We Provide the Cloud. You Change the World.

Recent Posts

Transforming Research

Realizing a Vision: Accelerated Cancer Research

In our current world, a good deal of energy in the health sciences has shifted toward understanding SARS-CoV-2, as well as treating and finding a vaccine against the disease it causes, COVID-19. Oracle for Research has been proud to collaborate with researchers working on these efforts. Yet, not all of our energy has been focused on COVID-19, as the same challenges that existed before the pandemic persist.  Over the past year, I have had the honor of working with researchers at the Lawrence J. Ellison Institute for Transformative Medicine at USC, who are part of an exciting experiment that envisions both the opportunity and the space to take a multidisciplinary approach in exploring and treating cancer.  Founded by Dr. David Agus, and supported by Oracle Chairman and Chief Technology Officer Larry Ellison, the vision is that interdisciplinary teams armed with modern tools, modern technology and modern solutions can redefine cancer care. The vision is to transform lives and enhance health. A major milestone on the road to realizing this vision was recently achieved. Nature published a groundbreaking study by Ellison Institute researchers that used Artificial Intelligence (AI) and machine learning to develop a proof of principle that has the potential to transform breast cancer diagnosis and treatment. The research team included a medical doctor, a biomedical engineer who is also a computer scientist, a pathologist, a theoretical physicist who specializes in signalizing dynamics in cancer cells, and a professor of medicine and practicing physician.  The project was funded in part by a grant from the Breast Cancer Research Foundation; the data came from tissue samples housed in different databases; and the analytical work and machine learning algorithms were developed using Oracle Cloud. Oracle technical advisors lent their expertise to optimize the use of Oracle Cloud technologies. In brief, the research team found evidence that with deep learning, tissue morphology can be used to successfully identify histologic H&E features that predict the clinical subtypes of breast cancer, potentially providing a viable and cost effective alternative to more expensive molecular screening. If their approach – using “tissue fingerprints” identified by deep learning to classify breast cancers – continues to prove successful, they will essentially democratize the diagnosis and treatment of breast cancer, making individualized medicine more accessible and more affordable for more women and men around the world. This, in and of itself, is rich reward for Dr. Agus’s early vision, and the Ellison Institute is just at the beginning.  We are only now getting a glimpse of what becomes possible when innovative research and cloud technology intersect and are used for good.  The Ellison Institute’s approach to research – bringing together multidisciplinary research teams, clinicians, patients, and technology – means successful outcomes potentially can happen much faster, and be found in unexpected ways. In pre-COVID days, I was fortunate to visit the Ellison Institute in person.  At the time, they were preparing to move from existing buildings at USC to their new, state-of-the-art facility, so I was only able to glimpse the reality that will come to pass when the move is complete.  I was excited to begin to understand their methodical approach of combining traditional “wet lab” research using tissue samples and microscopes with new, computational experiments.  Contrary to what I expected, the Ellison Institute isn’t just running computational simulations; instead, they are both leveraging data from “wet lab” experiments to qualitatively new kinds of results by using computational approaches and conducting novel experiments in computational labs that enable exploration beyond the limits of current “wet lab” technologies.  This isn’t just about getting to results faster – though that is certainly happening too – it’s about changing the kinds of questions scientists can ask and the ways they test hypotheses.  They are reimagining scientific experimentation. This is truly transformative work.   The Ellison Institute brings together the biologists, physicists, data scientists, researchers, doctors and patients. Oracle contributes the computational tools that enables these multidisciplinary teams to do their work faster.  I am excited about the possibilities that lie ahead. To learn more about the Ellison Institute’s revolutionary findings on breast cancer research, Oracle for Research will host the paper’s co-author Dr. Dan Ruderman in a live webinar on August 20th at 1pm PT. Registration is open to all. Attendees will have an opportunity to engage in thoughtful conversation with Dr. Ruderman and researchers worldwide.  

In our current world, a good deal of energy in the health sciences has shifted toward understanding SARS-CoV-2, as well as treating and finding a vaccine against the disease it causes, COVID-19....

Advances in Research

As Tropical Viruses Creep Northward, Visualizing a Potential New Vaccine

This article was written by Aaron Ricadela and was originally published in Forbes. As tropical diseases spread from their historical home territories into new regions including Europe and the United States, UK researchers equipped with high-performance cloud computing have designed a novel way to vaccinate against one of the most resilient to treatment. Scientists at the University of Bristol and the French National Center for Scientific Research are proposing using a lab-produced protein molecule that can act as a delivery system for future vaccines against the mosquito-borne illness chikungunya. The findings, which will be published in a paper this week in the journal Science Advances, show how the protein can be readily manufactured and stored for weeks at warm temperatures, making it easier to ship in regions where refrigeration is an obstacle. It’s also easy to produce in high volumes, an advantage as the disease spreads northward. “The sample is so stable you can transport it at room temperature—that’s the big deal,” says Imre Berger, a biochemistry professor at the University of Bristol and one of the paper’s authors. Designing the so-called vaccine scaffold, or delivery system, involved constructing detailed 3D images from cryogenically frozen samples scanned by an electron microscope, using high-performance cloud computing from Oracle. “You need to see what you’re actually engineering,” he says. “This is tailor-made for the cloud because every image can be processed in parallel.” The work could contribute to efforts to thwart tropical viruses that have spread beyond their usual zones. Chikungunya, whose name derives from the East African Makonde language and describes walking bent with pain, is related to dengue fever and causes high temperatures, joint pain, and exhaustion. It’s spread among humans by the bite of tiger mosquitos under the right conditions. There are no available treatments or vaccines, though the French biotech company Valneva in May reported promising results of a chikungunya vaccine trial of 120 healthy volunteers in the US. The illness belongs to a group of diseases, including Zika virus, that were previously found largely in sub-Saharan Africa, Asia, and India and have spread to the Northern Hemisphere as globalization and warmer climates push infected mosquitos far north of their normal ranges. A mysterious illness outbreak a dozen years ago in northern Italy among villagers who hadn’t traveled abroad turned out to be chikungunya. The disease spread to Florida in 2014. Zika too has broken out in the US with a rash of cases in 2016 and an infection in Texas two years ago. Deep Freeze To develop a vaccine candidate called ADDomer, scientists synthesized a protein that resembles a buckyball, or 12-sided molecule, that can carry an antigen, or substance capable of stimulating an immune response to a virus. Experiments showed ADDomer mimicked the virus’ behavior in mice and triggered immune responses, according to the Science Advances paper. The scientists’ innovation is the scaffold, which they showed can accommodate hundreds of different epitopes, the target to which an antibody binds in an immune reaction. Visualizing the proteins with help from Oracle Cloud Infrastructure was key to the molecule’s design and done at a fraction of what it would have cost to use an on-premises supercomputer cluster, according to Frederic Garzoni, a co-author of the paper and co-founder of Imophoron, a startup founded to commercialize the scientists’ approach. The ADDomer scientists studied the molecular structure of the synthetic protein using computer-generated images stitched together from exposures made by the University of Bristol’s cryo-electron microscope. The apparatus rapidly freezes samples with liquid nitrogen to nearly 200 degrees below zero Celsius to yield two-dimensional pictures. These can be constructed using special software into 3D images at nearly atomic resolution. “How the protein works is very strongly coupled to its 3D shape,” says Christopher Woods, a research software engineering fellow at the University of Bristol, who was involved in the computational work. By spending just 217 British pounds ($270)  on Oracle CPU and GPU power delivered as a cloud service, the team processed a large number of cryo-electron microscope images to generate a single 3D structure. “If you only need to generate an image occasionally it’s obviously much cheaper to do it in the cloud,” he says. Public cloud computing is increasingly augmenting or replacing traditional supercomputers in molecular biology, physics, and other scientific fields. Discover what you can accomplish with Oracle for Research.

This article was written by Aaron Ricadela and was originally published in Forbes. As tropical diseases spread from their historical home territories into new regions including Europe and the...

Advances in Research

The Woolcock Institute of Medical Research Explores Insomnia's Causes Using Oracle Autonomous Database

This article was written by Lisa Morgan and was originally published in Forbes. Doctors tell patients that a healthy lifestyle requires a nutritious diet, exercise, and adequate sleep. And they can give you lots of tips on the right food and activity level. But they understand much less about the causes of insomnia, how it affects individuals, and how to help those suffering from poor sleep. The Woolcock Institute of Medical Research in Sydney is using data science to discover how treatment can be tailored to a patient's insomnia characteristics. Specifically, Woolcock researchers are studying the brain signals of sleeping patients to understand the physiology of insomnia in greater depth. Using Oracle Autonomous Data Warehouse, Woolcock researchers can build a data model in as little as an hour versus the weeks or more it used to take using shared high-performance computer resources. By using machine learning to automate many of the steps in the data science process, Woolcock researchers can dive into problem-solving sooner. "With Oracle, we don't have to focus so much on the technical part, we can focus on what's needed to sleep,” says Dr. Tancy Kao, a data scientist at the Woolcock Institute. The institute is a network of more than 200 researchers and clinicians working to improve sleep and respiratory health globally through research, clinical care, and education. The institute is working with Oracle for Research, a global program that provides scientists, researchers, and university innovators with cost-effective cloud technologies, participation in the Oracle research user community, and access to Oracle’s technical support network.   Types of Insomnia At some point in their lives, most adults will complain of trouble sleeping. Acute insomnia tends to be circumstantial, such as the lack of sleep one gets when they're nervous about giving a speech or upset about losing a job. Chronic insomnia, which tends to affect adults age 40 to 60, is the inability to sleep well for three or more nights per week for at least a month, according to Kao. One form of chronic insomnia is the "wired and tired" brain that suffers from a less-than-normal sleep duration. These individuals have trouble falling asleep or staying asleep, so their total sleep time is about three to four hours per night versus healthier individuals who sleep seven to eight hours. "Paradoxical insomniacs" sleep for the same duration as healthy patients, but their EEG slow wave activity—deep sleep—is comparatively weak. Sleep duration and sleep quality are both important, although sleep quality matters more. When a person doesn't sleep well for prolonged periods, they are at a higher risk of anxiety, depression, high blood pressure, and heart disease. One-third of Australians will experience insomnia at some point in their lives, the Woolcock estimates. "There are all these people with insomnia that aren't treated effectively, so the data science is going to let us understand the condition better and look for new treatments that are going to be targeted based on our understanding of the condition," says Christopher Gordon, associate professor at the University of Sydney, and a contributor to the Woolcock's research. Right now, the two most typical treatments are pills to temporarily relieve symptoms, and cognitive behavior therapy (CBT) which identifies and treats the root causes of insomnia. Both treatments have limitations because people tend to take sleep aids for too long and not everyone is willing to go to or stay in therapy. "It really is a 24-hour condition, not just something associated with sleep," Gordon says. Data Science Helps Provide Answers The Woolcock Institute collects a lot of data about individual patients, including a detailed questionnaire about patients' perceived sleep patterns, medical history, work environment, and home environment. They collect two weeks of data from a wearable activity monitor, observations from a sleep lab, and diary entries that track behavioral factors, such as caffeine and alcohol intake. Data science is used to connect habits to sleep changes to identify which activities and the activity duration help or hurt their sleep. “For example, is it an indoor or outdoor activity? Are they spending more time chatting with friends and family, or did they consume more drinks?" says the Woolcock Institute's Kao. "This helps us understand who can get the most benefits from the activities and who cannot." A major source of data comes from patients spending a night in the institute’s sleep lab, hooked up to a high-density electroencephalogram (EEG) device. The device records brain activity from 256 electrodes every 2 milliseconds at a sample rate of 500 hertz. The result is millions and sometimes billions of data points per patient. The Woolcock uses Oracle Autonomous Data Warehouse, running on Oracle Cloud Infrastructure for data collection, preparation, and analysis. The team can then separate the different types of data, helping researchers understand the relationships among variables. Before using Oracle Autonomous Data Warehouse, Kao had to manually clean the data and study individual variables and their relationships. Only after that could she prepare the data for analysis, build models, and do the actual analysis. If there was data missing from a variable, she would have to determine what to do next. Kao likes how the Oracle Autonomous Data Warehouse offers suggestions for doing analysis and using machine learning, and she can decide whether a suggestion is helpful. "You can follow the suggestions to clean your data, explore the data plot of all the variables, and understand what the chain looks like,” Kao says. “It also gives you simple classifications, and you can decide whether the classification is reasonable or not and whether we should use it in machine learning." Previously, the Woolcock was storing data on servers and using a high-performance computer for analysis, and the process for using it was technical and time-consuming. With that previous system, "you have to use Linux commands to assign a task like modeling or machine training. If you want to visualize the result, you have to go back to the computer," Kao says. "When you submit one modeling task, it may take two or three days to come back with the results. It depends on how heavy, how demanding your machine learning is to do that." Building an entire model used to take one or two months, assuming it proved accurate. If the model wasn't accurate, it meant starting over. Now, Kao can build a model in as little as an hour without coding or having to understand mathematical modeling. Kao does know how to do all of that—she can work in R and Python, use Linux commands, and she understands mathematical modeling well—but Oracle Autonomous Data Warehouse saves her time by automating a lot of the manual work she and Gordon had to do previously. "We have tremendous amounts of data on [each] individual patient, and we need to be able to process that data quickly," Gordon says. "I can look at it through various mechanisms, different ways of visualizing it, come up with different ideas and then we can just literally click buttons to do the machine learning, and we can explore the models that we think are going on, and it gives us the answers right away." Data Visualizations Will Get More Sophisticated So far, the Woolcock has built a 2D visualization that shows the location in the brain of each of the 256 channels and the importance of individual variables as they relate to a specific patient’s insomnia. The goal is to build 3D visualizations, so the researchers can understand the path of individual signals as they travel from one part of the brain to another. Doing so could let them not only understand what’s happening in one part of the brain, but also how it might affect other areas, and whether it’s related to a given symptom. "Oracle can show you how the brain wave changed or moved across the brain overnight,” Kao says. “It helps us not only identify the parts of that brain that play a role in insomnia but how these areas talk to each other." Discover what you can accomplish with Oracle for Research. 

This article was written by Lisa Morgan and was originally published in Forbes. Doctors tell patients that a healthy lifestyle requires a nutritious diet, exercise, and adequate sleep. And they can...

Advances in Research

Ellison Institute Uses AI to Accelerate Cancer Diagnosis and Treatment

This article was written by Margaret Lindquist and was originally published in Forbes. Nearly 150 years after the introduction of modern tissue staining to detect cancer, doctors are still diagnosing the disease much the way they did then. They still look at each biopsy sample under a microscope and hone in on suspicious areas, although new technologies are now allowing them to identify the molecular markers, like DNA mutations, that indicate whether a patient would respond best to one therapy or another. But that diagnostic odyssey is beginning to change, thanks in part to groundbreaking research conducted at the Lawrence J. Ellison Institute for Transformative Medicine of USC. There, the introduction of artificial intelligence into cancer research and treatment has the potential to revolutionize both oncology fields, just as AI is doing in transportation (self-driving cars and self-flying planes), education (chatbot advisers), finance (predictive investment models), and a range of other sectors. “I asked a pathologist, ‘If there’s one thing a computer could do for you, what would it be?’ He said, ‘I want it to tell me where to look,’” says Dr. Dan Ruderman, director of analytics and machine learning at the institute and assistant professor of research medicine at the Keck School of Medicine at the University of Southern California. “Digital pathology is how the medical field is going to catch up with Silicon Valley. Computers are showing us that they can see things we can’t, so we’re taking these techniques that have been perfected for other domains and transferring all that knowledge over to pathology.” The institute is scanning slides that show slices of biopsies and analyzing that visual data with AI algorithms trained to recognize areas of concern. After enough training, the algorithm will be able to not only recognize cancer, but even recommend a course of treatment. The use of deep learning and neural networks will make sophisticated diagnoses available even in developing countries, where there may be only one doctor in a region who has experience with diagnosing cancer. “There’s a basic diagnostic slide that is taken for every patient—but not every place has someone who can read the slide and determine the subtype of cancer and be able to recommend a specific course of drugs,” Dr. Ruderman says. “If the AI can handle that, it will not only save these hospitals money, but it should also result in better patient care.” Unique Combination The Ellison Institute is unique among research facilities, Dr. Ruderman notes, because it combines an established research group with an established clinical group, both focused on discovering new ways to diagnose and treat cancer. The Ellison Institute is building a new, wireless 5G-enabled building in West Los Angeles that will be home to a community outreach team, a health policy think tank, educational sessions—even an art gallery. Most importantly, the new facility is designed to bring together researchers and patients, making it possible to follow a patient’s progress from diagnosis through outcome. “The notion of bringing patients and researchers together is really novel,” Dr. Ruderman says. “Patients are able to tour the research labs and talk to researchers about what they're doing. Researchers will be better able to understand what patients are going through.” Facilitated through the Oracle for Research program, the Ellison Institute’s extended research team also includes Oracle computing experts and technology resources, enabling Dr. Ruderman and other scientists to conduct computational experiments alongside their laboratory research. “Oracle for Research was developed to enable researchers to use the power of cloud computing to solve some of the world’s hardest problems,” says Oracle for Research vice president Alison Derbenwick Miller. “We provide access to Oracle Cloud and to technical advisors who collaborate with researchers. At the Ellison Institute, this further expands the patient-doctor-researcher team and broadens the approach to improving patient diagnosis and treatment.” Intense Workloads Are Just the Beginning The 3D imaging at the center of the pathology research is computationally intense and requires massive amounts of storage. For example, the data for just 1,000 patients, with 100 images per patient at 10 gigabytes per image, requires 1 petabyte—or a million gigabytes—of storage. Patient data comes not only from the patients at the clinic, but also from public sources such as The Cancer Genome Atlas and the Australian Breast Cancer Tissue Bank. To handle its intensive, scalable computing needs, the institute is using Oracle Autonomous Database, Oracle Cloud Infrastructure bare metal compute instances, and Oracle Cloud storage, as well as Oracle Cloud Infrastructure FastConnect for network connectivity. “We've been building neural networks using Oracle Cloud, and now that we are working in three dimensions, we have much, much more data,” Dr. Ruderman says. “It's going to be taxing the system a lot more and requiring a lot more computation.” “Using Oracle’s autonomous capabilities eliminates much of the manual IT labor that’s required for standard databases,” says Muhammad Suhail, Oracle Cloud Infrastructure principal product manager. “Dr. Ruderman’s team was using four CPUs, and they wanted to go to eight,” Suhail says. “All we did was tell them where to find the button. They clicked it, and now the data warehouse is running with eight CPUs without needing any downtime.” Dr. Ruderman says the institute stopped worrying about scaling issues once it started using Oracle Autonomous Database. “Now I can think more about science and less about technology,” he says. And there is more to come. The most revolutionary aspect of the institute’s work is still on the horizon: using AI not just for diagnosis, but also for experimentation—whereby, in effect, scientists are running experiments within the computer. For example, Ruderman says that the plan is to use artificial intelligence to interpret complex 3D patterns and determine whether those images provide more information than can be derived from 2D views. “We want to answer a very basic question—whether we can learn a lot more about a patient by looking at this added dimension in pathology,” says Dr. Ruderman. Dr. Ruderman is particularly excited about applying the institute’s research to better the lives of clinic patients. “We want to understand everything about the patient pipeline. Rather than focusing on just the piece that we think is important, we focus on all the pieces,” he says. “It gives us perspective on the patient's journey and how we can improve it. I hope that everybody on my team would be able to answer this question: ‘How does your work impact a patient's care?’” This unique vision of the future of patient care and research is a crucial benefit that Derbenwick Miller emphasizes. “We want to bring together patients, clinicians, researchers, and cloud computing into a single setting. The Ellison Institute’s approach seeks to rapidly advance and improve the diagnosis and treatment of cancer, and this should ultimately result in a better quality of life and prognosis for patients,” says Derbenwick Miller. Discover what you can accomplish with Oracle for Research. 

This article was written by Margaret Lindquist and was originally published in Forbes. Nearly 150 years after the introduction of modern tissue staining to detect cancer, doctors are still diagnosing...

Advances in Research

Research, Oracle Cloud and AI converge to help reduce the risk of diabetic amputations

Losing a limb is life changing…and expensive The loss of a limb can be devastating to a person’s life.  As well as the impact to mobility, independence and participation in day-to-day activities, it can also have a significant impact on a person’s relationships, community and social life.  To add to this, an amputation can radically change how a person views themselves and their future. Amputees often have to cope with ongoing health issues (e.g., pain), learn new skills, and adjust their expectations about their capabilities. Not only is an amputation a major life changing event for a person; the burden and cost of ulceration on the UK NHS is over £5bn pounds a year.  One in four people with diabetes will develop a foot ulcer in their lifetime. Of those, about a quarter will require a lower limb amputation as a life-saving procedure. Astoundingly, though, experts believe that with more proactive care up to half of all amputations could be avoided.  A significant challenge for at-risk individuals is accessing effective care early and having the information and tools to self-care thereafter.  To avoid eventual amputation, leg and foot ulcers and associated problems need to be treated quickly and correctly to reduce the risk of non-healing wounds, secondary health problems and deteriorating health1.  Researchers at Manchester Met have made a breakthrough Imagine what would change if AI could offer early detection of ulcers, and proactively refer patients for care.  For example, if AI could assist a patient, their carer, or a relatively low-skilled clinician, to identify early and monitor the progression of a foot or leg ulcer? Not only could the patient avoid an amputation; such a solution would also deliver significant time and cost-savings for health services. A clinical tool that is simple to use, widely accessible, and scientifically robust could relieve clinical burden and provide a paradigm-shift for diabetic footcare.  A team of researchers at Manchester Metropolitan University, co-led by Prof. Neil Reeves and Dr. Moi Hoon Yap, have been working on a solution to achieve just that.  Enabled by Oracle’s high performance computing, they have developed AI algorithms that use computer vision technology to identify a foot ulcer at various stages of its development.  Based on robust lab testing, the application, called FootSnap AI, can automatically identify diabetic foot ulcers and associated pathologies using deep learning.   Developed using thousands of diabetic foot images and subjected to extensive scientific peer review and published in a number of medical and computer vision journals234,  in lab trials, FootSnap AI has a high sensitivity (0.934) and specificity (0.911) in identifying diabetic foot ulcers from foot images.  Testing in a real world setting The NHS Manchester Amputation Reduction Strategy (The MARS Project) , which Oracle has also been supporting, will be shortly commencing a programme to test the efficacy of the technology in a real world setting.  The original performance of the standalone mobile app was constrained by its hardware capability and could only support a light-weight AI/deep learning model. Built on AI technology developed by Manchester Met and now using Oracle cloud infrastructure, FootSnap AI is scalable and can respond to new demands rapidly. The cloud infrastructure equipped with GPU can speed up the inference time and provide better accuracy in ulcer detection. “Understanding the treatment of ulceration and whether these wounds are getting better or worse is essentially pattern recognition.  Further, the real breakthrough will come if we - health professionals and patients - can identify these wounds much earlier and therefore initiate much more timely treatment.  This is where artificial intelligence is potentially a game changer,” says Naseer Ahmad, Consultant Vascular Surgeon at Manchester University NHS Foundation Trust. “Oracle Cloud has provided the framework for our AI architecture to be deployed to the cloud as a service to mobile clients. Oracle Cloud delivers an online enterprise scale solution where our data can be stored, processed, and monitored seamlessly using state of the art web technologies.​,” says Bill Cassidy, Research Associate at Manchester Metropolitan University, in Manchester, UK.  References: 1An NHS England study estimates that having effective care early, reduces leg ulcer healing times from approximately two years to just a few months and is 10 times less expensive. But many patients suffer unnecessarily for several years due to a lack of knowledge and not accessing the right care.  NHS England (2017). NHS RightCare scenario: The variation between sub-optimal and optimal pathways. 2 Goyal, M., Reeves, N., Rajbhandari, S., & Yap, M. H. (2019). Robust Methods for Real-Time Diabetic Foot Ulcer Detection and Localization on Mobile Devices. IEEE Journal of Biomedical and Health Informatics. 23(4), 1730-1741, doi:10.1109/JBHI.2018.2868656 3Yap, M. H., Chatwin, K. E., Ng, C. C., Abbott, C. A., Bowling, F. L., Rajbhandari, S., . . . Reeves, N. D. (2018). A New Mobile Application for Standardizing Diabetic Foot Images. Journal of Diabetes Science and Technology, 12(1), 169-173. doi:10.1177/1932296817713761 4Goyal, M., Reeves, N. D., Davison, A. K., Rajbhandari, S., Spragg, J., & Yap, M. H. (2018). DFUNet: Convolutional Neural Networks for Diabetic Foot Ulcer Classification. IEEE Transactions on Emerging Topics in Computational Intelligence. doi: 10.1109/TETCI.2018.2866254

Losing a limb is life changing…and expensive The loss of a limb can be devastating to a person’s life.  As well as the impact to mobility, independence and participation in day-to-day activities, it...

Research Computing

What is HPC in the Cloud? Exploring the Need for Speed

Welcome to Oracle Cloud Infrastructure Innovators, a series of occasional articles featuring advice, insights, and fresh ideas from IT industry experts and Oracle cloud thought leaders. High Performance Computing (HPC) refers to the practice of aggregating computing power in a way that delivers much higher horsepower than traditional computers and servers. HPC is used to solve complex, performance-intensive problems—and organizations are increasingly moving HPC workloads to the cloud. HPC in the cloud is changing the economics of product development and research because it requires fewer prototypes, accelerates testing, and decreases time to market. I recently sat down with Karan Batta, who manages HPC for Oracle Cloud Infrastructure, to discuss how HPC in the cloud is changing the way that organizations new and old, develop products and conduct cutting-edge scientific research. We talked about varying topics including the key differences between legacy, on-premises HPC workloads, and newer HPC workloads that were born in the cloud. Listen to our conversation here and read a condensed version below: Your browser does not support the audio player Let's start with a basic definition. What is HPC and why is everyone talking about it? Karan Batta: HPC stands for High Performance Computing—and people tend to bucket a lot of stuff into the HPC category. For example, artificial intelligence (AI) and machine learning (ML) is a bucket of HPC. And if you're doing anything beyond building a website—anything that is dynamic—it's generally going to be high performance. From a traditional perspective, HPC is very research-oriented, or scientifically-oriented. It's also focused on product development. For example, think about engineers at a big automotive company making a new car. The likelihood is that the engineers will bucket all of that development—all of the crash testing analysis, all of that modeling of that car—into what's now called HPC. The reason the term HPC exists is because it's very specialized. You may need special networking gear, special compute gear, and high-performance storage, whereas less dynamic business and IT applications may not require that stuff. Why should people care about HPC in the cloud? Batta: People and businesses should care because it really is all about product development. It's about the value that manufacturers and other businesses provide to their customers. Many businesses now care about it because they've moved some of their IT into the cloud. And now they're actually moving stuff into the cloud that is more mission-critical for them—things like product development. For example, building a truck, building a car, building the next generation of DNA sequencing for cancer research, and things like that. Legacy HPC workloads include things like risk analysis modeling and Monte Carlo simulation, and now there are newer kinds of HPC workloads like AI and deep learning. When it comes to doing actual computing, are they all the same or are these older and newer workloads significantly different? Batta: At the end of the day, they all use computers and servers and network and storage. The concepts from legacy workloads have been transitioned into some of these modern cloud-native type workloads like AI and ML. Now, what this really means is that some of these performance-sensitive workloads like AI and deep learning were born in the cloud when cloud was already taking off. It just so happened that they could use legacy HPC primitives and performance to help accelerate those workloads. And then people started saying, "Okay, then why can't I move my legacy HPC workloads into the cloud, too?" So, at the end of these workloads all use the same stuff. But I think that how they were born and how they made their way to the cloud is different. What percentage of new HPC workloads coming into the cloud are legacy, and what percentage are newer workloads like AI and deep learning? Which type is easier to move to the cloud? Batta:  Most of the newer workloads like AI, ML, containers, and serverless were born in the cloud so there already ecosystems available to support them in the cloud. Rather than look at it percentage-wise, I would suggest thinking about it in terms of opportunity. Most HPC workloads that are in the cloud are in the research and product development phase. Cutting-edge startups are already doing that. But the big opportunity is going to be in legacy HPC workloads moving into the cloud. I'm talking about really big workloads—think about Pfizer, GE and all these big monolithic companies that are running production workloads of HPC on their on-premises clusters. These things have been running 30 or 40 years and they haven't changed. Is it possible to run the newer HPC workloads in my old HPC environment if I already have it set up? Can companies that have invested heavily in on-premises HPC just stay on the same trajectory? Batta: A lot of the latest HPC workloads are the more cutting-edge workloads were born in the cloud. You can absolutely run those on old HPC hardware. But they're generally cloud-first, meaning that they have been integrated into graphics processing units (GPUs). Nvidia, for example, is doing a great job of making sure any new workloads that pop up are already hardware accelerated. In terms of general-purpose legacy workloads, a lot of that stuff is not GPU accelerated. If you think about crash testing, for example, that's still not completely prevalent on GPUs. Even though you could run it on GPUs if you wanted, there's still a long-term timeline for those applications to move on. So, yes, you can run new stuff on the old HPC hardware. But the likelihood is that those newer workloads have already been accelerated by other means, and so it becomes a bit of a wash. In other words, these newer workloads are built cloud-native, so trying to run them on premises on legacy hardware is a bit like trying to put a square peg in a round hole. Is that correct? Batta: Exactly. And you know, somebody may do that, because they've already invested in a big data center on premises and it makes sense. But I think over time this is going to be the case less and less. Come talk with Karan and others about HPC on Oracle Cloud Infrastructure at SC18 in Dallas next week in booth #2806.

Welcome to Oracle Cloud Infrastructure Innovators, a series of occasional articles featuring advice, insights, and fresh ideas from IT industry experts and Oracle cloud thought leaders. High...

Advances in Research

Critical Research Gets a Boost From Free Oracle Cloud Computing

This article was written by Sasha Banks-Louie and was originally published in Forbes. One research team is combining artificial intelligence and computer vision technology to help treat diabetics. Another is using 3D imaging to analyze rocks and predict their capacity to absorb carbon dioxide, and thereby reduce global warming. A third team created a platform used in designing new vaccines. These life-changing efforts are part of a program called Oracle for Research. Its goal is to help researchers take on some of the world’s most pressing problems and yield measurable results within the next five years. As part of the program, Oracle is providing researchers with cloud computing resources, technical support, and data expertise. Problems like those described above are data-intensive and require massive amounts of information to be processed quickly. Researchers affiliated with academic institutions or nonprofit research organizations worldwide can submit online their own projects for application to the Oracle program. “Granting access to high-performance computing power alone is not enough,” says Alison Derbenwick Miller, who runs Oracle for Research. “Most researchers are neither computing experts nor data scientists, so we give them access to a dedicated team of technical experts and architects to allow researchers to focus on what they know best—their research and their results.”  For Moi Hoon Yap, that research involves using artificial intelligence to help clinicians treat patients with diabetes. A professor of computer vision and artificial intelligence at Manchester Metropolitan University, Yap, Professor Neil Reeves, and their team of researchers are working with the UK’s National Health Services and Oracle for Research to develop FootSnap AI, a mobile app that lets diabetics and their doctors quickly diagnose foot ulcers. Diabetics frequently suffer nerve damage to their extremities that can cause a loss of foot sensation, so they might not notice a problem with their skin—“even when it’s breaking down or forming an ulcer,” Yap says. If such ulcers go untreated, they can infect the foot and require it to be amputated. FootSnap AI can respond to new demands rapidly, “with the cloud infrastructure speeding up the inference time and providing better accuracy in ulcer detection,” Yap says. To train its machine-learning algorithm, FootSnap AI ingests thousands of images of diabetic foot ulcers, supplied and annotated by podiatrists at Lancashire Teaching Hospitals NHS Foundation Trust. When a patient uploads an image of his or her foot to the app, the FootSnap algorithm looks for similar characteristics to those other images. The model runs on a virtual machine and an Nvidia P100 GPU on Oracle Cloud Infrastructure. Since upgrading to Oracle, “we’re not spending time maintaining servers anymore,” says Bill Cassidy, a research associate and lead application developer on the project. “It affords us a lot more time to do the real work of researching and writing papers about how to solve this health crisis.” Removing Carbon from the Atmosphere Another researcher in the program is Saswata Hier-Majumder, a professor of geophysics at Royal Holloway University of London, who is working on a project to capture carbon dioxide (CO2) in the atmosphere and permanently store it in rocks underground. With his team of PhD students, he has developed a simulation that analyzes digital images of rocks and predicts their capacity to absorb CO2 and organically remineralize it. The team takes images captured with 3D microtomography and runs them through its simulation engine to determine the pore volume of each fragment. A rock with 15% porosity might be able to hold and mineralize twice the amount of liquid CO2 as one with 5%. Royal Holloway University’s simulation also runs on Oracle Cloud Infrastructure, which lets researchers pick the amount of memory and threads needed to process the massive amounts of scanned images in a way that the team’s previous, on-premises computing options couldn’t. Says Hier-Majumder: “Oracle has helped us break the barrier of how much computational power we have in the lab.” A third effort involves researchers from the University of Bristol and vaccine-technology startup Imophoron. They tapped Oracle’s program to help build what they describe as a vaccine design platform. The platform provides an “atomic blueprint of the common nanoparticle scaffold we now use for all vaccine designs,” says Imre Berger, professor of synthetic biology at the University of Bristol and cofounder of Imophoron. Building that scaffold involved huge volumes of 3D images taken by an electron microscope and then processed using the high-performance computing capabilities of Oracle Cloud Infrastructure. Last year, the lab used the design platform for work on a vaccine against the mosquito-borne illness called chikungunya. Discover what you can accomplish with Oracle for Research. 

This article was written by Sasha Banks-Louie and was originally published in Forbes. One research team is combining artificial intelligence and computer vision technology to help treat diabetics....

Research Computing

ANSYS and Oracle: ANSYS Fluent on Bare Metal IaaS

If you’ve ever seen a rocket launch, flown on an airplane, driven a car, used a computer, touched a mobile device, crossed a bridge, or put on wearable technology, you’ve likely used a product in whose creation ANSYS software played a critical role. ANSYS is a global leader in engineering simulation. Oracle is pleased to announce its partnership with ANSYS. Oracle Cloud Infrastructure bare metal compute instances enable you to run ANSYS in the cloud with the same performance as you would see in your on-premises data center. Why Bare Metal Is Better for HPC Oracle Cloud Infrastructure continues to invest in HPC. Nothing beats the performance of bare metal. The virtualized, multi-tenant platforms common to most public clouds are subject to performance overhead. Traditional cloud offerings require a hypervisor to enable management capabilities that are required to run multiple virtual machines on a single physical server. This additional overhead has been demonstrated by hardware manufacturers to significantly affect performance [i]. Bare metal servers, without a hypervisor, deliver uncompromising and consistent performance for high performance computing (HPC). Instances with the latest generation NVMe SSDs, providing millions of IOPS and very low latency, combined with Oracle Cloud Infrastructure's managed POSIX file system, ensure that Oracle Cloud Infrastructure supports the most demanding HPC workloads. Our bare metal compute instances are powered by the latest Intel Xeon Processors and secured by the most advanced network and data center architecture, yet they are available in minutes when you need them—in the same data centers, on the same networks, and accessible through the same portals and APIs as other IaaS resources. With Oracle Cloud Infrastructure’s GPU instances, you also have a high performance graphical interface to pre- and post-process ANSYS simulations. ANSYS Performance on Bare Metal OCI Instances The performance of ANSYS Fluent software on Oracle Cloud Infrastructure bare metal instances meets and in some cases exceeds the raw performance of other on-premises HPC clusters, demonstrating that HPC can run well in the cloud. Additionally, consistent results demonstrate the predictable performance and reliability of bare metal instances. The following chart shows the raw performance data of the ANSYS Fluent f1_racecar_140m benchmark on Oracle Cloud Infrastructure's Skylake and Haswell compute instances. The model is 140 million cell CFD model. Visit the ANSYS benchmark database to see how Oracle Cloud Infrastructure compares favorably to on-premises clusters. Figure 1: ANSYS Fluent Rating on Oracle Cloud Infrastructure Instances Installation and configuration of ANSYS Fluent on Oracle Cloud Infrastructure is simple, and the experience is identical to the on-premises installation process. Bare metal enables easy migration of HPC applications; no additional work is required for compiling, installing specialized virtual machine drivers, or logging utilities. Although the performance is equal to an on-premises HPC cluster, the pricing is not. You can easily spend $120,000 or more on a 128-core HPC cluster [ii], and that's just for the hardware; that number doesn’t include power, cooling, and administration. That same cluster costs just $8 per hour on Oracle Cloud Infrastructure. That’s an operating expense you’re paying for only when you use it, not a large capital expense you have to try to “right-size” and keep constantly in use to experience the best ROI. Running on Oracle Cloud Infrastructure means that you can budget ANSYS Fluent jobs precisely, in advance, and the elastic capacity of the cloud means that you never have to wait in a queue. Scaling Is Consistent with On-Premises Environments When virtualized in your data center, CPU-intensive tasks that require little system interaction, normally, experience very little impact or CPU overhead.[iii] However, virtualized environments in the cloud include monitoring, which adds significant overhead on per node. Virtualization overhead is not synchronized across an entire cluster, which creates problems for MPI jobs, such as ANSYS Fluent, which effectively have to wait for the slowest node in a cluster to return data before advancing to the next simulation iteration. You’re only as fast as your slowest node, noisiest neighbor, or overburdened network. With Oracle Cloud Infrastructure’s bare metal environment, no hypervisor or monitoring software runs on your compute instance. With limited overhead, ANSYS Fluent scales across multiple nodes just as well as it would in your data center. Our flat, non-oversubscribed network virtualizes network IO on the core network, instead of depending on a hypervisor and consuming resources on your compute instance. The two 25Gb network interfaces on each node guarantee low latency and high throughput between nodes. As shown in the following chart, many ANSYS Fluent models scale well across the network.     Figure 2: ANSYS Fluent Scaling on an Oracle Cloud Infrastructure Instance The following chart illustrates greater than 100% efficiency with respect to a single core from 400,000 cells per core to below 50,000 cells per core. Figure 3: Efficiency Remains at 100% Even as Cells Per Core Drop Serious HPC Simulations in the Cloud Oracle Cloud Infrastructure has partnered with ANSYS to provide leading HPC engineering software on high performance bare metal instances so that you can take advantage of cloud economics and scale for your HPC workloads. Our performance and scaling with ANSYS matches on-premises clusters. It’s easy to create your own HPC cluster, and the cost is predictable and consistent. No more waiting for the queue to clear up for your high-priority ANSYS Fluent job or over-provisioning hardware. Sign up for 24 free hours of a 208-core cluster or learn more about Oracle Cloud Infrastructure's HPC offerings. And for more examples of how Oracle Cloud outperforms the competition, follow the #LetsProveIt hashtag on Twitter. [i] http://en.community.dell.com/techcenter/high-performance-computing/b/general_hpc/archive/2014/11/04/containers-docker-virtual-machines-and-hpc [ii] Example price card: https://www.hawaii.edu/its/ci/price-card/ [iii] https://personal.denison.edu/~bressoud/barceloleggbressoudmcurcsm2.pdf

If you’ve ever seen a rocket launch, flown on an airplane, driven a car, used a computer, touched a mobile device, crossed a bridge, or put on wearable technology, you’ve likely used a product in...

Research Computing

Exabyte.io for Scientific Computing on Oracle Cloud Infrastructure HPC

We recently invited Exabyte.io, a cloud-based, nanoscale modeling platform that accelerates research and development of new materials, to test the high-performance computing (HPC) hardware in Oracle Cloud Infrastructure. Their results were similar to the performance that our customers have been seeing and what other independent software vendors (ISVs) have been reporting: Oracle Cloud Infrastructure provides the best HPC performance for engineering and simulation workloads. Exabyte.io enables their customers to design chemicals, catalysts, polymers, microprocessors, solar cells, and batteries with their Materials Discovery Cloud. Exabyte.io allows scientists in enterprise R&D units to reliably exploit nanoscale modeling tools, collaborate, and organize research in a single platform. As Exabyte.io seeks to provide their customers with the highest-performing and lowest-cost modeling and simulation solutions, they have done extensive research and benchmarking with cloud-based HPC solutions. We were eager to have them test the Oracle Cloud Infrastructure HPC hardware. Exabyte.io ran several benchmarks, including general dense matrix algebra with LINPACK, density functional theory with Vienna Ab-initio Simulation Package (VASP), and molecular dynamics with GROMACS. The results were impressive and prove the value, performance, and scale of HPC on Oracle Cloud Infrastructure. The advantage of Oracle Cloud Infrastructure's bare metal was obvious with LINPACK, throughput is almost double the closest cloud competitor and consistent with on-premises performance. Latency is even more interesting: the BM.HPC2.36 shape with RDMA provides the lowest latency at any packet size and is orders of magnitude faster than cloud competitors. In fact, for every performance metric that Exabyte.io tested on VASP and GROMACS, they saw Oracle's BM.HPC2.36 shape with RDMA (shown as OL in the following graph) outperform the other cloud competitors. Below is a great example of both performance and scaling of Oracle Cloud Infrastructure on VASP. When parallelizing over electronic bands for large-unit-cell materials and normalizing for core count, the single node performance of the BM.HPC2.36 exceeds it's competitors and then scales consistently as the cluster size increases. The BM.HPC2.36 runs large VASP jobs faster and can scale larger than any other cloud competitor. Exabyte.io has provided the full test results on their website. Their blog concluded that "Running modeling and simulations on the cloud with similar performance as on-premises is no longer a dream. If you had doubts about this before, now might be the right time to give it another try." By offering bare metal HPC performance in the cloud Oracle Cloud Infrastructure enables customers running the largest workloads on the most challenging engineering and science problems to get their results faster. The results that Exabyte.io has seen are exceptional, but the results are not unique among our customers. Spin up your own HPC cluster in 15 minutes on Oracle Cloud Infrastructure.  

We recently invited Exabyte.io, a cloud-based, nanoscale modeling platform that accelerates research and development of new materials, to test the high-performance computing (HPC) hardware in Oracle...