Thread regarding IBM layoffs

IBM’s Structural Problem: New Leadership Won’t Fix the Economics

IBM’s problems are structural. Changing leadership will accomplish nothing unless they setup for a sale and get the value till it lasts.

Old R&D prestige does not override 21st century economics. IBM’s hybrid-cloud is not, by itself, evidence of cloud leadership. Customers running Red Hat software on AWS, Azure, or Google Cloud still pay for the underlying infrastructure alongside their software subscriptions. IBM must demonstrate that its additional layer delivers measurable value or risk operating an expensive software tollbooth.

Mainframe transaction capacity also cannot substitute for a compelling economic case. AWS reported more than 59 trillion DynamoDB requests during Prime Day 2026, while Aurora processed hundreds of billions of transactions. These results demonstrate that extraordinary scale is achievable on distributed cloud infrastructure.

IBM’s burden is straightforward. Prove that its platforms improve cost, performance, resilience, and delivery speed enough to justify their price. Historical prestige cannot carry that argument. Until the underlying value proposition changes, a leadership reshuffle risks putting a new face on the same strategic problem.


by
| 1 view | | 1 reply (17 days ago) | Reply
Post ID: @OP+1m5v7jjez

1 reply

Let's add Alibaba / Alipay and the amazing transaction rate they achieve during Singles' Day. In the US, most mainframes are locked down from Thanksgiving till after New Years. There are just too many potential single points of (major) failure to allow anything but standard operation. The hyperscalers (including in China) are much more resilient and fault tolerant.

IBM is in the very fortunate position of NOT really having to prove that the platform improves all those metrics. They would have to prove it if they were in the business of getting new mainframe customers but they don't really get a lot of those. Their existing customer base isn't loyal because they believe they are convinced by such arguments. They are loyal because they are stuck. Almost all customers are moving low hanging fruit off the platform and onto a distributed platform. Some are driving to fully replace the platform but this is usually a very expensive proposition. No customer is delighted with the prices they pay IBM. But as long as that price is a bit less than the cost of getting off the platform, they accept it.

The case where IBM does have to prove that the metrics improve enough to justify the cost is when a new machine (e.g., z17) comes out. That is probably going to get tougher down the road. IBM does rely on customers having aggressive refresh cycles. If a few extra customers skip a generation, it hurts. Fortunately, there is a fix for that too: IBM regularly announces End of Service on software products and generations of hardware. For many customers, especially in regulated industries, that forces their hands.

The other problem IBM has trying to prove things like those metrics justifying the price is credibility. Even internally, the stats and "facts" they spew are nonsense. (WCA4Z/WatsonX challenge anyone?) What they tell customers (e.g., LinuxONE is the cheapest way to run Linux) has been so suspicious for so many decades that every customer takes their "proof" with a grain of salt.

by
| | Reply
Post ID: @bc+1m5v7jjez

Post a reply

: