-
PCIe Hot Plug support standardization challenges in ATCA
Authors:
Miguel Correia,
Jorge Sousa,
António P. Rodrigues,
Paulo F. Carvalho,
Bruno Santos,
Álvaro Combo,
Carlos M. B. A. Correia,
Bruno Gonçalves
Abstract:
Throughout the last decade, the Advanced Telecommunications Computing Architecture (ATCA) solidified its position as one of the main switched-based crate standards for advanced Physics instrumentation, offering not only highly performant characteristics in data throughput, channel density or power supply/dissipation capabilities, but also special features for high availability (HA), required for l…
▽ More
Throughout the last decade, the Advanced Telecommunications Computing Architecture (ATCA) solidified its position as one of the main switched-based crate standards for advanced Physics instrumentation, offering not only highly performant characteristics in data throughput, channel density or power supply/dissipation capabilities, but also special features for high availability (HA), required for latest and upcoming large-scale endeavours, as is the case of ITER. Hot Swap is one of the main HA features in ATCA, allowing for Boards to be replaced in a crate (Shelf), without powering off the whole system. Platforms using the Peripheral Component Interconnect Express (PCIe) protocol on the Fabric Interface must be complemented, at the software level, with the PCIe Hot Plug native feature, currently not specified for the ATCA form-factor. From a customised Hot Plug support implementation for ATCA Node Boards, the paper presents an implementation extension for Hub Boards, allowing Hot Plug of PCIe switching devices, without causing bus enumeration problems. This paper further addresses the main issues concerning an eventual standardization of PCIe Hot Plug support in ATCA, such as the implementability of Hot Plug Elements and the generation and management of Hot Plug Events, aiming to stimulate the discussion within the PICMG community towards a long overdue standardized solution for Hot Plug in ATCA.
△ Less
Submitted 22 June, 2018;
originally announced June 2018.
-
The design and performance of the real time software architecture for the ITER Radial Neutron Camera
Authors:
N. Cruz,
B. Santos,
A. Fernandes,
P. F. Carvalho,
J. Sousa,
B. Gonçalves,
M. Riva,
C. Centioli,
D. Marocco,
B. Esposito,
C. M. B. Correia,
R. C. Pereira
Abstract:
The neutron detection system for characterization of emissivity in ITER Tokamak during DD and DT experiments poses serious challenges to the performance of the diagnostic control and data acquisition system (CDAcq). The ongoing design of the ITER Radial Neutron Camera (RNC) diagnostic is composed by 26 lines of sight (LOS) for complete plasma inspection. The CDAcq system aims at meeting the ITER r…
▽ More
The neutron detection system for characterization of emissivity in ITER Tokamak during DD and DT experiments poses serious challenges to the performance of the diagnostic control and data acquisition system (CDAcq). The ongoing design of the ITER Radial Neutron Camera (RNC) diagnostic is composed by 26 lines of sight (LOS) for complete plasma inspection. The CDAcq system aims at meeting the ITER requirements of delivering the measurement of the real-time neutron emissivity profile with time resolution and control cycle time of 10 ms at peak event rate of 2 MEvents/s per LOS. This measurement demands the generation of the neutron spectra for each LOS with neutron/gamma discrimination and pile up rejection. The neutron spectra can be totally processed in the host CPU or it can use the processed data coming from the system FPGA. The number of neutron counts extracted from the spectra is then used to calculate the neutron emissivity profile using an inversion algorithm. Moreover, it is required that the event based raw data acquired is made available to the ITER data network without local data storage for post processing. The data production for the 2 MEvents/s rate can go up to a maximum data throughput of 0.5 GB/s per channel. The evaluation of the use of real-time data compression techniques in RNC is also depicted in another contribution.
To meet the demands of the project a CDAcq prototype has been used to design and test a high-performance distributed software architecture taking advantage of multi-core CPU technology capable of co** with the requirements. This submission depicts the design of the real-time architecture, the spectra algorithms (pulse height analysis, neutron/gamma discrimination and pile-up correction) and the inversion algorithm to calculate the emissivity profile. Preliminary tests to evaluate the system performance with synthetic data are presented.
△ Less
Submitted 22 June, 2018;
originally announced June 2018.
-
FPGA code for the data acquisition and real-time processing prototype of the ITER Radial Neutron Camera
Authors:
Ana Fernandes,
Nuno Cruz,
Bruno Santos,
Paulo F. Carvalho,
Jorge Sousa,
Bruno Gonçalves,
Marco Riva,
Fabio Pollastrone,
Cristina Centioli,
Daniele Marocco,
Basilio Esposito,
Carlos M. B. A. Correia,
Rita C. Pereira
Abstract:
The main role of the ITER Radial Neutron Camera (RNC) diagnostic is to measure in real-time the plasma neutron emissivity profile at high peak count rates for a time duration up to 500 s. Due to the unprecedented high performance conditions and after the identification of critical problems, a set of activities have been selected, focused on the development of high priority prototypes, capable to d…
▽ More
The main role of the ITER Radial Neutron Camera (RNC) diagnostic is to measure in real-time the plasma neutron emissivity profile at high peak count rates for a time duration up to 500 s. Due to the unprecedented high performance conditions and after the identification of critical problems, a set of activities have been selected, focused on the development of high priority prototypes, capable to deliver answers to those problems before the final RNC design. This paper presents one of the selected activities: the design, development and testing of a dedicated FPGA code for the RNC Data Acquisition prototype. The FPGA code aims to acquire, process and store in real-time the neutron and gamma pulses from the detectors located in collimated lines of sight viewing a poloidal plasma section from the ITER Equatorial Port Plug 1. The hardware platform used was an evaluation board from Xilinx (KC705) carrying an IPFN FPGA Mezzanine Card (FMC-AD2-1600) with 2 digitizer channels of 12-bit resolution sampling up to 1.6 GSamples/s. The code performs the proper input signal conditioning using a down-sampled configuration to 400 MSamples/s, apply dedicated algorithms for pulse detection, filtering and pileup detection, and includes two distinct data paths operating simultaneously: i) the event-based data-path for pulse storage; and ii) the real-time processing, with dedicated algorithms for pulse shape discrimination and pulse height spectra. For continuous data throughput both data-paths are streamed to the host through two distinct PCIe x8 Direct Memory Access (DMA) channels.
△ Less
Submitted 15 June, 2018;
originally announced June 2018.
-
Real-time data compression for data acquisition systems applied to the ITER Radial Neutron Camera
Authors:
B. Santos,
N. Cruz,
A. Fernandes,
P. F. Carvalho,
J. Sousa,
B. Gonçalves,
M. Riva,
F. Pollastrone,
C. Centioli,
D. Marocco,
B. Esposito,
C. M. B. Correia,
R. C. Pereira
Abstract:
To achieve the aim of the ITER Radial Neutron Camera Diagnostic, the data acquisition prototype must be compliant with a sustained 2 MHz peak event for each channel with 128 samples of 16 bits per event. The data is acquired and processed using an IPFN FPGA Mezzanine Card (FMC-AD2-1600) with 2 digitizer channels of 12-bit resolution and a sampling rate up to 1.6 GSamples/s mounted in a PCIe evalua…
▽ More
To achieve the aim of the ITER Radial Neutron Camera Diagnostic, the data acquisition prototype must be compliant with a sustained 2 MHz peak event for each channel with 128 samples of 16 bits per event. The data is acquired and processed using an IPFN FPGA Mezzanine Card (FMC-AD2-1600) with 2 digitizer channels of 12-bit resolution and a sampling rate up to 1.6 GSamples/s mounted in a PCIe evaluation board from Xilinx (KC705) installed in the host PC. The acquired data in the event-based data-path is streamed to the host through the PCIe x8 Direct Memory Access (DMA) with a maximum data throughput per channel is 0.5 GB/s of raw data (event base), 1 GB/s per digitizer and up to 1.6 GB/s in continuous mode. The prototype architecture comprises an host PC with two KC705 modules and four channels, producing up to 2 GB/s in event mode and up to 3.2 GB/s in continuous mode. To reduce the produced data throughput from host to ITER databases, the real-time data compression was evaluated using the LZ4 lossless compression algorithm, which provides compression speed up to 400 MB/s per core. This paper presents the architecture, implementation and test of the parallel real-time data compression system running in multiple isolated cores. The average space saving and the performance results for long term acquisitions up to 30 minutes, using different data block size and different number of CPUs, is also presented.
△ Less
Submitted 21 June, 2018; v1 submitted 11 June, 2018;
originally announced June 2018.
