newsfilter
← Latest News
Earnings Call Transcripts·

Oracle & IBM Power Present: Modernize for AI resilience and security while optimizing costs and existing infrastructure - ILOG SA

Hello, everyone. Thank you so much for joining IBM webinar. Before we get started, I'd like to go over some housekeeping items. At the bottom of your screen, you should see a Q&A box where you can start submitting your questions for our presenters today. Later today, you will receive a link to the on-demand recording and a link to the slide deck. So without further ado, I will pass it over to Satya. Hello. IBM and Oracle together are super excited to tell our story of how we are collaborating with each other to deliver Oracle AI database 26 AI on Power 11 and AIX. My name is Satya Sharma. I'm an IBM Fellow and CTO in IBM Power. And my co-presenter here is Henry Bioram, VP of Product Management at Oracle, my friend and collaborator. Together, we are going to tell our story to you. So these are the different topics. I'll start with how Power 11 is the right platform for Oracle AI Database. Henry will go over Oracle 26 AI database capabilities and update. Then we have two demos planned. First demo is about how you start with the enterprise data and lead to AI power decisions. Second demo is how you secure in the quantum era Oracle 26 Ai on IBM power. Now let me start by saying that power is is the trusted choice for IT leaders running Oracle database. Together, we have 40 years of collaboration jointly. There are thousands of mutual customers who trust IBM Power with Oracle. And now we are in the 11th generation. We are at Power 11. I also want to highlight from my personal point of view, I've been working with Oracle since 2008. So I have a long history with this collaborative relationship with Oracle. Now, in the AI era, there are two different elements. First, AI speed requires that enterprises deliver faster outcomes, maintain high quality, and reduce cost. At the same time, as enterprises need to move with speed, you can create potential systematic risks. For example, you can create vulnerabilities or it can result in bad AI decisions, which can result in bad customer experience or even damage to your reputation. Now, clearly, software, which is Oracle and AIX, they play a key role in addressing these two elements, but the underlying hardware platform also plays a central role in enabling you to move with speed and mitigating your systematic risks. And we think that IBM Power 11 is absolutely the right platform for the agentic AIT world for your enterprise. Why do I say that? A power platform, starting from the silicon technology, which is processor to hardware systems to operating system firmware, and even hybrid cloud platform, all of these different layers are developed by IBM. And they are all integrated in a cohesive platform that cannot be said of any of our competitors. If you take our competitors, different layers are supplied by different vendors, and then they are cobbled together. This integrated platform results in great outcomes. I'm going to take three examples of these outcomes. First, 15x faster detection and resolution of capacity. What I mean by that is that we monitor capacity situation, and if additional resources need to be applied for more CPUs or memory, those actions are taken 15x foster as a result of agentic AI compared to traditional methods. Second proof point here is that for those developers who are new to the platform, the learning curve is 60% faster compared to traditional methods. And the third, for those customers who are making use of IBM Spire adapter, which is IBM accelerator adapter, in business-to-business sales recommendation, success rate is 98%. now what are the key elements to tackle ai speed i would say that there are three elements in terms of speed and uninterrupted operations first element here is unplanned downtime with power 11 we have explicitly stated that we have six nines of availability or uptime that translates to 31 seconds of outage per year. Second element here is we recognize that maintaining these systems from the patching point of view can be a painful task. So we have automated the entire system maintenance operations from hardware to firmware to hypervisor. And there is not an instance where customers need to plan for downtime to apply maintenance tasks. In many cases, these tasks are concurrent, which does not require reboot, whether it's hardware adapters or those sorts of things. Same thing from the firmware point of view. On occasion that it does require disruption, we use techniques like live partition mobility. So from customer planning point of you, there is never a time when you need to plan for downtime to take care of system maintenance. Third, from the cyber incidents point of view, in Power 11, we have a new capability called IBM Power Cyber Vault. This capability, if there is a ransomware event of some sort, our response time is less than a minute. So we can detect these events in less than A minute. And then we also provide the safeguarded copy so that we can recover in very little time. Now, to talk about the autonomous operation on IBM Power, I'm going to start with these four different classes of operations. From the build point of view, the entire process of building the infrastructure is policy-driven and entirely automated. This will eliminate any drift and it will eliminate any manual effort. Second, as the agents proliferate, we recognize that access controls becomes key with the scale. Third, we constantly monitor any vulnerability issues and, in fact, apply automatic remediation. So there is no action required on the IT administrator's part. And the fourth, like I talked on the previous slide, 15x faster detection and resolution of capacity constraints. So together, all of this delivers a great agentic AI platform. Now, with that, I'm going to hand over for a demo. It's about IBM Automatic Operations. you you you you you Okay, coming back to the topic. So what are the key elements of autonomous operations in IBM Power? First, these are native agents. We are not using some generic AI capabilities. These are purpose-built agents. Second, it's extensible by design. What I mean by that is we are making use of MCP servers, API, many other methods. which encompass infrastructure as well as application. You also recognize that from taking action point of view, humans want to be in the loop. So before we make the changes, we present options and the humans are able to approve these changes. So this accelerates issue resolution, improves availability, simplifies operations, automates operations, and so on and so forth. Now, I'm going to switch a little bit into Oracle specifics on IBM Power. What you see here on this slide is our history together since 2007. Oracle 11G, 12C, 18. So you can see that this is from 2007 up and last 20 years of history. I hope most of you are running 19C today on AIX 7.2 or 7.3. But what I'm happy to say today, as you'll see in the demo, is that 26 AI database on IBM Power on AIX 7273 is coming soon. I can personally vouch for the progress. I'm deeply engaged in the development project on sort of weekly and even daily basis. And I can say that we have made tremendous progress together from 26 AI development on IBM Power and AIX, and this capability will be made available to our joint users soon. Now, what are the different ways in which we are collaborating? First, there is executive engagement and resource allocation from feature prioritization point of view, day-to-day execution point of vue. We have added new human resources. We have additional power hardware for build and test purposes. We have replaced the previous generation compiler, XLC compiler, with the new open XLC compiler, which will improve the long-term sustainability of Oracle 26 AI on Power. We also have SIMD capabilities and MMA vector for AI workloads. We also recognize that you need compatibility with the applications which were built using previous compiler so we can mix and match old applications with the previous compiler and new compiler with 26 AI database. Third, we do joint testing. Oracle does testing of the product, and at the same time, IBM does testing of the project. And in fact, just to give you one area, in case of RAC, IBM Does considerable testing on an ongoing basis. IBM also has designed new test cases to verify 26 AI AI features. Fourth here, lastly, increasingly, we make use of open source products or packages, I should say. So in case of 26 AI, for example, OpenSSL and the ONNX runtime, so IBM is contributing to these projects as power optimized encryption and AI runtime. From a checklist point of view, here's a checklist of how to get ready for 26 AI, become current on 19c, become current on AI 72TL5 or 73, improve security posture by applying the right CVEs, upgrade to Power 11, get ready for how you are going to migrate to 26 AI and so on and so forth. So there is, in fact, the project is far enough along that we are now able to devise a checklist. I'm going to switch, you know, to a security topic. I would say three security issues in the industry today. First, AI systems, the vulnerabilities are being detected at far faster pace. And these can be in the source, some cases even in the binary. And when you chain multiple of these issues, the attack impact can be much larger. I'm sure you all are reading that many of these AI systems are also bypassing guardrails. Second is there's a risk to cryptography from quantum attacks. I'm sûr you are all reading about it. These asymmetric cryptography operations like RSA, ECC, they have the greatest quantum risk. and harvest now, decrypt later, whereby the bad actors could store the data today. And when the quantum gets ready a few years down the line, they can decrypter the data. That class of attacks is becoming a growing concern. As a result, there is increasing focus from the regulation point of view on PQC, post-quantum cryptography, starting critical infrastructure and government procurement. There's also now cyber resiliency regulation, for example, DORA. And there's a growing focus in many parts of the world around data sovereignty. We think that IBM Power Systems are best positioned to address these security challenges. First, we have always had secure by design in IBM Power. But now we are extending that to stay ahead of AI-based threats. First, we are benefiting from the Glasswing Mythos project. Second, the AI vulnerability testing is becoming part of the development lifecycle. There are even compiler options to detect against memory safety issues. From a PQC point of view, entire Power 11 stack from the hardware platform to firmware, This entire stack is now protected with the post-quantum, with the PQC class of algorithms and even extends to operating system and management tools. From the regulatory requirement point of view, now with IBM Power Vault, there is end-to-end cyber resiliency. And then from the autonomous operations standpoint, any security fixes that need to be applied, they are detected and applied without any overhead, without any work on your part. So with that, I will conclude my talk, and I'll hand over to my friend, Henry, to take you through the rest of the story. Great. Thank you, Satya. Thank you for the rundown and highlighting our collaboration. I'm going to spend a few minutes here talking about some of the newest features within the Oracle Database. Oracle DatabASE 26 AI is a major release for us. It's years in the making. So I'm willing to spend, like I said, a few moments walking down some of these new features and talking about kind of the intersection points between our database and IBM's power systems. I'm going to talk a little bit about some forward-looking statements. In fact, some of the stuff we've talked about today in terms of release dates is forward-thinking, but we do want to share this news because it is imminent, right? We are going to be having this release relatively soon. So 26 AI, as I said, is a major release for us. It's designed for the AI era. That means there are a lot of AI-native capabilities, but we have also focused on security, a similar focus on our part. Our release cycles typically are fairly synchronous, but one of the big impacts for us this past couple of years was the release of these foundation models that can detect bugs at an enormous rate and in depth that we've never seen before. So a large part of our engineering efforts are not so much focused on feature function, but on really going in and understanding where we have these vulnerabilities and patching them and making sure that the releases we put out have been hardened against these vulnerabilities. As I said, it's been a major release for us. There's over 300 new capabilities within 26 AI across the board, both in HA security, AI, and a number of other areas. I'm really going to focus kind of on four major areas. That's our adoption of AI in the database, our ongoing work to make sure that Oracle Database can support any workload at any data type, Some new capabilities on data access and integration, which are extremely important to have a rich AI solution. And then finally, but certainly not least, we'll talk a little bit about security as well. So new capabilities around AI. Some of these we've had for a little while, like AI vector search. Vectors are a native data type within the database. But in terms of other capabilities, we've implemented a number of new features, things like natural language to SQL translation, annotations to do data description, so AI is better informed when it generates code or SQL. We've implemented an MCP server. So again, as Satya said, we we've embraced a lot of these open connected connectivity standards. And then, more importantly, I think, for our customers, we've embraced agents and the agentic wave here. I'll talk a little bit about what that looks like. So agents now actually can run natively within the Oracle database. We can build agents either externally to access data in the database, but we also let them – we've enabled the ability to run in the database. And there are a couple of ways to build those. we have a product called Private Agent Factory, and also Select AI, which is attached to our cloud services. But what's important to know here is that building agents, whether externally or internally, as we've been talking about, guardrails are important. But one of the best things you can do to enforce guardrailes actually is to really constrain the agent itself. And we do that by building it in a way where it's essentially an API call into the database, where you can be sure that the SQL that's generated is exactly what you want, that the data it's accessing is secure, that it's not generating arbitrary SQL. So for us, agents are important, especially data-driven agents that process data in a predictable manner. To that end, we also include a number of very specific agents that we've built for the database. We have agents that are focused on database operations, so database agents that help you manage the database, triage issues in a similar way that Satya demonstrated for the power system, where you're getting diagnostic information and your DBAs can be assisted and resolve problems faster, but also agents that aren't allowing you to actually build solutions, agents that can analyze your unstructured data and build RAG solutions, knowledge agents that can actually look at your structured data, find insights, find patterns so that you are actually getting time to value much faster. You don't have to build customized agents. We actually give you a number of those out of the box. So those are some of the features we have today in terms of AI. Again, all packaged with the database. Also, these are included at no extra charge, these capabilities. So if you have a database license, you can immediately, when you implement 26 AI, you Can start leveraging some of these features. The next area I wanted to focus on is our converged story, our converge platform. We've been working for 30 plus years on a converged platform. If you only think of Oracle database as your transactional relational backbone for your ERP system, you're probably thinking of Oracle Database 10G. As Satya laid out those successive releases, in every one of those, we've included new capabilities. If you look at Oracle Databases today, even 19c, there's support for JSON. In 26AI, we introduced vectors. So really, there's no workload that you can't run on an Oracle database today. We support all of these processing capabilities and deployment methods and workloads. We've introduced a couple of really fantastic features here in 26 AI. One is something called JSON duality and the other is property graph. So you can take your relational database and you can immediately build graph representations and query graph relationships on top of your relational data, as well as support JSON. You can turn your relational data into objects and documents that your developers can then use in building their applications. So whether your developers are comfortable with tables or objects in terms of their programming methodology, you can support that directly within Oracle Database. And all of this is accessed via SQL. So your developers don't need to learn a new language. They don't need to run a whole new syntax. If they're comfortable with SQL, they can query both graph, vectors, relational, any of our object types or data types within the database. Also, with this release, we've introduced a Mongo API. That means you can actually move data, document data into an Oracle database and query it with a Mongo application. We've seen a lot of uptake with customers in this regard. They're trying to consolidate their data platforms. Data sprawl is a problem. When they're building AI solutions, they want to have access to all of these different data types. Agents can run different workflows or work paths all on the same database accessing this data. So a couple of new really cool features within Oracle 26 AI that I urge you to check out, supporting all of the new kind of ways to program and process data. I mentioned data access. So as things evolve in the marketplace, there are some new ways to kind of store data that have popped up over the last couple of years. And one of the biggest trends is the use of object stores and Iceberg as a storage format. So Iceberg is an open standard storage format that allows really any tool out there that supports the format to access that data and process it. So it's a way to kind of separate the storage of data from the processing of data. And again, it's an open standard. With 26 AI, we now support the access and processing of iceberg data. So we can directly query iceberg tables. We can process those in an Oracle database. Not only read, but we can actually write back to object storage. So this is a new capability, very important in data lakes and accessing the vast amounts of data that are being pushed into object stores. So that is a New Capability in 26 AI that allows you to really run the Oracle database anywhere and access data anywhere. We don't only read data kind of in object stores or within Oracle. we now have some ways to actually access data in any system, not just an Oracle system. We can access SQL Server. We can assess Teradata. We can ask us pretty much data anywhere with an Oracle database. We have a capability called a federated query, which actually lets us read data in place from almost any system in your enterprise. So, you are no longer forced to kind of build ETL processes and push data into a specific location with Oracle. We can process the Oracle database information alongside information in your other systems of record without ever having to move it. So this is a very powerful capability, the federated query capability that we've introduced. So now you can build AI solutions across multiple systems of record, do real-time agentic flows without having to move data into a single location. And finally, I'd like to touch on security before we get into the demos. So security is important to us. Again, we've been using AI within Oracle to help our developers, not just in terms of productivity. So generating code, doing code coverage analysis, building regression test suites, all with the assistance of AI. We've adopted it wholeheartedly within Oracle. But as I said, with the release of these foundation models, every software vendor has had to go back and reassess what their security posture is, what hidden bugs they may not have found in the past that have now come to the fore. If you read the trades or the news, you probably heard that the recent discovery of vulnerabilities across probably 70% of the world's websites has just kind of come out and they're aggressively trying to patch it. That was all found with the use of these foundational models. So we live in a very different world. Now, Oracle has always taken security very seriously. It's in our roots. It's în our DNA, you might say. And so this is an area where we constantly innovate. Now, for us, we've got a couple of new capabilities here with 206 AI. The first is what we call deep data security. This is a way to allow you to apply all of the different security policies, roles, privileges, cell, column, row-based security, all within the database. We think this is incredibly important in the new world where agents operate almost autonomously. You can't be sure that guardrails are going to stop an agent from generating SQL that is arbitrary or accesses data that it shouldn't. So one of the new features you'll hear about at AI World here in Vegas, next week, is a lot of information around deep data security in the world of AI. We've also had ZDLRA, our backup and recovery environment, now with immutable backups to protect against ransomware. And then where I want to spend a little bit of time, as I'll talk again about quantum-resistant encryption, but we've also incorporated a SQL firewall directly into the database with this release. So you have effectively a last line of defense within the Oracle Database itself against SQL injection attacks. Satya touched on this issue of harvest now, decrypt later. Quantum computing is showing a lot of promise, and it has people very nervous. And so there's been a lot effort here to address the potential for quantum computing to crack today's encryption. And what we mean when we say harvest now, decrypt later is hackers will use any vulnerability attack to just gather the data, even if they can't read it, right? They're going to gather the beta. They're gonna hold on to it until quantum encryption lets them decrypted at some later date. What's important for us to do as an enterprise and as a provider of enterprise products and platforms is to provide that level of encryption, the newest encryption libraries, so we delay the amount of time it takes should quantum computing get to the point where it can craft these algorithms. So Satya said, we jointly over the last year, maybe even year and a half, spent a lot of effort adopting the latest libraries, moving to open SSL from the prior provider. That's been something we've tested extensively. And by extending this window with stronger encryption, the value of the stolen data decreases, right? Data has a lifetime value. Certain pieces of data obviously are always valuable, such as social security numbers. But financial information ages. So the more time we can delay that period of when quantum encryption decryption might be applied, the better. So we have adopted the latest libraries directly into Oracle as well. As I mentioned, we have an in-database SQL firewall so that things like SQL ejection attacks can be addressed directly at the source of the data. If a firewall in the enterprise is bypassed, we have now kind of this last line of defense we can raise. CDLRA is our product that carries kind of your backup and recovery infrastructure. Again, we've implemented immutable air-gapped backups. These air-gapped backups mean that should your primary database be attacked and locked down by a bad actor, You can wipe that and restore it from a known good backup, and you can verify that the backups themselves are immutable. They have not been compromised as well. And our AirGap technology ensures that they really don't have an opportunity to attack the actual stored backups. So wrapping it up here from my side, I've talked about four major areas of investment with the Oracle database, starting with AI designed for data. A lot of capabilities now available, things like building agents. We've got free tooling to help you do that, run them inside, outside the database, natural language to SQL, support for MCP, whole range of capabilities. Support for new data types and ways to process data so that you don't have to spin up separate data platforms to do every kind of point solution processing that you might be interested in. You can take your relational data and apply these new processing techniques directly to it. We have ways to access external data from the database. Iceberg is a major storage format that we now support, as well as federated queries. And finally, I covered security in detail, always a primary focus for us. So with that, I'm going to wrap my part of the presentation up. We're now going to move to a couple of demos that show the partnership, the deep partnership that we share between Oracle and IBM and really give you a preview of some of the really cool things you can do with our joint solution. Thank you. you you you you you you Thank you. you you you you you you you You Henry, you are coming back. Henry, are you coming back? Yes. Okay. Yeah, you should be able to hear me now. Thanks. Sorry about that. All right. So we've, thank you everybody. We've run through both of our presentations here. Hopefully it's been informative. You've gathered some information to take the next steps and some actions. We're going to answer some questions here in a second. But first, I want to kind of go over a couple of calls to action we have coming up. So Satya, did you want to go over the IBM ones? Absolutely. So we are organized to participate in planning calls. There are these, you know, two different links. You can request a planning call around 26 AI on IBM Power, as well as a validation program. And if you want to learn more about Oracle workloads on IBM power, there's a link for that as well, which is number three here. With that, Henry, yeah. Yeah, thank you. I mentioned in my presentation, we have our big conference coming up next week, Oracle AI World in Las Vegas. Hopefully a number of you will be there. Certainly IBM has a massive presence as one of our major partners. We will keep our MOS note updated. So you see a link there and we encourage you to follow that. That's where we publish our schedules, our release dates, our platform support. So you can always get the latest and greatest off the website there. So hope to see you at AI World and answer any questions you may have in the chat. We've been keeping up, hopefully, with the questions that have been fielded. But if there are any other questions that we can answer, we'd love to. Yes, so we have been answering questions continuously. And, you know, we'll keep answering those questions. But maybe there are a couple of things here that are worth commenting on because there is a bit of a theme here. So first of all, will Oracle 26 AI be supported on Power 10 and 11? Absolutely. It will be supported from both platforms as well as AIX 7.2 and 7.3. I also want to touch upon, there are questions about, you know, what is the compelling reason? I hope what you got out of the session today is that 26 AI offers compelling AI capabilities and IBM Power enables you to move with AI speed and mitigate risks in an entirely automated way. So we hope that, you Know, this provides a compelling reason to start planning for it, start acting on a checklist and so on and so forth. Let's see here. There are some other questions here around spire cards. Spire cards are not required to make use of 26 AI on IBM Power. Let me see. And then there are sort of many other questions. Henry, you want to pick on anything specific from the questions? Yes, there's a long list. Hopefully we've answered some of these as I was in presenter mode. Ah, well, there is question number five. Question number five is the usual question. So when can we expect to see the release? So the release is coming in the first half of 2027. A lot of folks often ask me, why is it taking so long? If you've followed our release schedule, this is a bit different than what we normally have, right? We normally see successive platforms come out soon after we release Linux. But we haven't even actually released Windows yet, which is also a very common platform. And the reason for that is what both Satya and I covered, which is this impact of these foundational models and just the wave of security vulnerabilities that they've kind of detected. I have to say we're no different, right? And if you talk to any other software vendor, they should be telling you the same thing. And we applied these foundation models early on, and we detected a large number of previously undetected bugs. And what does that mean for our release cycle? Well, it doesn't make a lot of sense for us to release Oracle AI database on every other platform with known vulnerabilities. And so this cycle has taken longer for us to kind of get to the AIX and other releases because we didn't want to cut a label to port to every other platform with these bugs and vulnerabilities in place. And so this last kind of eight months or so have been really us focusing on remediating all of these issues and coming out with a label that we can confidently port and release into the marketplace with IBM's assistance with most of this remediation in place. So that's the delay. It's certainly not because we don't want to have 26 AI available on IBM. Again, we invest thousands of man hours, millions of dollars in hardware and farm equipment between the two companies to support this joint effort. So that is the timing and that is kind of the reasoning behind it. It is really to make sure you all get a quality release of Oracle Database on IBM systems. Henry, if I could also, if you want to comment on question 19 as well in the new question section. Yeah. Which? Will Oracle and IBM continue their collaboration beyond the end of 26 AI in 2031? Oh, I don't see any reason that we won't. I mean, some of our most valuable customers, our largest deployments are in AI apps. So, again, I know the delay has created this kind of perception or uncertainty. That's not because of any lack of effort on both of our parts to make sure that the release is available. Yes, yeah. You know, along these lines, the customer population running Oracle on AIX is extraordinarily large throughout the globe, right? So, absolutely, you know, to tackle, to address these customers' needs and requirements, I very much expect to be a very long-term collaboration between our two companies. I see some questions on AI and quantum computing risks. I hope that's considered answered at this point. I see a number of items on here, again, asking about the commitment, the release. Hopefully we've answered all of those. Oh, will there be any features of 26 AI not supported on Power but available on Exadata or other hardware? The answer generally to that is no. All features available on Oracle platforms are available on AIX. The only features that wouldn't be available are features that directly support Exadata, right? So some of the things like our storage indexing are native to Exadata storage. But any of the features I've talked about are available everywhere. So you should not see any kind of issue in terms of anything I've talked about or Satya's talked about today being limited in terms of its availabilities. And, oh, there's one, Satya, you might want to answer number 21, any integration between 26AI and the MMA accelerator and Power 1011? Yes. So, you know, the MMA along with the SIMD is sort of this class of accelerators on IBM Power. So we are absolutely in that SIMD slash MMA mode on Power 11. I also want to pick on this question around the open SSL encryption improvement. So the cryptographic libraries on IBM Power make extensive use of accelerators that are present in IBM Power 11 systems. So that's why the encryption overhead tends to be very low on IBM platform because of hardware acceleration. And in fact, this is where you will see benefit because by moving to OpenSSL, now all those benefits will be visible to all of Oracle users. By the way, this OpenSSl is also going to be made available on 19c, right, Henry? on 19c will make use of OpenSSL as well. So those benefits will be visible on that platform 19c as well Yeah, I'd like to add actually address number 22. It says as Oracle database AI capabilities are currently available only on cloud port platforms. Are there any plans for IBM AIX to be made available on cloud platform or in the future? So let me correct that statement. All of the AI capabilities I just listed in this presentation are available on every platform. When 26 AI goes out on IBM, all of those capabilities are available. And some of the capabilities that were kind of native integrations into our cloud services are also being made available on-premises. So we just released DBMS Cloud, our package that gives access to a number of these cloud capabilities for on-premises system. Our MCP server can be deployed anywhere. So, I do want to make sure that everybody's clear. Where there are some specific AI capabilities in our cloud, let's say like select AI is specific to our autonomous service, there are equivalents available for on-premises. So, you know, private agent factory can be run anywhere. Right. So I do want to kind of clarify that you should not be seeing anything that's kind of hamstrung or anything in terms of the capabilities. Yes, some cloud services we offer are only going to be available as cloud services. We can't really make those available on premises. But for the most part, everything we touch on is available. yeah right at the mark right so uh with one minute to go we should probably wrap up here yeah uh i there's one other i think you might have addressed this that was why should i upgrade to 26 ai if i don't want to run ai on the oracle data um well there are a number of reasons to run you know 26 ai not just ai so things like vector search graph spatial all of the other capabilities within Oracle aren't necessarily AI capabilities. We certainly expect you to want to run the latest release just for long-term support and patching. And as we've said, a lot of these security capabilities are in 26 AI. Things like the SQL firewall, access to the latest encryption libraries. Yes, we'll backport some of those to 19c, but it's always encouraged to run on the latest. any others you want to address I think we are good we are right at the mark so yeah we timed it well Henry we did thank you again thank you for the partnership Satya it's been always a pleasure over the years and you know it's been a great collaboration I think we've demonstrated an ability really to kind of push through some of these challenges like the encryption changeover and addressing some of these security vulnerabilities. So on behalf of IBM and Oracle, we want to thank all of the folks who joined us today. And I have to say, Henry, it's my fondest relationship dating back to 2007 with Oracle Engineering. And I take great pride in this extraordinary collaboration that we have established between two teams. And this year, I made 26AI my personal project. You did? I'm devoting an extraordinary amount of time. So I'm thrilled about this relationship. and I very much expect that it will continue for many, many, many years to come. Thank you all. My pleasure as well. Thank You, everyone.

Company
ILOG SA
Keywords
ILOG SA