 By that rule right to seven so 30% of this room is recovering physicists. I think that's true You can you can write it down all right Mike Miller of cloud and I'm the last thing between you and a drink So I'll try to keep it quick And give us a second here. So yeah, all right, so it's a lightning talk I'll try to be fun with this in the spirit of data chick and tell you why databases of service is not a thing That you're just gonna consider is weird, but a decade from now. It's our last hope So what do I mean by that? I promise a quantitative argument So I'm gonna start with a little bit of philosophy my first postulate I guess of humans we don't get Exponential growth right it's the reason that we continue to fill up our planets the reason that we continue to Grow actively into limited resources that folded with reality is gonna get us to this conclusion Which is that in a decade you are not gonna dream of installing your own database? Okay, this is my takeaway and now I'm gonna try to convince you of it So my apologies to the DBAs in the room, but I think that you know my goal is to extinguish your field Right, that's where we're heading and I'm just gonna be upfront and blunt about that because it's gonna allow you to do other things Right, you want to be the Instagrams of the world. So how do we get there? We're gonna start with a large chat around Clidra and when I honestly learned what it means to be Exponential this is a picture of one of my colleagues on the floor It's a five-story device at CERN a couple hundred meters underground This thing makes data at about an exabyte a second and I can get to that calculation for the physicist in the room later on But the interesting thing is you can't dream of recording that right you cannot dream of recording all that digitized data This is the thing that that actually samples that but there's a Clidra You know buried under the mountains that creates you know something like In every 10 nanoseconds there's a collision that lights up this this entire device and that Clidra itself You can think of it like a big flashlight Okay, it's got beams of particles and they get brighter and brighter year over year and the net results is about every two years through various Various kind of knobs you can turn and tune this thing doubles in the amount of data that it could potentially produce It took me a while to understand what that meant meant that every two years I would record enough data that represented the entire amount of particle physics data with a squint to the physicists in the room That means that everything we had ever discovered in physics could be replicated in two years, right? So as a grad student, it's like oh, I can just redo everything right with that data took me a while to understand what that really meant And that's playing out right now So this is you know I cited something to try to find the doubling time of the volume of data that we have digital data in the world and You know it's doubling something like every two years, okay? So right now I think in 2009 we're sitting at about one zeta bite 10 to the 21 So I think it's a trillion gigabytes right and so that means if you go two more years you've doubled that Okay, let's think about you can draw a line anywhere on this graph and say I care about here, right? I'm at 400. What does it take to go up to 800 zeta bites? Well, it means that on the left side I have the entire history of humankind and on the right side I have the entire history of humankind plus two years Okay, so two years. So as a vendor I am focusing on the new stuff, right? This is all new right and what we know is what's in our brains in our history, which is all of the old stuff So while we think it's crazy to move data on into the cloud, right when I started the company I said we're gonna take your database when to put it on the internet We're gonna put it everywhere right like like Akamai for your database and people said that's crazy Right, but the reality is that solves a lot of new problems, which I want to talk about now So these two things have broken our model of computing if your developers say with me the lamp stack is dead Ten years from now. It's not going to work right then this story. I just told you about data volume It's true of connected internet devices to right We've long since passed the place and time where the number of connected devices is larger than the number of humans on the globe Right, we're way past that so these two things conspire to break the way that we've built our stack Right and it is all being reinvented. That's why it's such an awesome time to be a founder in the space Name anything it's all being reinvented from scratch, right? And it's being reinvented not necessarily by the big names in the field But by people writing Ruby on Rails apps that have to solve these problems That's where no sequel came from in the first place And yeah, if Fusion IO was cheaper in 2008 no sequel may not have happened for a long time Right, it's just a response So they're these market pressures and the things that are coming out of this I think are two big examples are no sequel and cloud services, right? So our company in cloud and its database is a service. We ship with a mobile strategy Come see me if you want to know what that means But if you look at where this falls in the realm of other cloud products, you know our competition is Amazon More small company you may never have heard of cloud in we're probably the smallest company with a digit revenues That you've never heard of in the no sequel space But the reason is because this market is really really good Right and you want to be in a really good and fast-growing market And this is the fastest growing product that AWS has ever rolled out So if you're interested come join those in mobile gaming in the enterprise in big mobile And come talk to me afterwards over beer for some of these arguments. Thanks. All right