campusreview. com. au
ON CAMPUS
Doormats: mandating insecurity by requiring government access to all data and communications was co-authored by Creative Commons co-founder professor Harold Abelson, Cambridge security engineering professor Ross Anderson, and Microsoft Research senior cryptographer Josh Benaloh – along with 12 others. It outlines the security risks mandatory backdoors could cause. Dr Suelette Dreyfus, a University of Melbourne cybersecurity expert, has previously warned that backdoors could compromise university research worldwide.
Underwood commented that mandatory backdoors are an area“ outside my realm of expertise” and added,“ Dell is entirely focused on … making sure our customers’ data is secure and that those threats are something they don’ t have to worry about.”
Here, Underwood discusses with Campus Review how universities are already using supercomputers and how their potential is only beginning to be fully understood.
CR: Can you tell us about this partnership between Dell and Monash and the supercomputer there?
AU: It’ s been a multi-year partnership between Dell and Monash University. Over the last few years, Monash University and Dell have partnered to build a new form of high-performance computing systems. These systems have been designed jointly to be far more flexible, more cloud-like and more extensible for end researchers, with the focus on providing researchers with immediate access to data, and more capability for collaboration between Monash researchers and those from other universities and global research agencies.
More recently, we’ ve also announced the latest supercomputing system as part of this initiative, which is called M3 – also known as MASSIVE-3. MASSIVE-3 is built using the latest generation Dell HPC Systems and a unique type of technology called GPGPU, or General Purpose Graphical Processing Unit, in order to accelerate scientific visualisation and data science.
MASSIVE-3 was installed at Monash in February. What’ s it been up to since then? It’ s been doing some great science. I’ m excited that it’ s been deployed so quickly as well. It’ s servicing [ staff ] from the School of Engineering, School of Medical Science, and then the School of Maths and Stats as well. A broad range of sciences have been using MASSIVE-3. We’ re also seeing Monash doing some groundbreaking science around artificial intelligence and cognitive computing [ and ] taking advantage of these new GPUs as part of MASSIVE.
What future do you see supercomputers having in research? There’ s already an explosion of data taking place in the digital universe right now. In 2015, there were about 4.4 zettabytes worth of data in the digital world. We know this will grow to about 44 zettabytes by 2020.
Could you explain what a zettabyte is? A zettabyte would be roughly a billion times the amount of data a traditional personal laptop could hold. Just think about how much data supercomputers can sift through in order to help with understanding it and making predictions.
The medical faculty has been quite enthusiastic about MASSIVE-3. Why is that? With predictive analytics being able to sift through all of this data, the medical faculties are able to work on new ways and new types of medicines. By crunching all of this data together to understand the ways new drugs will be able to dock and bind in the human biology, but also by sifting through all encrypted patient records that have been deidentified in a secure manner, in order to start to understand where doctors and nurses can improve patient healthcare based on previous experiences with that patient, or hundreds of thousands of other patients who may have had similar symptoms.
Regarding universities’ research, is purchasing a supercomputer – for universities that can afford it – a trend that Dell has observed? Definitely a growing trend. It’ s something Dell’ s focused on. We’ re seeing that the evolving research and technical workloads in universities are pushing the limits of the current technology platforms. We know that these universities – in order to attract next generation researchers and compete against other universities on a global scale – need to do things faster with less risk. They also need to increase the amount of data they can store and ingest and analyse. They also need to make sure researchers can access these new supercomputers seamlessly, easily and securely, in order to take advantage of all the data, be it in fields of engineering or medical research or even in digital research and the arts sector.
We know that these science and research-focused universities are pushing and driving to have more processing power that will enable staff to accelerate their research output on a global scale. We’ ve recently announced, in partnership with the University of Sydney, that its Dell supercomputer, which is called Artemis, will be tripling in size within the next six to eight weeks.
So they’ ll be able to triple the amount of processing power their researchers on campus can access. That should help them drive their competitiveness on a global scale as well.
How do you keep these big supercomputers secure? We don’ t just look at the actual hardware of the supercomputer, we look at the software layer as well. That includes the operating system as well as the end-user tools, the file system the data is stored on, and then all of the security around that. We have a strong security division at Dell that we work with to make sure customers’ data, once it’ s sitting on these supercomputers, is locked up tight. Monash is a case study. You might have read that they worked closely with ANZ Bank. We gave Monash technology that allows for on-demand encryption of data, to make sure any data from ANZ Bank that’ s on site at Monash is entirely secured and encrypted, so hackers aren’ t able to take advantage of that data.
This highly valuable data surely draws hackers to these computers, right? It definitely does. Look, we do know that scientific data is valuable, both to hackers and to government agencies... Therefore, protecting the data is always a core focus of ours. For our high-performance computing, we have a saying:“ Data first.” By putting the data first, we make sure that we’ re focusing on enabling the data, securing the data and accessing the data.
We’ re focused on helping solve these big science challenges, at any scale, with agility and efficiency. We’ re focused on addressing all aspects of high-performance computing and making sure our clients can purchase the [ supercomputers ] with confidence – when, where, and how they would like to do so. ■
21