BlastFunction: A Full-stack Framework Bringing FPGA Hardware Acceleration to Cloud-native Applications

“Cloud-native” is the umbrella adjective describing the standard approach for developing applications that exploit cloud infrastructures’ scalability and elasticity at their best. As the application complexity and user-bases grow, designing for performance becomes a first-class engineering concern. As an answer to these needs, heterogeneous computing platforms gained widespread attention as powerful tools to continue meeting SLAs for compute-intensive cloud-native workloads. We propose BlastFunction, an FPGA-as-a-Service full-stack framework to ease FPGAs’ adoption for cloud-native workloads, integrating with the vast spectrum of fundamental cloud models. At the IaaS level, BlastFunction time-shares FPGA-based accelerators to provide multi-tenant access to accelerated resources without any code rewriting. At the PaaS level, BlastFunction accelerates functionalities leveraging the serverless model and scales functions proactively, depending on the workload’s performance. Further lowering the FPGAs’ adoption barrier, an accelerators’ registry hosts accelerated functions ready to be used within cloud-native applications, bringing the simplicity of a SaaS-like approach to the developers. After an extensive experimental campaign against state-of-the-art cloud scenarios, we show how BlastFunction leads to higher performance metrics (utilization and throughput) against native execution, with minimal latency and overhead differences. Moreover, the scaling scheme we propose outperforms the main serverless autoscaling algorithms in workload performance and scaling operation amount.

Cloud Computing Models for Business

The idea of cloud computing is not a new one, it has been developed and discussed for many years. Cloud computing is a model which allows to get access to the network upon request from the set of adjustable computing services, such as infrastructure, applications and storages. Cloud services and data storage products allow their users to store and share any type of document and file from any device connected to Internet. There are several types of cloud services, which can be subdivided into: SaaS (Software as a Service), PaaS (Platform as a Service), IaaS (Infrastructure as a Service). Besides, there are several deployment models, such as public, residential, hybrid or community cloud. Cloud computing models are based on modern process paradigm, which offers new alternatives to the companies of various ranges for implementation of innovative business models. With the help of these new business models small companies will be able to use cloud computing platforms and to increase gradually their computation capacities and data storage capacities depending on the requirements in real time mode, which creates a unique opportunity for market competition. Keywords— cloud computing, IaaS, OpenStack, PaaS, SaaS.

AuraRing

Wearable computing platforms, such as smartwatches and head-mounted mixed reality displays, demand new input devices for high-fidelity interaction. We present AuraRing, a wearable magnetic tracking system designed for tracking fine-grained finger movement. The hardware consists of a ring with an embedded electromagnetic transmitter coil and a wristband with multiple sensor coils. By measuring the magnetic fields at different points around the wrist, AuraRing estimates the five degree-of-freedom pose of the ring. AuraRing is trained only on simulated data and requires no runtime supervised training, ensuring user and session independence. It has a resolution of 0.1 mm and a dynamic accuracy of 4.4 mm, as measured through a user evaluation with optical ground truth. The ring is completely self-contained and consumes just 2.3 mW of power.

Action: A New Metric for Evaluating the Energy Efficiency on High Performance Computing Platforms (ranked on Green500 List)

The need for new and more reliable metrics is always in demand. In this paper, a new metric is proposed for the evaluation of high performance computing platforms in conjunction with their energy consumption. The aim of the new metric is to reliably compare different HPC systems concerning their energy efficiency. The metric provides a mean to rank supercomputers of similar capabilities, avoiding the misleading results of metrics like performance-per-watt, currently used for ranking systems, as in the Green500 list, where systems with totally different sizes and capabilities are ranked consecutively. An example of this misuse for two adjacent systems in the Green500 list, is discussed. A comparative study for the energy efficiency of three high performance computing platforms, with different architectures, using the proposed metric is presented.

A Data-Driven Identification Procedure for HVAC Processes with Laboratory and Real-World Validation

Linear system identification is a well-known methodology for building mathematical models of dynamic systems from observed input–output data. It also represents an essential tool for model-based control design, adaptive control and other advanced control techniques. Use of linear identification is, however, often limited to academic environment and to research facilities equipped with scientific computing platforms and highly qualified staff. Common industrial or building control system technology rarely uses these advanced design techniques. The main obstacle is typically lack of experience with their practical implementation. In this article, a procedure is proposed, implemented, and tested, that brings the benefits of linear identification into broader control system practice. The open-source DCU control system platform with its advanced control framework is used for implementation of the proposed linear identification procedure. The procedure is experimentally tested in the laboratory setting using a unique model of HVAC system as well as in real-world environment in an experimental two storey family house. Testing this novel feature of the control system has proved satisfactory results, while some of them are presented in graphical and numerical form.

GuPPy, a Python toolbox for the analysis of fiber photometry data

Fiber photometry (FP) is an adaptable method for recording in vivo neural activity in freely behaving animals. It has become a popular tool in neuroscience due to its ease of use, low cost, the ability to combine FP with freely moving behavior, among other advantages. However, analysis of FP data can be challenging for new users, especially those with a limited programming background. Here, we present Guided Photometry Analysis in Python (GuPPy), a free and open-source FP analysis tool. GuPPy is designed to operate across computing platforms and can accept data from a variety of FP data acquisition systems. The program presents users with a set of graphic user interfaces (GUIs) to load data and provide input parameters. Graphs are produced that can be easily exported for integration into scientific figures. As an open-source tool, GuPPy can be modified by users with knowledge of Python to fit their specific needs.