristics of two delay lines. In simulation study, the average time bin size was reduced from 24.7 ps to 12.6 ps and the size of widest time bin was reduced from 168 ps to 56 ps. The performance of the proposed TDC was measured with two input pulses that were generated from a pulse generator. The timing resolution of the TDC was 35 ps FWHM which was improved compared to single-TDL TDC pair (56 ps FWHM). The results of this preliminary study present dual-TDL calibration can effectively improve the time resolution of TDC. The subject of further study will be to extend the number of TDC channels in a fpga for TOF-PET applications.
Soft errors in the configuration memory of SRAM-based fpgas cause significant application disturbances. We demonstrate on Xilinx and Altera fpgas the feasibility of a very low cost and automated mitigation approach an...
详细信息
This paper explores the potential of partitioning the dynamic programming algorithm to utilise the capabilities of fpga platforms for parallel genome sequence comparison and assembly. We use this to solve the prefix-s...
详细信息
ISBN:
(纸本)9781467357623;9781467357609
This paper explores the potential of partitioning the dynamic programming algorithm to utilise the capabilities of fpga platforms for parallel genome sequence comparison and assembly. We use this to solve the prefix-suffix approximate matching problem to find overlaps between DNA strands in a given sequence. This is achieved by partitioning the basic dynamic programming (DP) algorithm into a series of discrete sub-DP calculations. Analysis of the error rate as a result of this partitioning by applying random sequences as input data is shown, and simulation results confirm good matching with the original algorithm with an error of less that 1.5% in the worst case. Optimisation for array implementation in fpga is shown and linear scalability for larger arrays is proven. Simulation results of the whole system confirm 98.8% similarity of the overlap adjacent matrix between error and error-free data.
This paper presents a new Dynamic Partial Self Reconfiguration (DPSR) flow for Xilinx fpgas. Leveraging the Xilinx fpga Editor and Plan Ahead tools, we provide two implementation approaches that enable partial reconfi...
详细信息
ISBN:
(纸本)9781479935253
This paper presents a new Dynamic Partial Self Reconfiguration (DPSR) flow for Xilinx fpgas. Leveraging the Xilinx fpga Editor and Plan Ahead tools, we provide two implementation approaches that enable partial reconfiguration for large configuration changes without Xilinx's paid tool. The flow is difference-based but still allows a modular design, which is made up of Partial Reconfiguration (PR) modules and a static design. It works regardless of the amount of difference between PR modules. We call this flow DPSR-LD, where LD stands for Large Differences. DPSR-LD is an enabler especially for Spartan-6 fpga family, as Xilinx currently supports PR on Spartan-6 only through the difference-based flow and only for small differences. DPSR-LD also includes an ICAP controller that makes DPSR possible and offers bitstream compression.
3D stacked integrated circuits hold great promise for increasing system performance, but difficulties in testing dies and assembling a 3D stack are leading to yield issues and slowing the large scale manufacture of th...
详细信息
ISBN:
(纸本)9781479915859;9781479915835
3D stacked integrated circuits hold great promise for increasing system performance, but difficulties in testing dies and assembling a 3D stack are leading to yield issues and slowing the large scale manufacture of these devices. We propose helping to mitigate these issues by repairing the stack with programmable logic in fpgas that have already been included in the stack for other purposes. Specifically, we propose bypassing the defective portion of a die by replacing the defective functionality with functionality on the fpga. In this paper, we focus on the replacement of selected defective functional units in an out-of-order microprocessor. Our simulation results show that not only can we salvage a device that would otherwise have to be discarded, but creating multiple copies of the defective partition in the fpga can allow us to regain performance even when the latency of the units in the fpga is longer than that of the original defective copy.
暂无评论