High performance FPGA-‐based implementa9on of the seismic modeling of the RTM algorithm
Hardware accelerators like GPGPUs and FPGAs have been used as
an alternative for the conventional computing architectures
(CPUs) in scientific computing applications and have shown
considerable speed-ups. In this context, this work, held at the
Federal University of Pernambuco, presents a solution that takes
advantage from FPGA’s flexibility to explore efficiently data reuse,
parallelization in both time and space domains for the first processing stage of the RTM (Reverse Time Migration) algorithm,
the seismic modeling.
To read the full paper, please click here:
Novo-G: Most Powerful Reconfigurable Supercomputer
It may be that the supercomputer, named Novo-G, developed at the Center for High-Performance Reconfigurable Computing (CHREC) is the world’s fastest reconfigurable supercomputer; likely faster than the Japanese supercomputer
touted as the world’s most powerful. This innovative new form of supercomputing has a dramatically different internal structure than existing supercomputers.
To read the full paper, please click here:
Throughput Analysis for a High-Performance FPGA-Accelerated Real-Time Search Application
The focus on real-time search is growing with the increasing
adoption and spread and of social networking applications.
Real-time search is equally important in other areas such as
analysing emails for spam or search web traffic for particular
patterns.
FPGAs have great potential for speeding up many types
of applications and algorithms. By performing a task in a
fraction of the time of a conventional processor, large energy
savings can be achieved. Therefore, there is a growing interest
in the use of FPGA platforms for data centres.
To read the full paper, please click here:
Acoustic Real-Time, Low-Power FPGA Based Obstacle Detection For AUVs
The underwater robots called Unmanned
Underwater Vehicles (UUVs) take over complex and dangerous
underwater missions that were previously performed by humans.
These vehicles operate in the unknown environments and make
their own decisions within the mission based on the readings of
the sensors, without any link with a human operator.
Independent of the mission, it is critical for the AUVs to be able
to avoid submerged obstacles such as cliffs, wrecks, and floating
mines. The AUV typically uses underwater imaging sonar that
has several drawbacks for obstacle detection purposes, and
therefore requires complex image processing algorithms. Due to
the imaging sonar limitations, addressing obstacle detection
using conventional software algorithms cannot meet an AUV’s
real-time, low power requirements. A low-power FPGA
algorithm for underwater obstacle detection that is based on
local image histogram entropy is proposed. The algorithm
maintains a real-time reliable performance while meeting the
AUV low power budget.
To read the full paper, please click here:
User review of Gidel ProcStar II and a first look at ProcSoC 3
I have used Gidel's ProcStar ASIC prototyping product since 2004, then moved
to their ProcStar II in 2006. This is my production experience with their
4 Altera Stratix II 180 FPGA hardware, and its associated development kit.
In addition, I've added my evaluation of Gidel's newest ProcSoC 3, which has
between 2-6 Stratix IV 820 FPGAs (we evaluated the version with 2 FPGAs).
We have a 4 M+ gate design to fit into ProcSTAR and 8 M+ design we will use
for ProcSoc. The designs will run from 100 to 600 MHz.
To read the full paper, please click here:
Bridging Parallel and Reconfigurable
Computing with
Multilevel PGAS and SHMEM+
Reconfigurable computing (RC) systems based on FPGAs are
becoming an increasingly attractive solution to building parallel
systems of the future. Applications targeting such systems have
demonstrated superior performance and reduced energy
consumption versus their traditional counterparts based on
microprocessors. However, most of such work has been limited to
small system sizes. Unlike traditional HPC systems, lack of
integrated, system-wide, parallel-programming models and
languages presents a significant design challenge for creating
applications targeting scalable, reconfigurable HPC systems. In
this paper, we introduce and investigate a novel programming
model based on Partitioned Global Address Space (PGAS), which
simplifies development of parallel applications for such systems.
To read the full paper, please click here:
FPGA based accelerator breaks new ground in Bioinformatics
PROBLEM: The act of aligning DNA, RNA, or protein sequences in order to identify regions of similarity is one of the most common and important tasks in computational biology.
CETIC Research Project in the field of microelectronics entitled as CANAPE .
This paper presents the results of the CANAPE project. The main objectives of this project are to study and to develop a prototype of FPGA based Hardware accelerator.
A 13 Weeks project in just 1 week, for less than $2K.
The goal of the project was to build a custom piece test equipment (system simulator) for a complex system.
The PROC board allows you to concentrate your efforts on the the application since the PCI interface,targeted application driver,
SDRAM interface board design for the latest Altera FPGAs and prototyping PCB(Printed Circuit Board) have been already implemented for you.
Record-Breaking - 1 day with PROC Board does more than 256 PCs * 14 days.
Application: Real-Time Simulator for a Powerful Error-Correcting Code.
This paper was published by Tong Zhang from Rensselaer Polytechnic Institute, Troy, NY.
The fastest real time LDPC decoding simulator ever reported in the open resource based on ProcSuperStar
Platform, was successfully implemented by researchers at Rensselaer VLSI System
Architecture Laboratory. The real-time LDPC decoding simulator achieved an average decoding throughput of 1Gbps with
frame error rate of less than 10-6 by simulating 8x1013 bits within only one day.
Compared to Researchers at Japan AIST which employed Super Cluster of 256 3GHz Processors to simulate 2x1013 bits in two weeks.
Hardware-accelerated protein identification for mass spectrometry.
An ongoing issue in mass spectrometry is the time it takes to search DNA
sequences with MS/MS peptide fragments (see, e.g., Choudary et al.,
Proteomics 2001; 1: 651-667.) Search times are far longer than spectra
acquisition time, and parallelization of search software on clusters requires
doubling the size of a conventional computing cluster to cut the search time in
half. Field programmable gate arrays (FPGAs) are used to create
hardware-accelerated algorithms that reduce operating costs and improve search
speed compared to large clusters.
A Dynamically Reconfigurable Processor for Dataflow Graph Execution.
This paper was presented by Lorenzo Verdoscia from the Research Center on Parallel Computing and Supercomputing - CNR.
The project was implemented using PROC Board hardware.
|