quantum computing quantum AI IBM artificial intelligence

IBM’s HPC Advantage And The Path To Quantum AI

IBM is pretty much the father of technology in the US. They, for the most part, were critical to the US gaining dominance in this segment, and they are one of the few companies thinking strategically enough to help assure that dominance in the future. The issue in the market is the rise of hedge funds to control most of the US technology firms, and the focus of those funds is on quarterly results, not strategic initiatives. In effect, and in my opinion, these hedge funds are effectively selling off the US’s future to make a buck.

IBM has resisted this trend and was designed over the more than a century it has existed to always operate strategically. This strategic focus is why the company was able to bring to market Watson before any of their peers even had an artificial intelligence (AI) initiative and why it is investing more than its peers in quantum computing.

These two elements add to IBM’s strategic effort targeting HPC (High-Performance Computing) in a way that could make the company almost impossible to compete with when this all blends with their existing OpenPower and System Z platforms.

Let’s talk about that this week.

IBM’s Unique Hardware Advantage

Most of the HPC space is defined by x86 computers, and this segment was just disrupted by the latest announcement by AMD, who effectively took the pure performance lead from Intel. Now I point this out only because AMD is lead by another ex-IBM employee (I’m ex-IBM myself) Dr. Lisa Su. But IBM stepped aside from this X86 based race and instead focused on a very Apple-like strategy (and you could argue that Apple got its strategy from the old IBM, kind of making this full circle) of assuring the hardware stack. This means that IBM, because they own more of the technology in their platform, much like Apple does, can typically provide higher security and better up-time. Since penalties for breaches are going from millions to billions of dollars in the HPC space and a down computer can never outperform a running machine, this focus provides some unique advantages to IBM’s platforms. Recently IBM bought Red Hat which should further increase their unique reliability and security advantages.

Their System Z platform is even more unique—more like a web service in a box—and the ability for this system to serve massive numbers of users is unprecedented in the market as is its security and up-time. This mainframe platform is likely the closest thing to an Apple product in the IT space in that virtually the entire system is a unique IBM design, the I/O capability is unmatched, and the security on the platform leads the segment.

The offset is the performance, but what if the system could be made to work smarter?

The AI Performance Advantage

When you add AI to the solution the result is more efficiency and, much like a racer taking a legal shortcut, the goal isn’t pure speed but finishing the task more quickly than the other racers. In one instance an IBM customer was able to reduce their workload by as much as 66% as a result of moving to the IBM platform. This is a very real tortoise and hare example where, thanks to AI, a system that on spec is slower but gets the work done more quickly because it more intelligently approaches the problem. This AI capability, as it matures, will provide additional automation performance advantages, and time to decision advantages for analytics. This last is likely the biggest performance boost because it really doesn’t matter how fast a system provides data if it takes an excessive amount of time to interpret and act on it and AI speeds up that last process where a typical HPC server, without AI, will not. Granted, for most, this last part is in the future, but then I am talking about IBM’s strategic plan, not just what they have in the market today.

The Quantum Booster

The final element of IBM’s impressive HPC strategy is its quantum effort. No other major company in the HPC space is investing as much in quantum computing as IBM. They even have a prototype system which looks like a prop from a science fiction movie. Quantum computing, for the jobs that it is suited for—like security, massive number crunching, and some types of modeling—is a massive game-changer. In theory a quantum computer can breach any existing encryption key in seconds, and can provide encryption that is—compared to today’s technology—relatively unbreakable.

While this is still several years out, IBM appears to be years ahead of anyone else, suggesting that when they can deliver the complete blended solution, they may own the top of the market.

Wrapping Up: IBM’s Strategic Edge

Chasing X86 was never going to be a winning strategy for IBM; it doesn’t play to their strengths, which are customer service, security, and reliability. So, they have figured out a way to boost their platforms significantly by using advanced technologies like AI and quantum computing. These, as they mature and make it to market, can, and will, provide the firm with massive solution performance advantages and, if the firm executes, effectively (and eventually) giving their products superpowers allowing them to jump over their competitors with a single bound.

This advantage showcases the huge potential advantages of thinking strategically and of the technologies that IBM is and will be increasingly connecting to their HPC solutions. It does, however, make you wonder both what people will do with these coming blended systems and what IBM will be doing next to massively disrupt the market.

Scroll to Top