VMware distributor Arrow says minimum software subs set to jump from 16 to 72 cores The French limb of global tech distributor Arrow has emailed VMware partners it serves with news of big price increases. In an email seen by The Register, Arrow explains that as of April 10th “the minimum number of cores required for VMware licenses will increase substantially, from 16 to 72 cores per command line.” We've …
Broadcom might get away with this in the short term, and short meaning next few years of IT strategy for many smaller customers. I guess they don’t want them. Whilst they may be hostage for now, the smaller customers will be redoing their roadmap strategy and removing VMWare down the road.
Good time to see how proxmox, openstack and xen emerge from this. And cloud of course.
Broadcom will make so much money from this in the short-to-medium term that the long-term will just be the icing on the VMware customer cake.
All the signs from Broadcom have indicated massive licensing increases and customers have been slow to move based on the stats released to the press. Customers have hoped it won't be as bad as some were forecasting.
So naturally, the next step for Broadcom is even larger pricing increases. Just like all the other times and IT company has acquired a cash cow.
If you don't have a strategy in place to move off of VMware or at least significantly downsize your VMware estate to just the hard-to-migrate VMs (assuming a 2-3 node x 72 core/per node cluster is a significant downsize).
Or maybe Broadcom will release a new product to assist you....like VMware lubricant....
Very accurate. We are heading elsewhere.
The "focus on the huge customers" fallacy has ruined so many previously successful systems and applications. Yes, smaller customers produce more support, but also many smaller customers often become larger customers and staff often move from smaller customers to larger customers. If these staff are used to using a different environment then they won't consider the "huge customer solution" and instead go with what they know instead. End result, "huge customer solution" becomes very niche and dies.
This post has been deleted by its author
We are exactly in the space this is being aimed at - smallish customer, double digit number of hosts, most ESX servers around 24C - 32C. I had been thinking we might have to suck up the price rises or drop down licence levels to the basic bundle, losing features we use like DRS etc, but this latest change is potentially a gamer changer in being enough of a "Foxtrot Umbrella" to persuade my management to accept the cost and risk of getting off VMware. Thanks Broadcom!
https://regmedia.co.uk/2025/03/28/screenshot_arrow_france_vmware_email.jpg
Example:
If a customer has a single-processor server with 8 cores, VMware by Broadcom will license 72 cores. If a customer has 5 dual-processor servers with 16 cores each (i.e, 160 cores), VMware by Broadcom will license 160 cores.
I mean from the example in the screenshot posted, and quoted above, it seems to not be 72-cores per physical server but per subscription/customer, which can be installed on multiple servers. Otherwise the example for 5 servers would show 360 cores needing to be licensed.
Don't get me wrong still think Broadcom are scum for what they are doing overall, but in enterprise environments with multiple servers that already go past a 72-core count, like you it seems, it shouldn't cause an issue (regarding this specific core change, not Broadcom's overall strategy, still many issues being caused by that)
What Broadcom are too blinkered to see, is that the VMware user base is an enormous community of champions, support and free marketing for your product. It's success and stability comes from a good dev team (many of whom have moved on) and word of mouth.... all those C-Levels on golf courses talking to each other about the fact that they run VMware and their platforms are stable, the users pitching into user groups, community discussions, sharing and genuinely caring for each other. That all disappears when only a relatively small percentage of their user base are left. When the community dies, you'll rely on your own support teams and where the hell are they nowadays?
Suddenly with no community helping champion and support your product for you - for free and keeping it stable, the bigger fish you so revere, will suddenly get poor support and ultimately drop you.
It's an absolute tragedy to watch a company I've so admired die a slow public death at the hands of this greedy, damaging acquisition engine who give zero f***s about technology in any way shape or form and have milked dry and destroyed from within, every company they have acquired.
They're finding the OTHER equilibrium point on the demand curve... one at lower price and high volume, the other high price low volume. Like the mainframe approach... they've never sold large volumes but price:per unit (often leased) is quite profitable. They're kind of ignored now but still making IBM billions a year. Or Oracle database for that matter... I have seen a very old ad from the early 1980s where it was like $29.95 or something like that LOL. I daresay they probably aim for the other end of the demand curve now too -- and still have plenty of customers despite MySQL. PostgreSQL and a choice of commercial DBs.
I hate to say it but Broadcom can probably successfully milk this for a looong time. Given the mainframe market. perhaps indefinitely.
I keep wondering why people persist in still considering Broadcom as a provider of anything (service, product, you name it).
At this point, what they are is an entity which spent 61 billion to make 900 billion. History has seen worse performance.
From there on, they can do anything they want. They can experiment with clients, kick out anyone they want, tinker with any setting they want. All they have to do is provide some simulacrum of activity - the more controversial, counterintuitive, illogical, shocking and against common business sense - the better. They can stay afloat if only by playing with their own stock.
They are like a handyman or a plumber who won the lottery, but have to keep the business open for tax purposes, not worrying about client satisfaction.
Or like that wedding photographer that got so much in demand that he started raising his prices to discourage clients, eventually levelling at around $90k per wedding for a sustained but not overwhelming flow of clients.
They can play with that flow to infinity in any way they want.
At this point, they have their own financial gravity pull - they attract money no matter what they do, and they have a wide berth in playing with the orbiting money's speed to keep it as fast as possible, but still just under escape velocity.
At the university here we repurposed the old scientific computing cluster as a cloud using VMware. The nodes each have a total of 32 cores so I'd naively expect a 72/32=2.25 increase in licensing costs.
Since we're also looking at a possible 10 percent general budget reduction due to lack of planning, it seems possible the solution will be to turn everything off but not back on again. For educational institutions Microsoft offers extended Windows 10 support at $1 per seat for the first year doubling each year after that. Will Broadcom offer similar discounts on legacy VMware for schools.
I'm not involved in the IT decisions, so my only problem is as a user of the cloud. Hopefully the discounts are worked out and operations will continue normally.
https://regmedia.co.uk/2025/03/28/screenshot_arrow_france_vmware_email.jpg
Example:
If a customer has a single-processor server with 8 cores, VMware by Broadcom will license 72 cores. If a customer has 5 dual-processor servers with 16 cores each (i.e, 160 cores), VMware by Broadcom will license 160 cores.
Its not 72-cores per physical server but per subscription/customer, which can be installed on multiple servers. Otherwise the example for 5 servers would show 360 cores needing to be licensed.
VirtualBox 7.1.6 runs my oldest VM, an Ubuntu 22.04 installation, just fine. That's the only VM that outlived my career, as it has all my old hobby project work in it's github instance hosted therein. I like to be able to back that up separately from everything else whenever I've put some time in on my pet projects (which hasn't been in a very long time now, as they're just kind of long in the tooth and I'm trying to decide what, if anything, to do with them all...)
The nut of it is that Oracle doesn't want my money unless I'm paying for 1000 seats of their software, so clearly they consider me too small to worry about - one guy running his own VMs on one box is kind of "under the radar" of being worth the effort of collecting money from - they'd spend more processing the bill than they'd earn, especially in this age of Trumpian and Canadian-response tarrifs. (Stupid approach. The last time the US did this, they entered the stockmarket crash of the 1930's and subsequent massive recession. But fools never learn, and Trump is the Ultimate Fool with his sidekick Tool...)
I can confirm the same information has been sent to our company by Dicker Data in Australia (without the core allocation example seen in the Arrow email). We have many small servers (8 to 16 cores). In less than two years we've gone from paying a little over $1k for a perpetual license to $5k for a subscription (only a few months ago) to $22k for a subscription (from April 10). Prices ex. GST
I'm not a virtualization expert by any stretch of the imagination so this comment will seem naive but for smaller companies, is it not feasible to hire a small team of open source folks and just get out from under this corporate greed software licensing cost horseshit altogether?
I am familiar enough with software, computers, and such to know that of course this is not as easy as it sounds but, for example, the tiny city I live in (about 2,500), instead of using off-the-shelf software or better yet open source software, they pay $60,000 a year to license custom software to run the water billing, police department, etc. Beyond ridiculous. I voiced this to the mayor and he said that they have decades of data in the current system and they make it almost impossible to get out. Yeah, no kidding. That's by design. It may be hard and painful to get out but it's certainly not impossible.
Proxmox is the current darling of the open source virtualisation community if you want to take a look.
The big issue is that even small organisations often have specific security/compliance/compatibility/support requirements that can be a lot harder to achieve on less "enterprise" strength product stacks.
On the other hand, there is very much a service offering opportunity there to professionally centrally manage large numbers of small scale Proxmox deployments for a reasonable price, probably for less money to the customer than the VMware licences would cost...
The old VMware had a complicated product line that was designed to get moderate to high dollars from big customers and less money from education and small customers.
This might have ignored the real cost of running a large and sprawling organization to cover all those customers.
The new rule is that all customers must be profitable.
Using low pricing to suppress the competition is no longer a thing.
Most everyone will be better off in the long run this way (since we can't go back to the old way).
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
