TSS = -7%
Power and storage = -6%
Z = +9%
IBM sooner or later has to admit they are just an Enterprise company
11 replies (most recent on top)
IBM is the next Lumen
This nets the quarter out nicely
https://blocksandfiles.com/2023/10/27/ibm-q3-2023/
There is an old saying in sales. Sell what’s on the truck today, and we’ll worry about tomorrow tomorrow. That’s what IBM is doing and given that Mainframes has 2 trillion in legacy (cobol) SW installed, that’s quite a field to farm. It’s only taken IBM 15 years to stumble upon the mainframe go to market golden goose, but it seems IBM has finally figured it out. We really need to thank AI for finally entering the modern age, and IBM realizing how to monetize it. The last 6 quarters of sales at IBM prove that their current go to market strategy is working. Like it or don’t, but they certainly seem to have turned the corner on the old HW/SW/sales model. Keeping mainframe and its cobol install base relevant is job 1, and AK has put in place a strategy that does that. NOW all he has to do is address the parts of IBM that don’t feed the mainframe/enterprise legacy (TSS, scale out offerings, niche services offerings (weather, Watson health, and GTS offerings being prime examples). Do that, and IBM becomes quite the cash generating machine.
https://finance.yahoo.com/m/efa27380-5c08-3ca5-bfbc-0badb651aa66/ibm-s-new-generative.html
When IBM AI has re-written all the COBOL code into Java, where is the future for the mainframe? Seems to me AK is shooting himself in the foot... but hey, isn't what IBM has been doing the past 10+ years?
The "AI" garbage made generally available yesterday is ... garbage. It doesn't even do what they promise it does let alone what they imply that it does. Don't take my word.... buy it and try it....and tell me that you love it.
I wonder how optimistic the Power and Storage sales teams are about their future given the consistent negative growth numbers? Riding it out until they reach the cliff? I would hate to have that cloud (no pun intended) over my head.
As an engineer, I shudder at the thought of trying to maintain a Java program that was originally ported from COBOL by AI. This is not going to end well.
Regarding the comment
"Unless IBM is using AI to do the conversion, AK labeling converting COBOL to java as anything to do with AI is just hilarious. "
Using AI do to this is apparently the plan
https://techcrunch.com/2023/08/22/ibm-taps-ai-to-translate-cobol-code-to-java/
https://www.ibm.com/products/watsonx-code-assistant
Given the investment required to bring P11 to market, IBM is going to find it very hard to justify a P11 investment without finding a willing partner who can also benefit from P11 while sharing development costs. IBM’s legacy Power base continues to shrink, and unless a partner can be found, IBM will most likely opt for a P10 shrink to buy more time. Remember approx 20% of the Power revenue comes from Power being embedded into IBM storage products and all lines that use Power (storage, OS/400, and Unix) are shrinking. Yes LINUX also runs on Power, but Linux on Power is a rounding error vs the legacy OS’s. Who may be viable partners? That is the billion dollar question. The partner will ideally have a commodity parts bin that they can leverage to squeeze profits out of Power and storage manufacturing. I suspect IBM will have to give up some of the “enterprise” manufacturing and sales profit on Power and storage to get a deal done.
"AK indeed mentioned that one of the AI focus areas was trillions of lines of “Legacy COBOL” code being converted to Java"
IBM is the leader in re-labeling everything to the latest fad buzzword of the month. Unless IBM is using AI to do the conversion, AK labeling converting COBOL to java as anything to do with AI is just hilarious.
During the analyst call IBM acknowledged that system Z sales continue to be stronger than expected. (+9%) The CEO and CFO continued to repeat, that IBM’s future plans revolve around continued System Z focus and adoption along with continued System Z innovation toward a Hybrid cloud and AI world. AK indeed mentioned that one of the AI focus areas was trillions of lines of “Legacy COBOL” code being converted to Java. The -6% distributed portion of Infrastructure division was characterized by Power delivering stronger performance than expected, but storage delivering weaker results than expected. The -7% TSS results were mentioned in passing as continued “weakness of TSS services” It is obvious IBM has a scale out vs enterprise problem. IBM continues to execute the “enterprise” portion of the restructuring, BUT the scale out portion of the restructuring continues to be a 750 million to 1 billion head wind. AK it’s time to lance the scale out boil that plagues Infrastructure.