NEWS HIGH-PERFORMANCE COMPUTING
Nvidia announces containerised applications for HPC and deep learning
At the SC17 conference Nvidia announced that it has begun optimising applications for HPC and visualisation of HPC workflows, in addition to partnerships with major cloud providers. Nvidia also revealed updates
LABORATORY INFORMATICS Artificially creating articles
Some 99 drafts of scientific papers have been generated so far by a manuscript writer launched three weeks ago, according to the electronic lab notebook company sciNote. The Manuscript Writer is described as
the first of its kind, allowing researchers to generate a draft of a scientific manuscript using data stored by the user on its software, and references that are accessible in open access journals. The device was developed by a team
of scientists and experts on machine learning and software development at sciNote (creators of sciNote, a free open source electronic lab notebook) and it has the ability to simplify the process of preparing scientific manuscripts by using the technological advances in artificial intelligence. Manuscript Writer is aimed at reducing the time needed to prepare initial content. It gathers the data scientists organise and
save in sciNote during their research, and presents it in the form of a manuscript draft. This allows scientists to save time on gathering their relevant data so they don’t need to start writing from scratch. Once they receive the draft, they can start editing and improving the text. SciNote says that some 54 per cent of all studies cannot be confirmed, because of a lack of data management. ‘There is no standardised approach and the amount of data doubles every three years,’ said Klemen Zupancic, PhD in biomedicine and CEO of sciNote. In this sense, electronic lab notebooks can help to reduce this rate. Once the data is systematically organised and traceable, Manuscript Writer can help out and save time needed to start writing the manuscripts. You can read more details about the
release of the manuscript writer addition to sciNote in a previous article, ‘sciNote adds AI capabilities’.
to its Nvidia GPU Cloud (NGC), which will provide containerised HPC, deep learning and HPC visualisation applications. The firm is targeting many
top HPC applications, including RELION a software package used to discover gravitational waves – a discovery that won the Nobel Prize for physics in 2017. During the first day of the
conference, Nvidia announced the top 15 applications and 70 per cent of top 50 HPC applications are now optimised for GPU acceleration. This data comes from an Intersect360 report which described Nvida as ‘critical to the future of scientific computing’. ‘GPU computing has reached a tipping point in the HPC market that will encourage continued increases in application optimisation,’ wrote Addison Snell and Laura Segervall, of Intersect360. In a pre-show talk, Nvidia CEO
and founder Jensen Huang noted every major computer maker and cloud service has turned to Nvidia Volta architecture to accelerate data-intensive workloads. Nvidia has begun to
development driven programs. IPVM looks at where the packaging industry is headed and provides insight into the ways that packaging connects to broader trends and drivers impacting business. Altair intends to apply the HyperWorks
suite to problems concerning technology, design, research and development and automation for the packaging industry. The recent acquisition of Carriots brings a new level of understanding to the Internet of Things (IoT), which the companies will also apply to the project. It is hoped that this will create a joint process that will
www.scientific-computing.com | @scwmagazine
contribute to business growth for both firms. ‘We see an opportunity to deliver a
joint-offering with Altair through our consulting programs, which inevitably expose customers to the HyperWorks platform. Altair can work with our clients to train them on the software, or take on the package development project entirely, to help us all achieve a common goal,’ said Brian Wagner, co-founder and principle at PTIS. ‘We believe that leveraging these capabilities will deliver a greater range of options, efficiency and speed to market.’
containerise HPC and HPC visualisation applications to enable more users to take advantage of GPU acceleration. Nvidia has released software and tools as part of the NGC container registry that allows scientists to deploy applications and HPC visualisation tools efficiently using cloud-based GPU. ‘One of the biggest market dynamics is the advent of AI,’ noted Intersect360, in the report. ‘Many organisations are looking to deep learning techniques to bring AI advancements to their products, services, or operations. These algorithms often rely on GPUs, to the extent AI has become a major growth driver for Nvidia.’
December 2017/January 2018 Scientific Computing World
25
PPhonlamai Photo/
Shutterstock.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28
