IBM (NYSE:IBM) is a company in transition. On the one hand, there are reorganizations and layoffs as the blue giant is transforming itself into a cloud-services company. On the other hand, there are flagship research projects such as Watson or the Quantum Experience as a foreshadowing of things to come in the next few years. I recently had the opportunity to ask IBM some questions about these projects and other things on my mind. This article contains said questions and IBM's answers.
Since my questions cover multiple research and product divisions at IBM, I did not interview a single person but rather had my questions forwarded to the relevant people. Readers are welcome to comment with more questions.
Q: Let's begin by talking about IBM's cloud strategy. IBM has a large number of cloud services from the Watson cloud to SoftLayer, IaaS, PaaS and SaaS. Where do you see IBM's specific strengths?
A: As you can imagine, we have a lot to say regarding cloud. The first phase of cloud was characterized by migration of IT to the cloud. This phase is still happening, with companies at varying stages on the migration path. IBM has participated fully in this phase of cloud and has built a $10.8 billion cloud business by understanding how to help clients take their existing IT infrastructure and move it into the cloud. The next phase of cloud will be about new innovation and competitive advantage. IBM works with the world's most recognized brands on cloud transformation projects. Examples include Marriott International, Shop Direct, US Army, Lufthansa, AIG, WhatsApp, Fitbit, Gamestop and Westpac and industry solutions such as Coca Cola, Halliburton, Kimberly Clark, Siemens, Ford, 1-800-Flowers or Lions Gate.
IBM is poised to lead the next phase because that plays to our unique strengths - hybrid, open, industry clouds infused with cognitive and analytics. Our differentiated cloud approach includes:
- Hybrid: Hybrid is the fastest growing segment in the cloud space. When it comes to cloud for every large enterprise, we hear from clients all the time that they don't want to move everything to the public cloud or keep everything in the private cloud. They are balancing data regulations and security and privacy for clients. Our partnerships with VMWare or SAP and Box are great examples of how we are giving clients the ability to utilize all of the investments they've made in IT, as is our recent announcement that we are cloud enabling all relevant IBM Software so clients can connect to any data, application or transaction system to the cloud. As part of that news, we introduced API Connect which allows any client to publish its IT as an API - making it easy to find, call and connect over the cloud. According to a recent UBS survey, 60% of all IT administrators and CIOs, mostly from enterprise-size businesses, agree that hybrid solutions make the most sense. Many analysts cite IBM as a leader in this market (i.e. Forrester Research named IBM a leader in its latest Wave report on hybrid cloud management solutions, while Synergy ranked IBM as a top performer and leader in the private and hybrid cloud space for 2015).
- Developers: IBM offers the easiest on-ramp for any developer or programmer to create cloud apps through Bluemix, IBM's Cloud platform. More than 20,000 developers a week are joining Bluemix, launching more than 112K apps a month. We reach new developers through the broadest support of open source cloud communities (openstack, cloud foundry, docker, spark) and ensure developers can interact across the most popular programming languages available -11M Swift developers, 13M Java developers, 11M node.js developers. We will be the most open cloud and as a result we will make it easiest to connect clouds to existing IT (ie the hybrid model).
- Bluemix is also bringing approximately 30 cognitive-based APIs to developers. Approximately 60 percent to 70 percent of all customers coming into IBM's cloud services tap into Watson. With IDC reporting that 80% of cloud applications in the future will be data intensive, intelligent apps fueled by data are a reality today. It's clear that Watson APIs will help differentiate them.
Industry Clouds infused with Analytics and Cognitive: The next phase of cloud will be industry cloud services which are higher value and higher margins. We are creating industry specific ecosystems on the Cloud, e.g. Watson Health and Watson IoT...We now have:
- One of the world's highest-volume IoT platforms in Weather.com
- One of the largest commercial health data clouds
- One of the largest commercial video platforms
- Blockchain as a Service on Bluemix, which is very hot and will be transformational; and
- One of the largest providers of spark-as-a-service and data-as-a-Service marketplace.
- We are also building strong developer ecosystems through Bluemix, where we are rapidly rolling out new innovations on the cloud. Data-as-a-service, Watson APIs, Block Chain, Quantum and IoT services are just a few examples.
Comment: This is really a quite comprehensive overview. I think the most interesting point is really the hybrid cloud. "Old" businesses outside pure tech companies and Silicon Valley want to enjoy the advantages of cloud-based infrastructure but cannot just move to Amazon or Azure.
Q: Google (NASDAQ:GOOG) (NASDAQ:GOOGL) has recently announced a broad cloud initiative and wants to make its machine learning services (e.g. Google translate, TensorFlow) a differentiator. Similarly, Microsoft has announced a special focus on NLP and a bot framework. Do you see Watson directly competing with these initiatives? I had a look at the Watson cloud API and it seems to contain a vast set of NLP, information retrieval and more generally deep learning features similar to what I expect Google to be offering in the future. Then again, Watson itself seems to be much more than just API endpoint in demos. What will we get from Watson in the future?
A: Watson is much more than an API endpoint or one piece of technology (machine learning, NLP, etc.). The Watson platform has dozens of cognitive services based on more than 50 underlying core artificial intelligence technologies, including deep learning, neural networks, NLP, machine learning and more. The reason we're able to offer such a broad and substantial set of services, rather than focusing on one or two particular technologies - like some competitors may - is driven in large part by our focus on science. IBM has a world-class research division where we invest $6B annually - that's a serious amount - and we bring the results of these efforts to technology throughout the company, including Watson.
Comment: Well, I guess I had to try - unsurprisingly, no specifics here about where Watson is going. Next attempt below.
Q: Will the Watson architecture always remain proprietary? I'm thinking of a comparison to the Hadoop ecosystem which is the backbone to many cloud companies today (e.g. HortonWorks). Where do you see your cloud strategy going here?
A: The Watson platform is open to developers around the globe and we have contributed SystemML, our machine learning system, to the Apache Open Source community (it's now referenced as Apache SystemML). In addition, we're making Watson services available to developers on other platforms, the first being Twilio and its more than one million developers - (here). Watson is a cloud service and fully accessible via APIs. It can be deeply integrated and tuned with a vibrant and growing ecosystem - most cloud services are accessed in this manner.
Q: For my readers who are invested in IBM: When do you expect the Watson division to contribute meaningfully to IBM's total revenues? Cloud revenue itself seems to have been growing handsomely.
A: Watson is part of IBM's analytics business, which was reported under the Cognitive Solutions Segment in our last earnings report. In Q1, we reported that Cognitive Solutions generated $4 billion. Over the last 12 months, our strategic imperatives, which includes analytics, has delivered $30 billion in revenue (37 percent of IBM's total revenue). We see cognitive computing as a real growth area, which is why we've invested $1 billion in Watson.
Q: IBM as a company is reorganizing itself, which includes "workforce rebalancing." How long do you expect this reorganization to take? Where should we expect IBM to be in 2020?
A: It's hard to give you an exact number as we look into the future. One thing to point out though is that at of the end of 2015 we had approximately 379,000 employees - that's up about 50,000 from 10 years ago. We're constantly innovating, and innovation requires new skills. We look to our workforce to have the skills to match as we move into new areas. We currently have 20,000 open jobs.
Q: Research projects: IBM Quantum experience - exciting stuff from a long-term perspective. Presumably too early to comment on any further product plans in the coming years?
A: Yes, the IBM Quantum Experience is very exciting. We believe quantum computing is the future of computing and has the potential to solve certain problems that are impossible to solve on today's supercomputers. As far as other product plans, IBM Research has been the innovation engine of the IBM corporation for more than 60 years. While we don't disclose specific future plans, our team of researchers will continue to invent the future of technology by making scientific and technological breakthroughs that will transform businesses and society for many more years to come.
Q: More generally, quantum chips seem to be dedicated hardware. Do you expect IBM to always do both hardware and software or is this confined to research projects? How does this fit into the vision of a cloud company?
A: IBM will always provide solutions that will best suit our clients and their respective businesses. Our systems hardware business had a strong 2015. Since the launch of the z13 system last year, we delivered mainframe growth of 35 percent. One key thing to note is that the z13 is specifically designed for the digital era - for cloud, mobile and advanced analytics. We are indeed a cognitive and cloud business, which is why we are anchoring our mainframes in corporate analytics and hybrid clouds.
Q: More research projects: It caught my eye a few years back when Craig Gentry joined IBM research. Any comment on where fully homomorphic encryption is going at IBM as a potential cloud differentiator? Is HElib being used in any products right now?
A: There has been a very rapid progress in the state of fully-homomorphic encryption since Craig's original breakthrough, and the technology seems on the verge of becoming usable for certain applications. That said, it is still not used in any product, and it may still be a few years before it does.
Comment: This question was primarily out of personal curiosity. Homomorphic encryption enables the execution of any query (or technically function) on encrypted data. This means data can be stored fully encrypted in the cloud and (encrypted) queries are run over encrypted data. The encrypted query result is returned to the client, who can decrypt it with his private key.
The whole point of this is that at no point the cloud service provider or any intruder at the server can see unencrypted data. Craig Gentry has provided a constructive proof for homomorphic encryption in his PhD dissertation, which yielded him a MacArthur fellowship and a position at IBM research. The main issue with homomorphic encryption is performance, and it's just not ready for prime time yet from my understanding. Once ready for widespread use, it will be hugely interesting to the cloud market.
I think this covers most of what's going on at IBM now that I could think of - and that IBM would be willing to answer without making material statements about its guidance. I think the central takeaway is really that in the cloud, IBM is largely targeting a different market than Google or Amazon (NASDAQ:AMZN) with its hybrid focus.
Further, I'm always excited to see interesting research come to fruition and as we have seen with the Quantum Experience IBM clearly has some long-term plays going on that have substantial potential. However, the issue of declining revenue is what investors care about in the medium term. What do readers think about IBM's portrayal of itself? Any further questions?
If you enjoyed this article, scroll up and click on the "follow" button next to my name to see updates on my future articles on technology, artificial intelligence and cloud computing in your feed.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.