52
R. Lulli et al.
put on the market high-end FPGAs based accelerator board, but the prices are still
today quite high (order of e10,000).
In the years, many-cores and FPGAs market became so crucial that these products
shared the same semiconductor lithographic processes, and they were composed of
the same number (order) of transistor than high-end CPUs. The price of these products
decreased, and the Software Development Kits (SDKs) evolved, letting to commonly
use them for various High-Performance Computing (HPC) applications (scientific
computing, industrial simulations, Big Data analysis, Artificial Intelligence (AI),
etc.).
7.1.3 New Design Guidelines for PC Based Spectrometers
In 2010, Dr. Eng. Stelio Montebugnoli, searching for a new approach to develop a new
line of spectrometers suitable for radio observations and following the computing
hardware evolution of the 2000s, proposed at the annual IAC in Prague (CZ) a new
idea for low-cost spectrometer design.
Dr. Montebugnoli proposed to design a low-cost, high-resolution spectrometer
for SETI researches using COTS components, like standard multi-core and/or multisocket PCs, industry-standard DAQ boards, and GPUs or FPGAs as computing
accelerators.
Dr. Montebugnoli concluded that “The algorithm described in this article for SETI
investigation can be implemented on several systems with a low and medium cost.
Taking into account the advantages and disadvantages shown, the user can find the
best solution (in terms of costs, performances, the difficulty of programming, I/O rate,
power consumption, processing bandwidth, etc.) to make its SETI back end. Even
with a small budget, it is possible to reach high-performance back end, especially
using GPU systems” [1].
7.2 The Design of the New Spectrometers
7.2.1 NewSpec: The First Approach for a New Spectrometer
Following the new design guidelines, the first approach to design a spectrometer
began in 2012 with the development of the NewSpec project. The base objective
of this project was to design and develop a low cost (order of e10,000), highresolution (Hz order resolution) SETI spectrometer to replace the dismissed Serendip
IV [15]. But, the final goal was to have a final spectrometer able to analyze all the
500 MHz Intermediate Frequency (IF) bandwidth coming from the first L band
receiver (1.35–1.45 GHz) of the 32 m parabolic dish (Medicina RT-32) without
another downconversion.
R. Lulli et al.
put on the market high-end FPGAs based accelerator board, but the prices are still
today quite high (order of e10,000).
In the years, many-cores and FPGAs market became so crucial that these products
shared the same semiconductor lithographic processes, and they were composed of
the same number (order) of transistor than high-end CPUs. The price of these products
decreased, and the Software Development Kits (SDKs) evolved, letting to commonly
use them for various High-Performance Computing (HPC) applications (scientific
computing, industrial simulations, Big Data analysis, Artificial Intelligence (AI),
etc.).
7.1.3 New Design Guidelines for PC Based Spectrometers
In 2010, Dr. Eng. Stelio Montebugnoli, searching for a new approach to develop a new
line of spectrometers suitable for radio observations and following the computing
hardware evolution of the 2000s, proposed at the annual IAC in Prague (CZ) a new
idea for low-cost spectrometer design.
Dr. Montebugnoli proposed to design a low-cost, high-resolution spectrometer
for SETI researches using COTS components, like standard multi-core and/or multisocket PCs, industry-standard DAQ boards, and GPUs or FPGAs as computing
accelerators.
Dr. Montebugnoli concluded that “The algorithm described in this article for SETI
investigation can be implemented on several systems with a low and medium cost.
Taking into account the advantages and disadvantages shown, the user can find the
best solution (in terms of costs, performances, the difficulty of programming, I/O rate,
power consumption, processing bandwidth, etc.) to make its SETI back end. Even
with a small budget, it is possible to reach high-performance back end, especially
using GPU systems” [1].
7.2 The Design of the New Spectrometers
7.2.1 NewSpec: The First Approach for a New Spectrometer
Following the new design guidelines, the first approach to design a spectrometer
began in 2012 with the development of the NewSpec project. The base objective
of this project was to design and develop a low cost (order of e10,000), highresolution (Hz order resolution) SETI spectrometer to replace the dismissed Serendip
IV [15]. But, the final goal was to have a final spectrometer able to analyze all the
500 MHz Intermediate Frequency (IF) bandwidth coming from the first L band
receiver (1.35–1.45 GHz) of the 32 m parabolic dish (Medicina RT-32) without
another downconversion.