 From around the globe, it's theCUBE with digital coverage of AWS re-invent 2020. Sponsored by Intel and AWS. Hello everyone and welcome back to theCUBE's wall-to-wall coverage of AWS re-invent 2020. We've gone virtual along with re-invent and we heard in Andy Jassy's hours long keynote, a number of new innovations in the area of storage and with me to talk about that is Mylon Thompson Bukebeck. She's the vice president of Block and Object Storage at AWS. That's everything, Elastic Block Storage, S3 Glacier, the whole portfolio. Mylon, thanks for coming on. Great to see you. Great to see you too. So you heard Andy, we all heard Andy talk a lot about reinventing different parts of the platform, reinventing industries and really kind of exciting and visionary talk that he put forth. Let's talk about storage though. How is storage reinventing itself? Well, as you know, cloud storage was essentially invented by AWS a number of years ago and whether that's in 2006 when S3 was launched or 2008 when EBS was launched. And we first came up with this model of pay as you go for durable attached storage to EC2 instances. And so we haven't stopped and we haven't slowed down if anything we've picked up the rate of reinvention that we've done across the portfolio for storage. I think as Andy called out, speed matters and it matters for how customers are thinking about how do they pivot and move to the cloud as quickly as they can, particularly this year. And it matters a lot in storage as well because the changing access patterns of what customers are doing with their new cloud applications, they're transforming their businesses and their applications and they need a modern storage platform underneath it. And that's what you have with AWS storage. Andy talked about some of the key releases particularly in block storage. It's actually kind of amazing what's been done with EBS is here. We launched GP3. GP2 was the previous generation general purpose volume type and we launched that in 2014. Again, the first type of general purpose volume that had this great combination of simplicity and price and just about everybody uses it for a boot or often a data volume. And with GP3, which was available yesterday with Andy's announcement, we added four times peak throughput on top of GP2 and it's a 20% lower storage price per gigabyte per month and we took the feedback, the number one feedback we got on GP2, which was how can I separate buying throughput and IOPS from storage capacity? And that is really important. That goes back to the promise of the cloud and it goes back to being able to pick what aspects do you wanna scale your storage on? And so with GP3, you can buy a certain amount of capacity and if you're good with that capacity but you need more throughput and more IOPS, you can buy those independently. And that is that fine grain customization for those changing data patterns that I just talked about and it's available for GP3 today. Yeah, I looked at that like my line is a knob that you could turn, it's okay, hey, just juice my IOPS and don't touch my capacity, I'm happy there. I don't wanna pay for more, love it. To add to that, it's a knob you can turn if you need it. We have more throughput and more IOPS as a baseline capacity for your storage capacity than we did for GP2. But then you can tune it based on whatever you need not just now but in the future. So given the pandemic, I mean, how is that affected? I mean, everybody's talking about going to the cloud because where else you're gonna go but how is that affected what customers are doing this year? And does it change your roadmap at all? Does it change your thinking? Well, I have to say there's two main things that we've seen. One is it's really accelerated customers thinking about getting off of on premises and into the cloud. It's done that because nobody really wants to manage the data center. And if there's ever a year you don't wanna manage the data center, it's 2020. And it's because particularly with storage appliances it takes a long time to acquire let's just take storage area networks or cents. Super expensive, you get a fixed amount of capacity you have to acquire it, it takes months to come in you got to rack and stack then you gotta change all your networking and maintain it. A lot of customers don't wanna do that. And so what it's done for us is it's really accelerated our thinking and you saw it yesterday in Andy's keynote as well of how do we build the first sand in the cloud? And we launched IO2 in August of this year we introduced the first nines of durability again reinventing how people think about durability and their block storage, but just this week we now have IO2 block express with 256K IOPS 4K megabytes of throughput and 64 terabytes of capacity that's sand level performance and it's available for preview because IO2 is going to be your sand in the cloud. And that is a direct correlation to what we hear from customers which is how can I get away from these expensive on-premises purchases like sands and combine the performance with the elasticity that I need. So that's the first thing how can we accelerate getting off of these very rigid procurement cycles that we have and having to manage a data center it's not just for EBS it's for S3 as well. The second thing we're hearing from customers is how can I have the agility? So you talk to customers as well you talk to CIOs and CTOs it's been a crazy year in 2020 and it was one thing that a company has to do it's pivot it's really figure out how are you going to adjust and adjust quickly? And so we have customers like Ontario telehealth network up in Canada where they went from 8,000 to 30,000 users because they're doing virtual health for Ontario and we have other customers who, you know that's a pivot that's an increase and we have other customers like AppsFlyer where their goal is to just save money without changing their application and they also did a pivot they used an intelligent hearing storage class which is the most popular storage class S3 offers for data lakes and they were able to make that change which they've 18% on their storage cost no change to their application just using the capabilities of AWS and so this ability to pivot helped really make us think and accelerate what we're building as well and so one of the things that we launched just recently for intelligent hearing is we added two new archival tiers to intelligent hearing and those archival tiers just like intelligent hearing automatically watches every object in S3 storage in your data lake and gives you dynamic pricing based on if it's frequently accessed in a month or infrequently accessed you can turn on an archival tier and if your object, your parquet file for example isn't accessed or your backup isn't accessed for 90 days intelligent hearing will automatically move it to glacier characteristics of archival or to deep archive and give you the same price a dollar a terabyte per month if your data isn't accessed in 180 days it's done automatically and it means you save up to 95% in cost on that storage and so if you think about those two trends how can I get away from getting locked into those on-premises hardware cycles how can I get away from it faster for SANS and other hardware appliances and then the other trend is how can I pivot and use the innovation and the reinvention in our storage services to just save money and be more agile in these changing conditions? So I got to ask you a follow-up question on sand in the cloud because when you think of sand you think of switches you think of complexity but I get that you're connecting to the performance of a sand but you guys are all about simplicity so how did you, what's behind there? Can you take us under the covers? Did you guys build your own little storage network because it's cloud it's got to be fast and simple. That's right. When we're thinking about performance and cost we go down to the metal for this stuff we think about unit cost at a very fine grain level and when we're building new technology that we know is gonna be the foundation for everything we're doing for that high performance we went down to the protocol level we're using something called SRD it's all rolled up under the hood for Block Express and it's the foundation of that super, super high performance. As you know, there's a lot of engineering behind the scenes in the cloud and for what we've done this year as part of that reinvention we've reinvented all the way down to the protocol layer. Love it. Let me ask you that the two things that come up in our survey when you talk to CIOs they say two priorities. Security is actually second cloud migration actually popped up to the top. So where does storage fit in that whole notion of cloud migration? Storage is usually where a lot of people start. You know, luckily with AWS you don't have to choose between security or cloud or migration. Security is job one for every AWS service. And so when customers are thinking about how do I move an application? They got to move the data first. And so they start from the data what storage do I use? What is the best fit for the storage and how do I best secure that storage? And so the innovation that we do on storage always comes with that combination of migration, the set of tools that we provide for getting data from on-premises into the cloud. We have tools like AWS Datasync which do a great job of this. And then we also look at things like how do we continue to take the profile of security forward? And one example of that is something we launched just this week called bucket keys S3 bucket keys. And it drops the cost of using KMS for service site encryption with S3 by over 90%. And the way it does it is that we've integrated those two services super closely together so that you can minimize the amount of costs that you make for very, very frequent requests because in data lakes you have millions and billions of objects. And our goal is to make security so cost effective people don't even think about it. That also goes for other parts of the platform. We have guard duty for S3 now. And what that does is security anomaly detection automatically to track your access patterns across S3 and flag when something is not quite what it should be. And so this idea of like, how do I not only get my data into the cloud but then how do I take advantage of the breadth of the storage portfolio but also the breadth of the AWS services to really maximize that security profile as well as the access patterns that I want for my application. Well, we hit the major announcements and unfortunately we're out of time but I really would love to have you back and go deeper and have you share your vision of what the cloud storage piece looks like going forward. Thanks so much for coming in theCUBE. It's great to have you. Great to be here. Thanks Dave, see ya. See you later. And keep it right there, everybody. You're watching theCUBE's coverage of AWS re-invent 2020, we're right back.