Paul Ziots - Director of Investor Relations
Andrew Mendelsohn - Senior Vice President of Oracle Database Server Technologies
Kash G. Rangan - BofA Merrill Lynch, Research Division
Oracle Corporation (ORCL) Executive Access for Investors Conference April 12, 2012 12:00 PM ET
Good day, everyone, and welcome to the Oracle Executive Access for Investors Conference Call. Today's call is being recorded. At this time, I'd like to turn the call over to Paul Ziots, Director of Investor Relations. Please go ahead.
Thank you, Carissa. Hello, everyone, and thank you for joining us today for Oracle's Executive Access for Investors, an educational webcast series hosted by Oracle. Today is Thursday, April 12, 2012.
Joining us today is Andy Mendelson, SVP, Oracle Database Server Technologies; and Equity Research Analyst, Kash Rangan, from Bank of America Merrill Lynch. Today, Andy will discuss Big Data, why this is good for Oracle and Oracle's key products and technologies, including the Oracle Big Data Appliance.
Please note that Andy will not be discussing any information today that is not already publicly available. Also, please note that we'll be conducting a separate web event on another hot topic: Analytics, and Oracle's Exalytics In-Memory Machine. That will be later this month.
At the conclusion of Andy's presentation, we'll turn the webcast over to Kash, who will kick off the Q&A. You may submit questions at any time during the presentation by clicking on the Ask a Question tab above the webcast slides. Please keep in mind, we will not comment on business in the current quarter.
As a reminder, the matters we'll be discussing today may include forward-looking statements and as such, are subject to the risks and uncertainties that we discuss in detail in our documents filed with the SEC. Specifically, the most recent reports on Form 10-K and 10-Q, which identify important risk factors, which could cause actual results to differ from those contained in forward-looking statements. You're cautioned not to place undue reliance on these forward-looking statements, which reflect our opinions only as of the date of this presentation. Please keep in mind that we're not obligating ourselves to revise, update or publicly release the results of any revisions of these forward-looking statements in light of new information or future events. Lastly, unauthorized recording of this webcast is not permitted.
And now I'll turn it over to Andy.
Thanks, Paul. Before we get into Big Data, I thought I'd just let -- alert you guys to the fact that recently, Gartner just published their relational database market share data for 2011. And so it's interesting in this era of Big Data to look at what happened, according to Gartner, over the last year. And we're very happy that -- to report that Gartner's saying the database market's very healthy, growing over 16.3% per year at a $24 billion market. We're also very happy that they've announced that Oracle grew over 18% in the database market, and we took market share, and our #2 and #3 competitors, Microsoft and IBM, actually lost market share in that market. And the competitor, Teradata, that's actually known as a big data warehouse company, they actually didn't grow nearly as fast as we did in this very vibrant relational database market. And so feel free to go out and talk to the Gartner analysts and get more details on what happened over the last year.
Okay. So with that, let's move into the Big Data space. So everybody is seeing a lot of buzz around Big Data, and everybody is trying to figure out what does this mean for our customers and our industry, and let me just start off by sort of explaining this with an analogy. So we all know cars. Henry Ford invented the Model T 100 years ago. And today, Ford makes cars, and they have much better engines, and they have their full sensors and computer systems. But at the end of the day, they're still cars. Cars have evolved a lot over the last 100 years.
And Big Data should be thought of in the same way. So today, we have our information systems for business intelligence, and people call them data marts and data warehouses, and they're loaded with all this transactional information from our transaction processing systems like E-Business Suite and other application vendors. And that information is really valuable, the crown jewels of company, that transactional data. It's not going away. But what people want to do with Big Data is that they just want to look at capturing new kinds of information to enhance and enrich that transactional data that they're currently using.
So, for example, if you're a retailer, you might want to go out to Facebook and pull out information about -- from your customers' Facebook pages if they're willing to friend you, and most of that information is completely worthless, right? All the pictures of babies and families and all that stuff. You don't want to keep that in your relational databases. But the fact that somebody actually had a baby is really interesting to a retailer, right? They can use that to sell -- upsell baby bottles and baby toys and everything else to baby.
So what's really important about Big Data is to understand there's a lot of this data, most of it's completely worthless to the business, but they're really these gems, these nuggets of information, like the fact a customer just had a baby, you want to take that information, you want to integrate it to your existing transactional data that you've got in your data warehouse and really use that to better -- make better business decisions and make more money for your company. Okay.
So with that, let's go to the next slide. And the basic idea with Big Data is that across all the various industries, there is new kinds of data that people are looking at using to augment their transactional data, to use that to grow their business, make better business decisions.
And on this slide, we just highlight 5 different kinds of Big Data. We go through -- there's health care, there's a lot of remote patient monitoring and sensor data manufacturing. Organizations are using sensors to gather a lot of information about the manufacturing process. Of course, everybody knows that our cell phones have GPSs and are gathering location data about our every movement. Retailers are looking at social media, like I just mentioned, and trying to understand what their customers are up to and doing things like sentiment analysis.
I thought I'll just walk you through health care a little bit because everybody really understands this really well. So what's this all about? So you're out in the tennis court, you're playing tennis. Suddenly, your arm doesn't feel quite right. You sit down, you're short of breath. You go to the doctor, the doctor looks at you and you're perfectly fine. He can't see anything wrong.
And so in the era of Big Data, what are we doing now? So we put sensors. The doctor puts a heart monitoring sensor on the patient. And he's out there. And again, most of that data being gathered is completely worthless. It's just says the patient's heart is just perfectly fine. And then he goes out and plays tennis again and it happens again. And this time, the sensors are capturing what's going on. And this is a really good example of Big Data. That information about the event that happened when you started feeling pain, that's what the doctor wants to look at. He doesn't want to look at all that other data that you've been capturing all week that just says you're perfectly fine. And this is a very common thread in Big Data. There's a lot of it. Most of it's completely worthless to the business. You want to sift through that Big Data, you want to pull out the nuggets, in this case, about your potential heart problems. That's what you want to look at, that's what you want to integrate with your existing health care systems about this patient. You don't want the volumes of data that just says the patient's perfectly fine.
Okay. So let's go to the next slide. And the next slide is sort of what we and the industry analysts have sort of extracted away from Big Data. What are the key elements that make something Big Data that's really new? And they call that -- this the 4 Vs: volume, velocity, variety and value. And let me just go through this a little bit, because I think all of these are -- people think Big Data, they think, "Oh, it must -- it's just big. There's a lot of it." And that's true, but people like to think, "Oh well, existing information. There's just so much data, the existing information systems and databases can't deal with it." And that sort of -- that's crazy, right? The existing information systems and relational databases out there can deal with humongous amounts of data. We have petabyte data warehouses running on Oracle Exadata today.
Customers have no problems dealing with the volume of data, okay? But that is a characteristic of what people like talking about. Velocity is another thing people like talking about. So for example, this sensor data I mentioned, it's constantly streaming large amounts of data from this sensor on my heart to monitor what I'm doing. Or in a manufacturing process, there's a lot of this data, it's streaming, it's really fast. GPS data is the same way. So people like talking about velocity. And again, it's all about there's this large volume of data, and we can deal with large volumes of data. There's really nothing special about that.
And then what's really new here is variety. So there are new kinds of data, as I mentioned, beyond the traditional transactional data that people are using to make business decisions. And there's social media data like Facebook, there's Twitter, there's blogs, there's sensor data from smart meters and from manufacturing sensors or Health Mart monitors, things of that sort. So there's new kinds of data that people are gathering, and this is what people refer to as variety. This is really the core of what people are getting at in Big Data, just new kinds of information. We want to gather it. Some of it's structured, some of it's less structured, more textural in matter like Facebook pages or whatever. But the bottom line, it's just data, and we want to have information systems that are able to analyze that data and capture the valuable parts of it.
And then finally, value. And so between variety and value, I think you really get the real essence of Big Data. So like I mentioned, this Big Data, there's a lot of it and high volumes of it, and the real challenge is that most of it is -- has very low business value. And so the real challenge is you want to sift through this data, you want to find the gold nuggets that's really valuable, and that's the data you want to integrate into your data warehouse, integrate with your transactional data and use that to make better business decisions.
And so the real key isn't that relational databases can't store this data. Relational databases can store all this data. The real key is that most of the data is worthless to the business. You don't want to store it. You want to only capture and store the really valuable information, and that's what you want to store in your high-performance relational databases.
Okay, so let's go to the next slide. And so why is -- why are people so excited about Big Data? Well, it's the next generation of -- a next evolution of our business intelligence information systems. And for all these systems, what's the goal? The goal is to grow our revenue, to grow the bottom line, cut expenses, et cetera. And so McKinsey has done some studies and on this slide, we just point out some of the data they point out, about how Big Data can be used to improve businesses, okay? And you can feel free to look at this study in more detail.
Okay. So in the next few slides, I'll just very briefly walk you through what we're doing at Oracle in our products to deal with Big Data. And what we like to do -- at the high level, what Big Data is all about, as I said, it's about -- it's an evolution of our information systems for business intelligence, right? And the goal is to make better business decisions, right? To grow our revenue, to cut our expenses, et cetera.
And we look at it as a life cycle of data. So the Big Data comes in through an acquisition phase, and then you want to, what we call, organize the data or sift through the data, looking for those gold nuggets of valuable information for the business. And then you want to load that data into your data warehouse. You want to analyze it and integrate it with your existing transactional data and then use that to make better business decisions. Okay. So those are the 4 main phases.
And let's go to the next slide. And on this next slide, which is a bit of an oversimplification, I just tried to position all the various software technologies that Oracle is deploying to provide a Big Data solution for our customers. And I've organized it according to these 4 phases: acquisition, organization, analysis and business decision.
And there's a little color-coding here. Red means basically software that Oracle has built and our existing Oracle products, and the gray are the open-source components we're getting from Apache Hadoop. And so why don't I just go through this a little bit from left to right. So the Big Data is coming in, and we show with these icons, from social networks, maybe, or from sensors and Twitter, et cetera, and we have an array of tools for capturing that information. What's new in the Big Data space is HDFS, Hadoop -- which is the Hadoop file system. It's a distributed file system, which some people are using for capturing large amounts of this Big Data.
Oracle is also providing what we call the Oracle NoSQL database, which people are using also in this acquisition phase sometimes. NoSQL databases are next-generation key-value stores. Key-value stores have been around for 40 years. You may remember them from your -- those of you who have been -- go back to the mainframe days where they have these things called Index Sequential Access Methods. They basically store a key in a value, and they let you look up by a key and get the value back. What's new there is these are now very scalable. Because instead of having one index, essentially, you can now have hundreds of indexes, and you do a hashing technology layer to distribute the data across the hundred key-value store indexes.
That technology, like I said, has been around a long time. There's an established market for it, it's not very big. Oracle is actually the biggest player in that market. We have a product called Berkeley DB, which is the leading key-value store, and we've now expanded Berkeley DB into what we now call the Oracle NoSQL database to have a distributed key-value store.
And then finally, of course, all the transactional data is coming in through enterprise applications like Oracle E-Business Suite, Siebel, PeopleSoft, SAP, et cetera. Lots of Oracle databases are under the covers there, gathering that information, and that's sort of where all the data is coming from in the Big Data world. Okay?
And then the data is flowing through to what we call the organization phase, and there is where Hadoop has a major role. So what is Hadoop? So Hadoop consists of this distributed file system that we mentioned earlier, plus it has what's called a MapReduce engine, which is just a Java development platform for building parallel applications.
So Hadoop is an interesting technology. It's an open-source technology provided by the Apache Foundation. We work with Cloudera at Oracle. Cloudera has a distribution of Hadoop that we are using in our Big Data Appliance, that I'll talk about on the next slide. And like a lot of new technologies, there's a lot of hype around it. But it's got a lot of maturity problems, and I'll just go through a couple of the issues around Hadoop.
Number one, it is a development platform for very sophisticated Java developers to build parallel applications. And so one of the big problems around Hadoop is a skill-set problem. I talk to customers all the time, and they just don't have developers who know how to write these MapReduce programs. And so one of the big challenges of Hadoop is sort of to raise the level of discourse around Hadoop so you don't have to have rocket-scientist Java parallel programming developers, but you can code at a higher level. And one of the interesting things that people are looking at now is like, "Well, how do you do that?" Well, in the relational database world, there's this language called SQL, or SQL, that we all provide.
Well, in the Hadoop world, they're saying, "Hey, that SQL thing is actually pretty cool." And if we provided SQL on top of Hadoop, we could open this up to lots more people that they could actually write queries and reports on Hadoop. And so in the Hadoop world, there's a project called Hive, which is actually a very simple SQL engine. And I sort of agree with them. If Hadoop is going to be successful, it's probably going to have to have SQL. And then pretty soon, it's going to look like an analytic relational database. And that's very interesting, future for them to go, but I guess we all already know that market. That's called relational database market.
The other issues around Hadoop are just the typical stuff for a new platform. It doesn't have very good technology for security. It doesn't have very good technology for disaster recovery. Of course, relational databases do all that kind of stuff already really well. And so it's got some growing up to do in that phase.
And then let's talk about the red boxes in this column. So this is where Oracle is trying to add some value to the Hadoop world. And I'll just briefly talk about a few of these things. Oracle Data Integrator is a code generator that Oracle uses today as an ETL tool for data warehousing. It is able to generate MapReduce programs automatically for customers, based on data flows and transformations using a graphical user interface.
So Oracle Data Integrator is another way of raising the level of discourse for developers to not require incredibly sophisticated MapReduce programming skills. That's part of our offering in this Big Data space.
We also have a loader for Hadoop. So once you get the data you want and you want to load it at a high performance into an Oracle data warehouse, we have the Oracle loader for Hadoop, which is a MapReduce program that my developers have written to very rapidly move the valuable gems of information you find in the Hadoop world into an Oracle data warehouse at very high performance.
And finally, the last thing is something we call Oracle Direct Connect for HDFS. And this is actually very cool. What we're doing here is we're saying Oracle has a really good SQL engine. And the people in the Hadoop space really want SQL. And so what we are able to do with Oracle is use our technology for what we call external tables, to basically let the Oracle SQL engine actually reference data in the Hadoop file systems, pull it in via standard SQL programming constructs, query constructs, and let our developers write SQL that actually spans across data that's in the Hadoop file system and in the Oracle relational database, join it together, sort it, do all kinds of analytics, and we think it's something a lot of our customers are going to be interested in.
Okay. So let's move to the next column, the analysis phase. And here is what I talked about before. We found those gems of data using our Hadoop MapReduce engine. We want to load them into the Oracle database at a really high performance. We want to integrate it with our transactional data, enrich that transactional data, and then we want to analyze it. And, of course, Oracle has great tools for doing that sort of thing. And the one thing I just want to call out is in the Oracle database, we have very powerful in-database analytic capabilities. So there's not only SQL, which is the obvious analytic technology that comes with our relational databases. Oracle has also recently integrated R, which is a very powerful statistical and mathematical analytics development environment. We've integrated R into the database. It runs in-database and does very high-performance analytics on the Big Data in an Oracle database.
We also, of course, have all the predictive analytics data-mining algorithms that customers are looking for. Plus, we actually have very powerful spatial and other unstructured data analytics technologies built into the database. And so all this talk about relational databases aren't good for unstructured data is all nonsense. 15 years ago, relational databases became object-relational databases, and the whole objective of that was to make relational databases good stores not just for structured data but also for unstructured data. So if you look at typical relational databases today, it's very common if more than 1/2 the data in those databases are unstructured.
You go out to banks. Banks have scanned all the checks into their relational databases. The IRS stores all the tax returns in relational databases where it -- relational databases are loaded with spatial geo-coded information, text documents, XML documents, everything you can imagine. So the relational databases are very powerful platforms for storing and analyzing all kinds of data, both structured and unstructured.
And then finally, as we get -- as we want to analyze the data, Oracle and lots of other vendors have query and reporting tools. We have analytical applications, various sorts, data discovery tools. And basically, all that stuff, it will be used and employed against the Big Data.
Okay. So let's go to the next slide. One of the unique things Oracle has been doing over the last few years is, since we bought Sun, is we have been building what we call engineered systems. These are systems that combine our hardware and our software together to deliver very unique capabilities for customers as far as incredibly high performance, really good time-to-value, which means the customer buys the system, they roll it into their shop, and they can be up and running in a few days, kind of thing. So these are proven to be very popular. And in the Big Data space, I just wanted to make sure everybody understands, we have a complete set of engineered systems. We're helping customers very rapidly deploy Big Data solutions, and we have everything from the Big Data Appliance that I'll talk about a little bit more in our next slide. And Exadata, of course, which all of you, I'm sure, are very familiar with, which is our very high-performance platform for running all kinds of workloads, including Big Data warehouses at really high performance.
And then, finally, Exalytics, which is our latest engineered system for doing business intelligence and for running Oracle's very powerful business intelligence tools like BIE [ph] and our new Oracle Endeca data discovery technology. And we'll be having a separate briefing on that technology, so I won't go into that in any more depth.
Okay. So let's go to the next slide and drill down a little bit on the Big Data Appliance, because that's the main topic for today. So what is that? So the Big Data Appliance is our engineered system for running Hadoop and the Oracle NoSQL database at very high performance and very good time-to-value. So if you're an Oracle customer today and your business wants to get into the Hadoop space, you can just call us up, and we will give you an -- sell you a rack, which is the Big Data Appliance, a rack which is essentially an 18-node Hadoop cluster-in-a-box. It's got lots of cores of memory, lots of -- lots of cores, lots of memory, lots of storage. And if you look at the price per terabyte of this box, we are a very aggressively priced technology. It's not a premium-priced product like some people seem to think. It's a very competitively priced product, and our price per terabyte is very good.
We're using standard, off-the-shelf Sun commodity Intel servers. And the key thing in the Hadoop world is people don't want just 18-node clusters, some people want to ride in 100-node clusters or 200-node clusters. And this Big Data Appliance is a building block. You start with one and if you need more, you buy another one and you connect it over the InfiniBand interconnect that we use here, and you scale out. And it's very easy to scale out, to build out very large clusters running Hadoop or the Oracle NoSQL database. One of the big benefits, of course, is that if you buy this from Oracle, we give you a single source of support for all the software, all the hardware that you use here, and it's something that a lot of our customers are very interested in.
And what's the business value? It's what we mentioned before. You can use the Hadoop engine, NoSQL engine to gather information, Big Data information, to sift through that information, to find the gold nuggets. I mentioned earlier, we have now integrated R with the Oracle Database to provide in-database analytics using R. Well, R is also available on our Big Data plans for doing analytics in the Hadoop space. You can integrate R with your Hadoop parallel MapReduce programs.
And then finally, one of the big values of this technology is to integrate with our existing data warehouses. So those of you who have heard about Exadata know that over the next few years, every Oracle customer who has a Big Data warehouse or data mart is going to be moving to Exadata. The benefits are just incredibly compelling from a price/performance standpoint.
So one of the big focuses of the Big Data Appliance is to be a very good companion with an Exadata data warehouse. So we use the same InfiniBand technology we use in Exadata to be the interconnect for the Big Data Appliance and to make it very easy to connect the Big Data Appliance to an Exadata data warehouse at very high performance. And I mentioned earlier, we have this loader for Hadoop technology. One rack of a Big Data Appliance can load 15 terabytes per hour of data into the Oracle Exadata data warehouse over this InfiniBand interconnect. We talked earlier, we have very -- we have this direct connect for HDFS capability that lets the Oracle SQL engine reach out into the HDFS file system and analyze Big Data. And we talked about the Oracle Data Integrator already as well.
And just to close off the discussion, I just want to talk a little bit about what our customers are up to. So we're talking to a lot of customers about what they'd like to do in this new world of Big Data information systems. And I just thought I'd mention 3 of the customers and what they're doing and -- just to give you some insight into what's going on out there.
So the customers are in 3 industries: insurance, travel and games. So insurance as an industry we all understand pretty well. We all have car insurance. And this particular customer already has an Exadata data warehouse, and they're already capturing, of course, all their transactional insurance information about their customers, their accidents, their policy information, et cetera. And what they would like to do is enhance that data with a new kind of data that you can get from cars. So cars are now loaded with sensors that are capturing your every movement of what's going on out there and it's called car telematics data. And what they would like to do is use that information to actually study the actual driving behavior of their customers and use that to better understand maybe what their insurance rates should be, what their driving habits are and maybe even help customers be better drivers. And this is actually a very classic-use case. So they're interested in augmenting their Exadata system with a BDA, Big Data Appliance. They're very interested in our R technology as well. Okay?
Next customer is in the travel industry. They run websites for customers who are looking at doing travel of various sorts. Today, of course, they're already capturing all the transactional data about their customers, what are the trips they're buying. And what they would like to do is augment that information with what's going on in their websites. They want to capture the web logs, they want to get social media data to better understand what their customers are up to, what are the trips they're anticipating maybe going on and combine that information with their existing information about their customers' previous transactions and use that to help make better promotional offers to the customers and grow their business.
And then the last customer is in this game space. So gaming is becoming a huge industry, as you all know. And this company is in the business of selling game consoles of various sorts and Internet games, and they already have a big Exadata data warehouse already that's analyzing that information. And they would -- they're looking to augmenting their Exadata data warehouse with a BDA, and they want to use it to what you might expect: better understand what the customers are doing out in the games. They want to understand relationships between customers. One of the really interesting things in games is that people play games with each other. And you want to understand the social networks of people who are playing with each other because it's likely that if one person in that network wants to do something, the others will want to do the same thing. So they can use that information to better upsell information in this game space.
So that's 3 good examples. And I can tell you, just talking to customers, there's a lot of interest in this space. And it's, as I mentioned earlier, it's a real evolutionary technology. Customers have their existing big BI information systems, their data warehouses, their data marts, and they're really excited about augmenting their transactional data with this Big Data to help them grow their business.
So we'll just conclude on the last slide, just to summarize what we're up to. Oracle is very unique in this space. We not only have a complete set of software for dealing with Big Data information problems, we also have a complete set of engineered systems as well so customers can very rapidly roll out for their businesses complete deployments of Big Data hardware and software solutions.
As I mentioned earlier, our technology deals with both structured and unstructured data. We deal with SQL and, of course, we also have NoSQL technologies as well. And we provide -- because of this unique hardware/software combination, we provide the fastest time-to-value for customers. If you're an IT guy and you want to roll out a Big Data system, you can just call up Oracle and you can deploy the system incredibly fast by rolling out our engineered systems. And of course, Oracle will provide single-vendor support across all the hardware and software that you're using in your Big Data environment.
And with that, I think we'll move into a question phase.
Right. Thank you, Andy. Before we go over to Kash, I just want to remind everybody that you can ask a question online by clicking the tab, Ask a Question. With that, Kash, please go ahead and start the Q&A.
Kash G. Rangan - BofA Merrill Lynch, Research Division
Sure thing. But first all, thank you, Andy, Paul and Ken, who may be on the line, for giving us a chance to come in and spend some time with you guys. I think a year back, Andy, we spoke about the database market itself. So it looks like -- listening to you more and more, it feels like this Big Data evolution reinforces the core of what Oracle does really well and adds more new opportunities around it. I think there's a lot of confusion and misperception as to what exactly Big Data is, so thanks for doing this webcast with us. And for investors on the line, we published a report today on Oracle and how it fits into Big Data and what Oracle is exactly doing on the Big Data side. So with that, one item I wanted to get your input on is Cloudera. Why did you choose to work with Cloudera? I think there are certain other distributions out there in the marketplace, just your thoughts on that?
Well, we're very -- in the open-source space, we're very interested in providing technologies that sort of conform with the open space ethos. Just like in the Linux space, we have the standard Linux distribution based on open source. It's not proprietary. And the thing we liked about Cloudera versus some of the other distributions out there is they are not a proprietary distribution. If you store all your data using the Cloudera distribution for Hadoop, you're using the standard HDFS file system. And if at some point in the future you decide, "Well, Cloudera was nice, but I want to use somebody else," it will be easy for you to move because you're not locked into some proprietary technology. So we really liked Cloudera's openness. Also, we think their expertise in this space is really good. They have some of the leaders of the Hadoop engineering community, open-source community, working at that company. And so we thought they were -- they'd be a great partner for us to work with in this Hadoop space.
Kash G. Rangan - BofA Merrill Lynch, Research Division
Got it, got it. And the second thing I want to touch up on is Hadoop and NoSQL databases have been designed for running on large commodity clusters, with the horizontal scalability being a key emphasis there. I'm wondering, when we think about the Big Data Appliance, what is the differentiating value proposition for the Appliance? It looks like there is a lot of activity around Oracle, with other competitors are doing similar initiatives. What is the right way to think about your differentiation here?
Well, we understand that the Hadoop community of customers -- or one of the things they like about it is the fact they can use commodity servers to build these clusters at very good price/performance. And so when we built this system, we decided to just use standard off-the-shelf Intel servers for building out these Hadoop clusters just like everybody else. And we knew that the pricing has to be aggressive in this space. We can't charge some premium ridiculous pricing for something where people -- one of the attractions are is that it's low cost. And so if you look at our cost per terabyte of this technology, it's very well priced. We think it will be very attractive to our customers. And then we add a lot of value on top of that. Oracle has end-to-end support for all the software and hardware components here, which our enterprise customers really appreciate, and nobody else is doing that. And then one of the big features, of course, is our integration with other Oracle systems. I talked about Exadata; I talked about Exalytics. So our installed base customers who are using Oracle Database, they might be using Exadata and BIE [ph] and Exalytics. This is a natural fit for them to expand into the Big Data space from their existing BI system. And then finally, I mentioned we have a bunch of software that my group is building. We're building MapReduce software for our customers. We have the Data Integrator capability that does code generation for Hadoop, that makes the development of Hadoop MapReduce programs easier. We also have our SQL engine extended so they can reach out into the Hadoop space again to make Hadoop more accessible to the masses who know SQL but don't know how to do MapReduce programming. And we have R integrating into our Hadoop offering, which is another big plus for customers to want to do analytics in the Hadoop space. So we think we have a very interesting offering. It's got both a hardware side and a software side and a support side that I think is very -- going to be very attractive to our customers.
Kash G. Rangan - BofA Merrill Lynch, Research Division
Got it. And I think some of the use cases you talked about, the 3 examples, through your point -- to the example of how Big Data Appliance ties into your core business and ties into the use case of Exadata, data warehouse, database implementations, so we're starting to see that, although you announced that the product itself, just at the user conference 6 months back, it's interesting to see how quickly the use cases are operating in the marketplace.
Kash G. Rangan - BofA Merrill Lynch, Research Division
I wanted to just drill into unstructured data a little bit, because it's a relatively less-understood topic. There's been a lot of activity in this space. What are some of the key problems that you think Oracle can address with unstructured data? What are the future areas of innovation in this market that Oracle can address?
Yes, and I already talked about this a little bit earlier. We started -- the classical relational database model that was invented 30 years ago is what, somehow, people seem to think it still is. The classical relational databases stored numbers and dates and strings in rows and columns, and that's all they did. Right? But about 15 years ago, there was a big revolution in relational database technology. It was called object-relational databases. And the whole objective of that was to extend relational databases to store all kinds of data, not just rows and columns of numbers and strings. And as I mentioned earlier, relational databases has -- have now been extended, and the technology is very mature. You can store what we call LOBs or files in a relational database, and you can read and write those files faster than you can through file systems, believe it or not. Relational databases are really, really good, very highly optimized for dealing with unstructured data, and that's just the raw unstructured data. On top of that, we have a lot of value-add. We have text indexing in the database so you can do unstructured keyword text searches using SQL. We have integrated products on top of that. We have a solution called Secure Enterprise Search that actually lets you use a relational databases as a search engine. We have spatial technologies so you can ask questions in relational databases using geo -- about coordinates. You can say, "Show me all the stores within 5 miles of this location." So cell phone GPS information is all over relational databases. And so there's a lot of value added -- we have text mining, actually, algorithms in the database. You can look at all your documents. You can classify them. You can do all kinds of stuff. We have data mining algorithms as well. So relational databases are very enriched structures for storing structured and unstructured data and for analyzing structured and structured -- unstructured data. So that's just the Oracle Database side. Oracle also has other technologies, and the most recent one is a product we actually bought, an acquisition called Endeca. We have the Oracle Endeca Information Discovery tool. This tool is also a very flexible technology that can analyze both structured and unstructured data. It's a combination of sort of a traditional search -- keyword search-based technology and BI, sort of "look at the data and drill down into it" technology that's very unique in the industry. It's a big part of our Big Data offering, and it actually is being optimized to run on our Exalytics platform that we announced and launched just -- very recently. So that's another big offering into that space. And then Oracle is very big into analytics. We have vertical industry analytics, we have horizontal analytics through our BI apps products. And in this whole space, we're very focused on all the data customers want, the structured and unstructured, and we're very excited about this. It's a great opportunity for us as captures -- customers want to capture all kinds of information. We're an information company, we want to help them manage all their data whether it's structured or unstructured.
Kash G. Rangan - BofA Merrill Lynch, Research Division
All right. Next question is actually going to be the synergy between your core business and unstructured data, and I think you answered that very elegantly. So maybe move on to in-memory databases, a lot of talk, at least from Wall Street and perhaps the industry too. In-memory databases, what are your thoughts? Is it an opportunity or a threat for Oracle?
Yes. So in-memory databases are an interesting technology. And at some level, there are 2 things I want to get across. One is sort of going back to my car analogy. Cars have evolved a lot over the years, and sort of one of the things people have sort of figured out is that you can get really good performance at low cost with these things called turbochargers. You stick them on the engine, the engine runs a lot faster. And everybody -- they're expensive so you don't put them in all the cars, but you put them in some of the cars. And you can think of in-memory database technology as the same sort of thing. It's another feature of relational database. They can turbocharge the performance of the database. But like a lot of things, it's not for everybody. So for example, there's an economic side to in-memory databases that has been true for many years, and it's going to continue to be true for the next 5, 10 years. Disk drives, if you look of the cost per terabyte, are about 100x cheaper than main memory. And then in between, there's a flash technology called Flash, which is about 10x the cost of main memory. And so what customers always are going to want to do is they're going to say, "Okay, I want to get the best price/performance out of my technology." And some customers -- most customers, especially as we move into the era of Big Data, are going to say, "Heck, I can store this data really cheap on hard disks, so I'm going to store all my data on hard disks." And this is the approach Oracle is taking in our Exadata platform. And then as that data gets warmer, when you start reading it, we transparently move that data into Flash. And then as it gets really hot, we want to move that data into memory, and that's where the in-memory database technology, in-memory column store technology that you hear a lot about, takes effect. And so all the relational database vendors are working on in-memory column store technology. It's a really nice feature. It's going to turbocharge databases. But pure in-memory databases where you say, "Okay, all that data in my database has to sit in memory?" That's really expensive. It's like saying in the car industry, everybody's going to buy Ferraris. And we all know everybody doesn't buy Ferraris. And why don't they? Because they're expensive. And it's the same thing in the database space. People are driven by price/performance. And inside a given company, some of those analysts are going to be given the Ferraris. They're going to get those in-memory databases where all the data is in memory, and other analysts are going to get the economy cars and the mid-sized cars. And so we think the right strategy, and this is a strategy we're doing at Oracle, is dealing with a memory hierarchy. And we think in-memory database -- pure in-memory database where every -- all the data fits in memory has a place there, just like Ferraris have a place in the market. But it's a market niche. It's not going to be everybody's going to be using that stuff. So we think the right approach is to build a product, like we've done with Exadata, that deals with all forms of memory from disk drives to Flash to main memory and has the right balance to give our customers the right price/performance that they want.
Kash G. Rangan - BofA Merrill Lynch, Research Division
Got it. One thing was -- there's some talk in the industry about the convergence of OLTP, the Online Transaction Processing market and the analytical market down to one platform that can handle both. What are your thoughts on that?
Well, that's a very interesting space because Oracle is very unique in that space. Today, if you buy our database machine, or any other Oracle database, we can run OLTP in that database. We can run data warehousing. We can do OLAPs. So Oracle, already today, in our current product, is the only product out there that can actually combine all these forms of processing in one database. If you look at what all the other vendors are doing, they are very specialized. So IBM says, "Yes, buy [indiscernible] for doing your Big Data warehouse. But for OLTP, buy that DB2 product over here." And Microsoft has the same thing. They have a parallel data warehouse product that's completely separate from their transaction processing product. Teradata only does data warehousing. They don't do transaction processing at all. So we completely agree with that vision, and we already do it today. We're very unique in that space. And we will continue evolving and enhancing our products to continue to do better at OLAP and OLTP and data warehouse, all on the same database. That's been our strategy for 20 years, and it's going to continue being our strategy in the future.
Kash G. Rangan - BofA Merrill Lynch, Research Division
So one other thing that I wanted to touch upon was just to -- we would like to appreciate how strong the entrenchment of the relational database is. There are some that mentioned -- that are talking about the risk of replacement, how easy it seems to be. I don't know how they get this vision, easy it is to replace a database and certify another application for another database. Can you talk about -- how do you -- how should one view this entrenchment of the relational database, and why is it so tough to unseat from within the industry?
Well, the relational database market is extremely competitive. And we are where we are as the market leader because we have the best technology, and that's -- and we are the most innovative company. And if we don't keep our guard up and keep innovating and bring out new technologies like enhanced column store databases, et cetera, we're going to lose our position there. And so -- I mean, it's -- there's a stickiness, as you sort of referred to when a customer is using Oracle, and they want us -- move to another database. There's cost doing that. And so yes, there's cost there. But the main reason people stay with Oracle is not just that it's expensive to move to another database, it's that we have great technology that helps them solve their business problems better than anybody else's database, and that's what this game is all about. Whoever has the best technology is going to win, and that's the game we've been playing for 30 years. And we're going to continue playing that game. And so if somebody wants to unseat us, they have to come out with a better product basically. And we love competing with other companies, that sort of gets our blood flowing. And so we're eager to take on all these new competitors in this market. And we've been doing -- as you saw from the Gartner Market Share data, we think we're doing a really good job of that and taking market share, and we hope to continue that moving forward.
Kash G. Rangan - BofA Merrill Lynch, Research Division
And our blood is flowing. As analysts, we love the fact that there's a lot more dynamics in an industry, there's more debate, et cetera. So back to you, Paul. I think we're...
Thanks, Kash. Let's squeeze one in here from the web before we conclude. Andy, one person is asking about -- he's referring to Mark's -- actually Mark Hurd's talk in OpenWorld in Japan recently. He's talking about the growth, explosive growth in data expected from now until 2020 and something like 20x growth. And this person's wondering about what your thoughts are regarding the contribution of Big Data to that growth.
Well, if you -- the interesting thing about growth of database is that if you go back to like the year 2000, it's only 12 years ago, people thought a terabyte was a really big database. That was humongous. And now, 12 years later, a petabyte, which is 1,000x bigger, is what people think is humongous. And so I have no doubt that what Mark's talking about, this whole exponential growth of data, is continuing. And of course, this is great for our business. The more data people have to manage, the more databases people want and the more interesting technologies we can come out with to help them manage that data. So that -- there's no doubt that this is a trend. There's been this exponential growth in the size of databases, and it's going to continue. And it's great for our business, and we're -- it's a really great challenge for my developers to come up with technology to help manage these huge databases.
Thank you, Andy. That's a great way to end up. Big Data is good for Oracle. And with that, we like to thank you all for joining us today on the web. Also, a special thank you to Kash for leading the Q&A portion of the call and for asking the questions that investors most typically want to hear about Oracle and Big Data. So this concludes our call here. If anybody has any additional questions, please contact Oracle Investor Relations. Thank you for joining us today.