As VMware says goodbye, leadership thinks Broadcom buy is a win

"You write what you need to win."

That's what I was told during VMware's Explore conference last week by a very senior Virtzilla executive who was directly involved in efforts to have UK regulators approve Broadcom's acquisition of the virtualization giant. Among the tactics deployed to win approval was an extraordinary document that argued VMware is failing to execute its multi-cloud vision, customers aren't buying it, and only Broadcom's tender care can ensure Virtzilla thrives and survives to present an alternative to hyperscale cloud operators.

Implicit in "You write what you need to win" was a belief that the acquisition is a good thing – it means the tactic increased the likelihood the deal will be done. At Explore I also heard execs enthuse about what Broadcom's investment will mean for VMware and its customers.

If "You write what you need to win" was therefore a negotiating tactic – and not post-approval bluster – it should be a relief to VMware's staff, partners, and customers, because it means the virtualization pioneer is in good shape.

But all remain nervous. VMware staff expect the acuisition will lead to job cuts and therefore imagine a ruined Christmas after Broadcom takes the reins in late October. Partners have reason to be optimistic – Broadcom has repeatedly assured them of their important role in its strategy – but of late the semiconductor manufacturer has mentioned global system integrators as the vanguard of its plans. Small channel orgs have mixed feeling about the world-girdling services giants: sometimes they nimbly evade them, and sometimes they're outmatched.

I understand consultancies who advise on VMware have not heard the end of queries from customers who feel they need to hedge against Broadcom bringing unpleasant change.

It's not hard to see why so many are fearful, for two reasons.

One is that Broadcom has a history of shrinking back office and sales operations – even beyond the sensible consolidations that come with acquisitions. While it has promised investment at VMware, it has also promised swift and substantial profit boosts. Those seldom come without cuts that impact individuals, or the customer experience, or both. And it's not hard to find VMware customers who already grumble about service and support.

Another reason for trepidation is that VMware has spent the last few years claiming to be the company users can rely on to tame the complexities of multicloud operations – writing what it takes to win – but has not fully delivered.

At this year's VMware Explore it took a decent step towards multicloud mastery with the debut of the Edge Cloud Orchestrator – an entirely new cut of the VMware stack that can run on very modest processors. It flips management paradigms by arranging updates of edge infrastructure when conditions on the edge make it possible and sensible to ask for them, rather than directing changes from the datacenter core.

That's a big change – but also an admission VMware wasn't really up to the job before.

A similar change came to VMware's Tanzu portfolio – a sprawling collection spanning a Kubernetes distribution, automation tools, infrastructure-as-code tools, plus networking for microservices. Tanzu has gained an "application engine" and "intelligent services" that bring more automation, plus functions like observability, so the portfolio can be used across more of the development lifecycle.

Again, that's welcome. But again, VMware's spiel for the last few years has been that Tanzu is a fine way to get cloud native software running at scale, then manage its operations.

• VMware edges towards its multi-cloud vision, with a side quest to virtualize AI
• VMware sees no need to Arm itself for multi-architecture multi-cloud
• Here’s how VMware hopes to spend $1B Broadcom R&D budget boost
• SmartNICs haven't soared so VMware will allow retrofits in old servers

I mentioned to execs that VMware's new crop of code is delivering on promises the biz has previously said it had fulfilled. They told me that building software can be a bit like painting a bridge: by the time you finish the job, the end you started at already needs a fresh coat.

That's a fair observation. But from people who write what they need to win, also one that's hard to accept.

VMware will today announce probably its last ever quarterly results as an independent listed company. I've previously argued the virtualization pioneer should be celebrated for carving out a viable niche and defending it very well for an extended period.

That niche will soon be Broadcom's to defend. And writing what it needs to win won't defend it.

Just like VMware wasn't ultimately able to win its bid to stand alone – status that it won just 21 months ago, after finalizing its divorce from Dell. At the time it said independence would create "flexibility to partner even more deeply with all cloud and on-premises infrastructure companies to create a better foundation that drives results for our customers."

And writing that didn't translate into a win. ®