Search

Summary The use of Graphics Processing Units (GPUs) for scientific computing has become mainstream in the last decade. Applications ranging from deep learning to seismic modelling have benefitted from the increase in computational efficiency compared to their equivalent CPU-based implementations. Since many inverse problems in geophysics relies on similar core computations – e.g. dense linear algebra operations, convolutions, FFTs – it is reasonable to expect similar performance gains if GPUs are also leveraged in this context. In this paper we discuss how we have been able to take PyLops, a Python library for matrix-free linear algebra and optimization originally developed for singe-node CPUs, and create a fully compatible GPU backend with the help of CuPy and cuSignal. A benchmark suite of our core operators shows that an average 65x speed-up in computations can be achieved when running computations on a V100 GPU. Moreover, by careful modification of the inner working of the library, end users can obtain such a performance gain at virtually no cost: minimal code changes are required when switching between the CPU and GPU backends, mostly consisting of moving the data vector to the GPU device prior to solving an inverse problem with one of PyLops’ solvers.

The MPEG-G standardization initiative is a coordinated international effort to specify a compressed data format that enables large scale genomic data to be processed, transported and shared. The standard consists of a set of specifications (i.e., a book) describing: i) a nor-mative format syntax, and ii) a normative decoding process to retrieve the information coded in a compliant file or bitstream. Such decoding process enables the use of leading-edge com-pression technologies that have exhibited significant compression gains over currently used formats for storage of unaligned and aligned sequencing reads. Additionally, the standard provides a wealth of much needed functionality, such as selective access, data aggregation, ap-plication programming interfaces to the compressed data, standard interfaces to support data protection mechanisms, support for streaming and a procedure to assess the conformance of implementations. ISO/IEC is engaged in supporting the maintenance and availability of the standard specification, which guarantees the perenniality of applications using MPEG-G. Fi-nally, the standard ensures interoperability and integration with existing genomic information processing pipelines by providing support for conversion from the FASTQ/SAM/BAM file formats.In this paper we provide an overview of the MPEG-G specification, with particular focus on the main advantages and novel functionality it offers. As the standard only specifies the decoding process, encoding performance, both in terms of speed and compression ratio, can vary depending on specific encoder implementations, and will likely improve during the lifetime of MPEG-G. Hence, the performance statistics provided here are only indicative baseline examples of the technologies included in the standard.

An appropriate complexity analysis stage is the first and fundamental step for any methodology aiming at the implementation of today's (complex) multimedia algorithms. Such a stage may have different final implementation goals such as defining a new architecture dedicated to the specific multimedia standard under study, or defining an optimal instruction set for a selected processor architecture, or to guide the software optimization process in terms of control-flow and data-flow optimization targeting a specific architecture. The complexity of nowadays multimedia standards, in terms of number of lines of codes and cross-relations among processing algorithms that are activated by specific input signals, goes far beyond what the designer can reasonably grasp from the "pencil and paper" analysis of the (software) specifications. Moreover, depending on the implementation goal different measures and metrics are required at different steps of the implementation methodology or design flow. The process of extracting the desired measures needs to be supported by appropriate automatic tools, since code rewriting, at each design stage, may result resource consuming and error prone. This paper reviews the state of the art of complexity analysis methodologies oriented to the design of multimedia systems and presents an integrated tool for automatic analysis capable of producing complexity results based on rich and customizable metrics. The tool is based on a C virtual machine that allows extracting from any C program execution the operations and data-flow information, according to the defined metrics. The tool capabilities include the simulation of virtual memory architectures. This paper shows some examples of complexity analysis results that can be yielded with the tool and presents how the tools can be used at different stages of implementation methodologies.

Mesenchymal stem cells (MSCs) have been found to be useful in the management of different models of neurological diseases. In the present study, we tested the possible protective effect of MSCs on sensory dorsal root ganglia (DRG) explants exposed to the toxic effect of CDDP, a widely used anticancer drug. DRG explants cultured on a collagen layer and exposed to NGF for 2 h (differentiating neurons) or for 5 days (fully differentiated neurons) were treated with CDDP and subsequently co-cultured with MSCs. MSCs were able to support the survival of both differentiating and fully differentiated DRG neurons up to 2 months after the drug treatment, reducing the CDDP-induced death of DRG neurons. MSCs were, however, unable to restore the correct length of DRG neurites compromised by CDDP treatment. The positive effect on neuronal survival was exerted through the contact between DRG and MSCs, and not mediated by neurotrophic factors released by the MSCs. Our observations could represent a starting point for designing a neuroprotective strategy to limit CDDP induced neuropathy without interfering with its anticancer properties.

Background Several studies report increased minor neurodevelopment dysfunctions in children born very low birth weight (VLBW). Usefulness of preventive early intervention programs to improve neurodevelopment outcome of VLBW infants is still under investigation. Aims To evaluate the effects of an early post-discharge developmental mother–child intervention program on neurodevelopment outcome at 36 months in VLBW infants. Study design Prospective study. Subjects 36 VLBW infants ([mean (S.D.)] birthweight = 864 g (204 g); gestational age = 27.9 weeks (2.4 weeks)), consecutively born January–August 2001, randomized in intervention and control groups. Outcome measures Neurodevelopment assessment at 36 months of chronological age with use of the Griffiths Mental Development Scale and related subscales. Results At 36 months of chronological age, as compared to controls, children in intervention group exhibited higher scores in personal–social subscales ([mean (S.D.)] = 101.4 (9.3) vs. 92.9 (12.1), P = 0.02), eye–hand coordination (92.7 (4.5) vs. 87.1 (9.9), P = 0.041), practical reasoning (98.6 (8.2) vs. 89.4 (10.1), P = 0.01). Development Scale were 97.6 (5.5) and 92.4 (9.9), respectively, in intervention and control groups (P = 0.074). Conclusions Early post-discharge developmental mother–child intervention program may have a positive effect on later neurodevelopment outcome of VLBW children.

The complexity of nowadays, algorithms in terms of number of lines of codes and cross-relations among processing algorithms that are activated by specific input signals, goes far beyond what the designer can reasonably grasp from the "pencil and paper" analysis of the (software) specifications. Moreover, depending on the implementation goal different measures and metrics are required at different steps of the implementation methodology or design flow of SoC. The process of extracting the desired measures needs to be supported by appropriate automatic tools, since code rewriting, at each design stage, may result resource consuming and error prone. This paper presents an integrated tool for automatic analysis capable of producing complexity results based on rich and customizable metrics. The tool is based on a C virtual machine that allows extracting from any C program execution the operations and data-flow information, according to the defined metrics. The tool capabilities include the simulation of virtual memory architectures.

This paper presents some examples of architectures for the implementation of wavelet codecs. Compared to previous works, the novelty of the approach is based on a joint optimization of both the algorithmic and architectural features, according to the overall system hardware and software strategy. The results of the analysis, although general for the different considered architectures, allow specific optimization of system performance for both dedicated ASIC design and embedded software implementation based on available system resources, such as execution speed and cache performance, while minimizing power consumption.

The increasing complexity of processing algorithms has lead to the need of more and more intensive specification and validation by means of software implementations. As the complexity grows, the intuitive understanding of the specific processing needs becomes harder. Hence, the architectural implementation choices or the choices between different possible software/hardware partitioning become extremely difficult tasks. Automatic tools for complexity analysis at high abstraction level are nowadays a fundamental need. This paper describes a new automatic tool for high-level algorithmic complexity analysis, the Software Instrumentation Tool (SIT), and presents the results concerning the complexity analysis and design space exploration for the implementation of a JPEG2000 encoder using a hardware/software co-design methodology on a Xilinx Virtex-IITM platform FPGA. The analysis and design process for the implementation of a video surveillance application example is described.

The increasing complexity of processing algorithms has led to the need of more and more intensive specification and validation by means of software implementations. As the complexity grows, the intuitive understanding of the specific processing needs becomes harder and harder. Hence, the architectural implementation choices or the choices between different possible software/hardware partitioning become extremely difficult tasks. Moreover, it is also desirable to understand and measure the algorithm complexity at the highest possible level near to the algorithmic level so as to be able to take the more appropriate actions. Automatic tools to perform such analysis become nowadays a fundamental need.In this paper, the requirements of a suitable algorithmic complexity evaluation technology are discussed, with a particular emphasis to the problem of the analysis of multimedia systems and signal processing algorithms. A brief review about limitations and weaknesses of existing tools is given, specifying the characteristics of ideal "complexity evaluation systems". A new approach is described, called here Software Instrumentation Tool, SIT, yielding an automatic software tool able to extract information not depending on the simulation platform, keeping into account specific input data and resulting in a good and useful measure of the desired high-level algorithmic complexity.