Publications of the IPE expert group for embedded parallel systems
Caselle M., Chilingaryan S., Herth A., Kopmann A., Stevanovic U., Vogelgesang M., Balzer M., Weber M.
in IEEE Transactions on Nuclear Science, 60 (2013) 3669-3677, 6510495. DOI:10.1109/TNS.2013.2252528
X-ray computer tomography is a powerful method for nondestructive investigations in many fields. Three-dimensional images of internal structure are reconstructed from a sequence of two-dimensional projections. The polychromatic high density photon flux of modern synchrotron light sources offer hard X-ray imaging with spatio-temporal resolution up to the micrometer and micrometers range. Existing indirect X-ray image detection systems can be adapted for fast image acquisition by high-speed visible-light cameras. In this paper, we present a platform for custom high-speed CMOS cameras with embedded field-programmable gate array (FPGA) processing. This modular system is characterized by a high-throughput PCI Express (PCIe) interface and efficient communication blocks. It has been used to develop a novel architecture for a self-event trigger that increases the effective image frame rate and reduces the amount of received data. Thanks to a low-noise design, high frame rates in the kilohertz range, and high-throughput data transfer, this camera is well suited for ultrafast synchrotron-based X-ray radiography and tomography. The camera setup is accomplished by high-throughput Linux drivers and a seamless integration in our GPU computing framework. © 2013 British Crown Copyright.
Mattiazzo S., Battaglia M., Bisello D., Caselle M., Chalmet P., Demaria N., Giubilato P., Ikemoto Y., Kloukinas K., Mansuy C., Marchioro A., Mugnier H., Pantano D., Potenza A., Rivetti A., Rousset J., Silvestrin L., Snoeys W., Wyss J.
in Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 718 (2013) 288-291. DOI:10.1016/j.nima.2012.10.098
We present a monolithic pixel sensor developed in the framework of the LePIX project aimed at tracking/triggering tasks where high granularity, low power consumption, material budget, radiation hardness and production costs are a concern. The detector is built in a 90 nm CMOS process on a substrate of moderate resistivity. This maintains the advantages usually offered by Monolithic Active Pixel Sensors (MAPS), like a low input capacitance, having a single piece detector and using a standard CMOS production line, but offers charge collection by drift from a depleted region and therefore an excellent signal to noise ratio and a radiation tolerance superior to conventional undepleted MAPS. Measurement results obtained with the first prototypes from laser, radioactive source and beam test experiments are described. The excellent signal-to-noise performance is demonstrated by the capability of the device to separate the peaks in the spectrum of a 55Fe source. We will also highlight the interaction between pixel cell design and architecture which points toward a very precise direction in the development of such depleted monolithic pixel devices for high energy physics. © 2012 Elsevier B.V. All rights reserved.
Potenza A., Bisello D., Caselle M., Costa M., Demaria N., Giubilato P., Ikemoto Y., Mansuy C., Marchioro A., Mattiazzo S., Moll M., Pacher L., Pacifico N., Pantano D., Rivetti A., Silvestrin L., Snoeys W.
in Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 718 (2013) 347-349. DOI:10.1016/j.nima.2012.10.020
The LePix project aims at developing monolithic pixel detectors in a 90 nm CMOS technology ported on moderate resistivity substrate. The radiation tolerance of the base material, which is an order of magnitude higher doped than standard high resistivity detectors, and which underwent the full advanced CMOS process, has been investigated. Diodes of about 1 mm2 and pixel matrices were irradiated with neutrons at fluences from 1012 n/cm2 to 2 × 1015 n/cm2 and characterized using CV and IV measurements. Matrices have also been irradiated with Xrays and withstand at least 10 Mrad. © 2012 Elsevier B.V. All rights reserved.
Caselle M., Chilingaryan S., Herth A., Kopmann A., Stevanovic U., Vogelgesang M., Balzer M., Weber M.
in 2012 18th IEEE-NPSS Real Time Conference, RT 2012 (2012), 6418369. DOI:10.1109/RTC.2012.6418369
X-ray computed tomography (CT) is a method for non-destructive investigation. Three-dimensional images of internal structure can be reconstructed using a two-dimensional detector. The poly-chromatic high density photon flux in the modern synchrotron light sources offers hard X-ray imaging with a spatio-temporal resolution up to the μm-μs range. Existing indirect X-ray image detectors can be adapted for fast image acquisition by employing CMOS-based digital high speed camera. In this paper, we propose a high-speed visible light camera based on commercial CMOS sensor with embedded processing implemented in FPGA. This platform has been used to develop a novel architecture for a self-event trigger. This feature is able to increase the original frame rate of the CMOS sensor and reduce the amount of the received data. Thanks to a low noise design, high frame rate (kilohertz range) and high speed data transfer, this camera can be employed in modern synchrotron ultra-fast X-ray radiography and computed tomography. The camera setup is accomplished by high-throughput Linux drivers and a seamless integration in our GPU computing framework. Selected applications from life sciences and materials research underline the high potential of this high-speed camera in a hard X-ray micro-imaging approach. © 2012 IEEE.
Stevanovic U., Caselle M., Chilingaryan S., Herth A., Kopmann A., Vogelgesang M., Balzer M., Weber M.
in Conference on Design and Architectures for Signal and Image Processing, DASIP (2012) 383-384, 6385417.
The first prototype of a high-speed camera with embedded image processing has been developed. Beside high frame rate and high through-put, the camera introduces a novel self triggering architecture to increase the frame rate and to reduce the amount of received data. The camera is intended for synchrotron ultra-fast X-ray radiography and tomography, but it’s concept is also suitable for other fields. © 2012 ECSI.
Birk M., Balzer M., Ruiter N., Becker J.
in 2012 International Conference on Reconfigurable Computing and FPGAs, ReConFig 2012 (2012), 6416735. DOI:10.1109/ReConFig.2012.6416735
With the rise of heterogeneous computing architectures, application developers are confronted with a multitude of hardware platforms and the challenge of identifying the most suitable processing platform for their application. Strong competitors for the acceleration of 3D Ultrasound Computer Tomography, a medical imaging method for early breast cancer diagnosis, are GPU and FPGA devices. In this work, we evaluate processing performance and efficiency metrics for current FPGA and GPU devices. We compare top-notch devices from the 40 nm generation as well as FPGA and GPU devices, which draw the same amount of power. For our two benchmark algorithms, the results show that if power consumption is not considered the GPU and the FPGA give both, a similar processing performance and processing efficiency per transistor. However, if the power budget is limited to a similar value, the FPGA performs between six and eight times better than the GPU. © 2012 IEEE.
Balzer M., Birk M., Dapp R., Gemmeke H., Kretzek E., Menshikov S., Zapf M., Ruiter N.V.
in 2012 18th IEEE-NPSS Real Time Conference, RT 2012 (2012), 6418198. DOI:10.1109/RTC.2012.6418198
3D ultrasound computer tomography (USCT) is a new and promising method for early breast cancer diagnosis. An ultrasound computer tomograph was developed by the Karlsruhe Institute of Technology (KIT) and provides a resolution of 0.2 mm. The main components are the semi-ellipsoidal aperture with 628 ultrasound (US) emitters and 1413 US receivers and the 480 input channel data acquisition system. An additional external computing system is used for the time consuming image reconstruction. To reduce the reconstruction time different implementations on massive parallel computing architectures like multicore processor, GPUs and FPGAs were tested. One realization applies reconfiguration of the 60 FPGAs inside the DAQ system. The investigations show significant acceleration of the reconstruction time up to a factor of 15.8 for the latest FPGA generation and 17.6 for a state of the art GPU. © 2012 IEEE.
Haas D., Mexner W., Spangenberg T., Cecilia A., Vagovic P., Kopmann A., Balzer M., Vogelgesang M., Pasic H., Chilingaryan S.
in PCaPAC 2012 – 9th International Workshop on Personal Computers and Particle Accelerator Controls (2012) 103-105.
X-ray imaging permits to spatially resolve the 2D and 3D structure in materials and organisms, which is crucial for the understanding of their properties. Additional temporal resolution of structure evolution gives access to dynamics of processes and allows to understand functionality of devices and organisms with the goal to optimize technological processes. Such time-resolved dynamic analysis of micro-sized structures is now possible by aid of ultrafast tomography, as being developed at the TopoTomo beamline of the synchrotron light source ANKA. At TopoTomo, the whole experimental workflow has been significantly improved in order to decrease the total duration of a tomography experiment down to the range of minutes. To meet these requirements, detectors and the computing infrastructure have been optimized, comprising a Tango-based control system for ultra fast tomography with a data throughput of several 100 MB/s. Multi-GPU based computing allows for high speed data processing by using a special reconstruction scheme. Furthermore the data management infrastructure will allow for a life cycle management of data sets accumulating several TByte/day. The final concept will also be part of the IMAGE beamline, which is going to be installed in 2013. © 2012 by the respective authors.
Balzer M., Caselle M., Chilingaryian S., Herth A., Kopmann A., Stevanovic U., Vogelgesang M., Rolo T.D.S.
in SEI 2012 – 103. Tagung der Studiengruppe Elektronische Instrumentierung im Fruhjahr 2012 (2012) 121-132.
Birk M., Guth A., Zapf M., Balzer M., Ruiter N., Hubner M., Becker J.
in Conference on Design and Architectures for Signal and Image Processing, DASIP (2011) 67-74, 6136856. DOI:10.1109/DASIP.2011.6136856
As today’s standard screening methods frequently fail to diagnose breast cancer before metastases have developed, earlier breast cancer diagnosis is still a major challenge. Three-dimensional ultrasound computer tomography promises high-quality images of the breast, but is currently limited by a time-consuming synthetic aperture focusing technique based image reconstruction. In this work, we investigate the acceleration of the image reconstruction by a GPU, and by the FPGAs embedded in our custom data acquisition system. We compare the obtained performance results with a recent multi-core CPU and show that both platforms are able to accelerate processing. The GPU reaches the highest performance. Furthermore, we draw conclusions in terms of applicability of the accelerated reconstructions in future clinical application and highlight general principles for speed-up on GPUs and FPGAs. © 2011 IEEE.