WASHINGTON--(BUSINESS WIRE)--Amazon Web Services (AWS) CEO Andy Jassy discusses how cloud-based services have buttressed the at-home economy in the COVID-19 era; how machine-learning tools are being deployed to speed the development of vaccines and the natural synergies between the tech and energy industries in the latest edition of the CERAWeek Conversations series.
In a conversation with Daniel Yergin, vice chairman, IHS Markit (NYSE: INFO), Jassy talks about the origins of AWS; the rapid growth of cloudfrom primarily used by smaller startups to being embraced by large companies and the public sector; and how providing reliable, scalable, cost-effective infrastructure allows people to innovate.
The complete video is available at: http://www.ceraweek.com/conversations
Selected excerpts:Interview Recorded Wednesday, June 24, 2020
(Edited slightly for brevity only)
It really spans the gamut in terms of customers. In the first half dozen years or so of AWS you mostly saw startups using the platform.
What's happened over the last seven or eight years, but particularly over the last five, is that the enterprise and the public sector have really started to adopt AWS and the cloud in a very pervasive way. You see it in every imaginable vertical business segment: In the energy sector; you see it in financial services; in healthcare; in the public sector we have 7,000 government agencies worldwide using AWS; 10,000 educational institutions; 25,000 non-profits. So, its a very broad and diverse customer base.
One of the biggest benefits of the cloud [is] the ability to move quickly and to innovate on behalf of your customers quickly. When you have data, as most enterprises do, that live all over the place in all these different silos, it's incredibly difficult to actually do much with that data or to do analytics on that data.
We started working with Shell on their subsurface data universe. What they really want to do was have this one place where they could run lots of analytics in almost endless simulations. The data scientists had all kinds of ways they wanted to use this data, so we built this with them.
They appropriately thought about it and said: we want to actually democratize this data set so that lots of other companies can do the same type of exploration and, as importantly, will get a number of ISVs (independent software vendors) and systems integrators who will build to enable us to do more in the analytics and then what we can do with this data. So, they really opened this upits just totally changed access to that data and what you can find from it and what you can do and how you can save money and time and energy in the exploration process.
You can look at Woodside which has gas plants all over the world. They were trying to innovate on the traditional way that you ran these [plants] and you did maintenance and you operated them. So, they went about using machine learning and AI to build an intelligent asset that tried to fuse together all of this historical data and how the plants were used and how they were operated and fusing that data with thousands of real time sensors on their various assets.
What they're able to do now is, they have all this historical data, they built models to analyze and predict what's going to happen based on historical data, they have real time inputs, they have a visual representation of what's happening in their plants and then they're able to run all these simulations on different scenarios, such that, now they're able to predict when there might be a failure or maintenance issue up to a week before it actually happens.
Moderna built really a digital manufacturing platform on top of AWS and they were able to use things like EC2, which is our compute service, and S3, which is our storage service, and Redshift, which is our data warehouse service.
Then they used our machine learning to build a number of very deep neural networks that allowed them to be able to predict structures of different strains of what they're trying to solve.
They were working on mRNA and they were trying to build a vaccine for SARS-COV-2 and they were able, by using machine learning [to] predict these structures and predict what they would have in various simulations. They were able to build a vaccine in 63 days in what normally takes 20 months. Under any circumstances that would be remarkable, but if you especially think about the environment in which we live today, it's very important.
They didn't have to buy the hardware. They didnt need the data center. They didnt have to build the infrastructure software. They just used our compute and our storage and our data warehouse and our machine learning capabilities to scale up and scale down as they needed to really in rapid fire be able to build something much faster than they otherwise would.
Energy companies have technology infrastructurewhether they use the cloud or don't use the cloudthey have on-premises technology infrastructure that they're going to use. If you look at research that's done, it's 3.6 times more energy efficient to run technology infrastructure in AWS and the cloud than to run it on-premises.
If you actually add the carbon intensity for the same task, it's 88% more energy efficient to run an AWS than it is on premises. Wed rather try to be a part of the solution and help energy companies be more energy efficient with the technology infrastructure they're going to use, but also save money and time so they can spend their innovation cycles working on the renewable projects that allow us to all get to 100% renewable like we want.
I'm not sure I know any companies that aren't focused on the environment right now and who don't understand how important that is to our future. There's such an urgency to get there, but if you're going to run technology infrastructure it is so much more energy efficient to be able to do so in the cloud than on-premises and that's because we have much better utilization in the cloud of servers and more efficient power. If you can be 3.6 times more energy efficient with the same workloads you have to run, but in the cloud, that changes what your own carbon footprint looks like.
Amazon by most people's standards in the 1990s and early 2000s was a company that innovated at a very rapid clip. And yet at the time we were very frustrated with the speed with which we were innovating. It felt like it was taking so long.
We asked people how much of [their] time when building these features or capabilities is spent on what really differentiates the capability versus that undifferentiated heavy lifting of the infrastructure. We were stunned to hear that people spent about 80% of their time on that undifferentiated heavy lifting. That was a real realization to us that, if you built reliable, scalable, cost-effective infrastructure services, it would allow Amazon to innovate at a much faster clip than we were before.
We thought about Amazon, which is a very strong technology company, and if we had that issue, we figured a lot of other companies probably had that same issue.
It changes what's possible with companies. If you can turn capital expense into a variable expense, usually that's attractive to people. In the cloud you dont have to lay out all that capital for the data centers and the hardwareyou get to just use the infrastructure as you need it. And then variable expenses is a lower expense than what most companies can do on their own because we have such giant scale that we pass on to customers in the form of lower prices. Weve lowered our prices on almost 80 different occasions in the last 10 years.
Anyone who needs to provision infrastructure knows that you have to guess ahead of time how much you need. If you guess too little, you end up having an outage if it turns out that you need moreso nobody guesses too little, everybody provisions for the peak. Theres a reason they call it the peak, you rarely sit at the peak.
In the cloud you don't have to guess. You provision what you need. If turns out you need more, you seamlessly scale up and if it turns out you hit a peak, you just give it back to us and stop paying for it. But the number one reason that people have adopted the cloud and the way it's most changed a lot of businesses is just the agility and speed with which you can innovate and change your own customer experience.
The first is who you hire. If you hire people who are content moving slow and operating in an incumbent pace, you will move slow. Whereas we overly index on who we hire on builders. We think about builders as those who like to invent, but also those who look at a customer experience and try to figure out what's wrong with it and then have this hunger to change that and to reinvent that.
Then we try to organize them in as small and separable and autonomous teams as we can, so they own their own destiny. If teams own all the resourcesthere's no more finger pointing between engineering and product management, operations; they're all in the same team, and they're all owning it togetherthey take charge of their own destiny.
The last piece is, if you want to invent and you want people to take chances and work on new initiatives, you have to be able to tolerate failure. None of us like failure at Amazon. But if you don't allow people to take chances and have an opportunity if it doesn't work out to be able to move on to something great, it's a self-fulfilling prophecy: none of the good people work on the new things because it will be too risky.
We talk a lot about how everything we build starts with the customer and then moves backwards from there. If you sat in our product development meetings you would see that it's almost like the customer sitting at the table. It's always with a customer lens on it. It's easy to talk about being customer focused and a lot harder to actually execute on that. I think the company is unusually focused on customers.
Just in the last few days we've announced this $2 billion fund to fund renewable projects that change the speed with which we can all get to carbon zero. It's a big priority for the company. It's a big priority for AWS and for the rest of Amazon and we're looking for as many partners as we can to help accomplish it.
Watch the complete video at: http://www.ceraweek.com/conversations
About CERAWeek Conversations:
CERAWeek Conversations features original interviews and discussion with energy industry leaders, government officials and policymakers, leaders from the technology, financial and industrial communitiesand energy technology innovators.
The series is produced by the team responsible for the worlds preeminent energy conference, CERAWeek by IHS Markit.
New installments will be added weekly at http://www.ceraweek.com/conversations.
Recent segments also include:
A complete video library is available at http://www.ceraweek.com/conversations.
About IHS Markit (www.ihsmarkit.com)
IHS Markit (NYSE: INFO) is a world leader in critical information, analytics and solutions for the major industries and markets that drive economies worldwide. The company delivers next-generation information, analytics and solutions to customers in business, finance and government, improving their operational efficiency and providing deep insights that lead to well-informed, confident decisions. IHS Markit has more than 50,000 business and government customers, including 80 percent of the Fortune Global 500 and the worlds leading financial institutions. Headquartered in London, IHS Markit is committed to sustainable, profitable growth.
IHS Markit is a registered trademark of IHS Markit Ltd. and/or its affiliates. All other company and product names may be trademarks of their respective owners 2020 IHS Markit Ltd. All rights reserved.
Read More..