So how big is big? So a manufacturing plant with 30,000 sensors, sampling at 100 milliseconds, producing 16 bytes of data per sample, yields? [SOUND] You do the math on that, and comes out to 48,000 bytes per second or 4.1 gigabytes a day. If that was going on 24/7, we end up generating 1.5 terabytes a year but that's nothing [LAUGH]. This database called the Multi-wave Infrared Atlas of the Galactic Plane, data collected by astronomers. That database is 45 terabytes. Now we're starting to get there. I like this one. Just up, this is one of the changes I made today, went and did some poking around when I was at work on my lunch hour. And then when I got home, I updated the slide. Pratt and Whitney's, they make jet engines. They got the Geared Turbo Fan engine, get this, 5000 sensors in one jet engine. I always knew they were highly instrumented but I always figured there was like 2 or 300 sensors. There's 5,000 sensors in one jet engine. You can go read the link on aviation week about this. That's the source of this data. So I did the math and I said, well, so it's generating 10 gigabytes of data per second. [LAUGH] That's an enormous amount of data. A two hour flight, it generates 72 terabytes. That's just one engine. There's two of those and it's 144 terabytes for a 2 hour flight, from here to Chicago. I don't know if they stream all that data to a server farm bolted under the bottom of the airplane [LAUGH]. I don't know where all that data goes. I suspect they're selective about what they pick. If they wanted to, they could save that much data. I don't know where all that data is stored on an airplane. I don't know enough about aircraft construction and what's reasonable. because every additional drive that they put in there, that's added weight, and then it's less baggage and less passengers. And just as a data point, one Watson system typically has 16 terabytes of RAM, I thought that was pretty high for a system, but that doesn't surprise me, it's in the range. I thought it might have been 4 or 8 terabytes of RAM, not 16TB, typically. So you have all heard of terabytes before. You know that they are a 1000 gigabytes. And a petabyte is a 1000 terabytes. And an exabyte is a 1000 petabytes. And I heard a senior director, he's either director or VP at Microsoft, recently talk, and in his estimate, the Internet contains about 170 exabytes worth of data today. I have a statistic to share with you. When I was working at Seagate in 2011, we had these quarterly all hand meetings, we all go down to this big meeting room and we get an update on how a company is doing and so forth. And I don't remember who was giving the presentation but it doesn't matter. And this was the one, that was at the end of 2011, and the executive that was talking to us said, so who's uploaded a video to YouTube. And all these people put their hands up, right? Yep, yep. He said, so here's something to think about. If you added up all of the video if you took all the TV shows and all the movies that were created from the very first black and white soundless movies. And you turned all that into digital video and added all that up to the beginning of 2011, more video was uploaded to YouTube in 2011 than all the video going back in history. And I thought, wow, [LAUGH] that's a lot of data. And it's great news for people in the storage industry because customers have this insatiable desire for storage. And it's like we, as a worldwide society, we don't want to throw anything away. How many pictures have you taken and you've uploaded to Facebook or Snapchat or whereever, Pinterest. And you know what, you take pictures and you post them and you write something about them. And Facebook comes showing them to you, ones, remember this three years ago, and then you get to see it again, I totally forgot about that. And I never would have gone to even look at those pictures unless Facebook reminded me of this event three years ago, or five years ago, or however long it was. It's like we don't want to throw anything away. So, we keep generating more and more and more data. And the curve looks exponential to me, which is really great for being in the storage industry. Okay, so his estimate was 170 exabytes. And then a zetabyte is 1000 exabytes, and there's a term after this one, but I quit going because that's just an enormous amount of data. But these are some sizes, and who's heard the term Kibi before? Raise your hand if you've heard the term Kibi. [SOUND] Yes, you have? Okay, cool. Going off script. I don't remember what year it was that someone brought a lawsuit against Seagate because there's the, Drive was advertised to have, I don't remember what the exact size was, but it was supposed to have 512, Gigabytes of storage. When in fact it only had 500, so there was this lawsuit. [LAUGH] And I don't know if there was a payment made or how it settled. But it, I don't like that color, it created a new unit [COUGH] [LAUGH]. So when you see KB, That is exactly 1,000 bytes. If you see it with a little i in it, KiB, That's 1,024. So this is 10 to the 3rd, and this is 2 to the 10th. And guess what, it continues. So we've got megabytes, Mebibytes, Gigabytes, And gibibytes, that's how it's pronounced, jibby, jibby bytes. We've got terabytes, these are all base 10. Tebibytes, and on up. So there's a difference. So what we used to know as kilobyte being thinking 1,024, when I was in university, is now a base-10 number. So this is 10 to the 3rd, 10 to the 6th, 10 to the 9th, 10 to the 12th. And this is 2 to the 10th, 2 to the 20th, 2 to the 30th, 2 to the 40th.
