IBM Previews New Power Tech At TechXchange Event
October 21, 2024 Timothy Prickett Morgan
I wish we could tell you what was up, but we don’t know what it is. But we do know that all of the top tech brass at Big Blue related to the Power Systems platform are going to be participating in the keynote address at IBM’s TechXchange 2024 event in Las Vegas – what we used to call PartnerWorld – on Tuesday.
Like a bad AI model with not enough parameters and not enough data broken down into tokens, we have to try to infer from first principles what IBM might be announcing and therefore what we might see announced before the end of the year.
First and foremost, we do not think that any announcements that IBM is talking about at TechXchange 2024 will have to do with the Power11 processor and the future systems expected to be based on it that are due sometime next year, perhaps around June if the rumors are right.
We admit, it is odd that IBM did not talk about Power11 at the Hot Chips 2024 conference back in August, and we think that there is an outside chance that this might happen behind closed doors at TechXchange for key partners and customers. But we do not expect much in the way of details about Power11 and its systems and this week’s conference.
What we do know is that Steve Sibley, vice president of product management and Bill Starke, distinguished engineer at Power Systems and the chief architect of the Power10 and Power11 processors, will be featured speakers at the opening keynote. Rebecca Gott, chief technology officer of Power Systems, will also be giving presentations, and so will Bargav Balakrishnan, who is identified as vice president of banking and industry modernization infrastructure at Power Systems.
First of all, we didn’t know that Power Systems had a chief technology officer and have never heard of that title being used before. If you look at Gott’s profile on LinkedIn, she has been a distinguished engineer at IBM since 2017, working on various System z mainframe platforms before moving over to Power Systems in September 2022. Gott is also an expert on blockchain. In her Power Systems role, Gott is “responsible for the technology roadmap across silicon, firmware, operating systems and solutions for IBM Power Systems” and “leads the cross-stack technical use case development and differentiation in IBM Power strategic segments, sustainability, cross-IBM platform collaboration, and development of the Power roadmap across the stack including system icons, OpenShift, security, AI, and hybrid cloud management.”
Balakrishnan also hails from the Systems z line, and according to his LinkedIn profile was an SRAM designer for mainframe processors. He rose up the ranks and eventually became the offering manager for System z in 2013. In 2017 he was named chief of staff to the general manager of the System z general manager and was tapped to be vice president of product management for IBM Systems in August 2022. The word on the street is that Balakrishnan will be taking on some of the duties usually performed by Sibley, who is being tapped to find new ways to grow the Power Systems business. None of this is reflected in their respective titles on LinkedIn.
Starke hinted that he was going to be talking about “exciting new technologies” in the TechXchange 2024 keynote:
But that could mean just about anything.
Gott was equally vague in her posting: “Join me at IBM TechXchange conference the week of Oct 21st to learn about latest innovations on IBM Power and how AI is being explored and used within the enterprise.”
There are certainly some things that IBM could do in the AI space even before Power11 is launched. It would be interesting to see how the on-chip matrix math and vector accelerators in the Power10 chip could be used to do retrieval augmented generation (RAG) processing against corporate data using Watsonx models and without the need for GPU processing. There could be enough oomph for this to be useful for production applications. And if not, IBM could create hybrid CPU-GPU machines to augment the AI processing of Power iron – even using inference accelerators in outboard servers linked through memory coherent fabrics. It would be interesting to see a mix of Power10 processors and AMD accelerators as a counter balance to Nvidia’s hegemony running the ROCm library and driver stack from AMD, the Python language, the PyTorch framework and Llama 3.2 models from Meta Platforms. All of this software is open source, unlike most of the Nvidia AI Enterprise stack.
That’s what I would do.
Anyway, whatever Big Blue is up to, we will let you know when we find out.
RELATED STORIES
GenAI Interest ‘Exploding’ for Modernization on IBM i and Z, Kyndryl Says
Some Thoughts On Big Blue’s GenAI Strategy For IBM i
How To Contribute To IBM’s GenAI Code Assistant For RPG
IBM Developing AI Coding Assistant for IBM i
The Time Is Now To Get A GenAI Strategy
Top Priorities in 2024: Security and AI
Thoroughly Modern: Proceed With Caution With AI In The Landscape Of Cybersecurity
IBM i Shops Are Still Getting Their Generative AI Acts Together
IBM To Add Generative AI To QRadar
How Long Before Big Blue Brings Code Assist To IBM i?
Generative AI Is Part Of Application Modernization Now
Sticking To The Backroads On This Journey
With Fresche’s New CEO, There Are No Problems, Just Solutions
Enterprises Are Not Going To Miss The Fourth Wave Of AI (The Next Platform)
IBM Introduces watsonx For Governed Analytics, AI
Technology Always Replaces People While Augmenting Others
Thanks to Timothy Pricket Morgan for his thoughts on the IBMExchange conference possibilities now in Los Vegas.
I reviewed the conference topics and the focus appears to be on AI. and the future., rather than what IBM needs to do now to actually grow and thrive and succeed.
After seeing iBM continure to slide ever more in annual revenue and in stock maklet capitalizaton behind SAP, Apple, and many others who offer complete usable solutions, it is clear to me that IBM simply must offer complete software and hardware solutions that new customers can actually immediately use successfully without expensive IT staffs.
Like the wildly successful IBM MAPICS product that made IBM successful and the wildly successfu IBM Installed User Program (IUP)s that provided tens of thousands of new IBM customers with the best in class IBM software applications including application source code together with excellent IBM hardware in decades past.
IIM today is in serious decline and is wasting critial time and effort on “Modernizing” as it loses its competitive position , and promising AI and C++ as its savior rather than excellent usable solutions thet customer needs and wants, like SAP does, is a disaster..