‘We have essentially solved the hardware shortage and the hardware cost issues with a software solution,’ says Krish Prasad, head of Broadcom’s VMware Cloud Foundation (VCF) division.
As customers grapple with memory shortages and rising server prices, VMware says it has built a platform that solves those two pressing issues: VMware Cloud Foundation.
“We have essentially solved the hardware shortage and the hardware cost issues with a software solution,” VMware’s Krish Prasad, who leads Broadcom’s VMware Cloud Foundation (VCF) Division, tells CRN.
“Memory costs are skyrocketing. With VCF 9.0, we introduced something called advanced memory tiering,” Prasad said. “What that allows you to do is to solve this exact problem of skyrocketing cost of DRAMs and availability of DRAMs by saying, ‘Hey, you don’t need to use as much DRAMs. You can tier DRAMs with NVMe storage. And then once you do that, ESX through software, will make sure that you’re not seeing a significant performance penalty.”
Prasad said VMware is seeing a significant “tailwind” with customers buying VCF due to the global issues around rising server prices and memory shortages.
“This whole shortage is causing customers to rethink the value of virtualization and do more virtualization. So that’s the underpinning of our VCF cloud,” he said. “So it’s been a huge VCF tailwind in recent months for us.”
VCF 9.0 Is The Answer To Memory Shortages, Hardware Cost Concerns
VMware Cloud Foundation 9.0 is the Palo Alto, Calif.-based company’s flagship unified private cloud platform launched in 2025. VCF 9.0 treats servers, storage and networks as fluid software pools or resources with baked-in cybersecurity and enterprise-wide compliance guardrails.
The platform features Advanced Memory Tiering with NVMe, which offloads cold data from expensive DRAM to lower-cost NVMe storage to improve server consolidation.
“This is one of the things that customers love about VMware and traditionally has been very much in the DNA of VMware—which is to foresee problems that might happen in the future,” said Prasad, general manager and senior vice president of VCF.
“This is a great example of that, where the memory tiering capability was designed and developed over multiple years, became available in VCF 9.0. Since then, the memory [prices and shortages] went through the roof, and customers are now able to take advantage of it with VCF,” he said.
In an interview with CRN, Prasad takes a deep dive on the VCF 9.0, if he believes the memory and hardware shortage with continue throughout 2026, and VMware’s technology roadmap ahead.
Is VMware being impacted at all by the industry’s memory shortage or rising server prices? Or is it a tailwind for you?
Because of all the huge investment that is happening in the AI space, there’s a shortage of hardware. There is a shortage of memory or DRAMs. And the memory costs are going through the roof.
Similarly, the server costs are going through the roof.
And not only that, you have to be in line to get access to servers because there is a capacity crunch because they cannot meet the demands that AI is driving across the world.
So what is happening is customers are back to thinking about the basics. The basics for VMware was about doing more with less, right? That is from the day one. If you think about virtualization, it was about taking a server and maximally utilizing it by running other virtualized servers on top of it. It’s really about consolidation and really utilizing what we have more effectively.
Because they cannot get hardware at the pace they need to, they have to take advantage of the hardware that they have, but more efficiently.
So what customers are doing is they are packing their existing virtualized infrastructure much more. And there is no other platform in the world that allows you to pack more than VMware.
They are taking their bare metal servers that they have in the data center, converting them to virtualized servers, so they can get much more capacity out of it.
So overall, this whole shortage is causing customers to rethink the value of virtualization and do more virtualization. So that’s the underpinning of our VCF cloud. So it’s a huge VCF tailwind in recent months for us.
How is VCF the answer in terms of rising memory costs and shortages?
If you think about what’s happening around memory costs, it has gone up multiple folds.
Memory costs are skyrocketing. With VCF 9.0, we introduced something called advanced memory tiering.
What that allows you to do is to solve this exact problem of skyrocketing cost of DRAMs and availability of DRAMs by saying, ‘Hey, you don’t need to use as much DRAMs. You can tier DRAMs with NVMe storage. And then once you do that, ESX through software, will make sure that you’re not seeing a significant performance penalty.”
We will take care of the memory management across those tiers.
So do the cheaper NVMe devices tiered with the lesser amount of DRAM and then you don’t see the difference in how your applications run. We have essentially solved the hardware shortage and the hardware cost issues with a software solution.
So VCF is saying, ‘Hey, you can solve it some other way. You don’t need to really be going after this much weight in DRAM. Use the memory tiering capability that came out in the product to address that.’ And that’s resonating very well with customers.
How was VMware able to get ahead of these IT shortages and price increase issues?
It came before the customers ran into this significant problem, because memory is very critical for applications to run and maintain the performance and all of that.
This is one of the things that customers love about VMware and traditionally has been very much in the DNA of VMware—which is to foresee problems that might happen in the future.
This is a great example of that, where the memory tiering capability was designed and developed over multiple years, became available in VCF 9.0. Since then, the memory thing went through the roof, and customers are now able to take advantage of it with VCF.
So it has been another big tailwind in terms of our adoption of VCF 9.0.
So what are customers saying when they buy VCF 9.0?
Customers are going to the basics, which is, ‘Hey, I need to consolidate more. I need to densify my infrastructure. I need to do more with less because I’m not able to get more.’
Then the whole answer is around virtualization, right? And VCF is the best platform in the industry for doing that.
So we see this as a place where we can play a critical role in helping our customers go through this crisis.
Do you see these shortages and price increases going away soon?
It’s not going to go away real fast.
Because the demand and the shortages and all of that is driven by AI. And AI is not going away anytime soon. Everybody is building out big data centers and all that kind of stuff.
By the way, that also is going to drive up the cost of the public cloud.
Because the public cloud vendors are the ones who are building out these mega data centers and all of that for AI.
When you do that kind of investment, at some point, it will reflect in the pricing across the board. I don’t see the public cloud cost issue going away anytime soon.
As the leader of VCF, what is VMware’s technology roadmap in 2026 and beyond?
When you think about our roadmap, you should always think about what our core value proposition is.
The first one is about reducing the total cost of ownership for the customer. For example, for storage, we have vSAN.
If you compare it to the customer using any kind of external storage, vSAN is significantly better in terms of total cost of ownership. It’s not even close. It’s much less. So you pay much less for storage if you use vSAN than using any other storage vendor.
The second one is about container modern workloads.
More and more of the AI workloads are all containerized. We are the only platform in the industry that has both containers and VMs running on the same platform, on a single platform. If you bring another vendor, then you have two platforms. If they have us and some others, then everything gets much more complex.
We provide all types of workloads running on a single platform—whether it’s AI, whether it’s other modern apps, whether it’s VM-based traditional apps—all runs on the same operational models. You train your people once, it’s much more efficient to do it that way, and much more secure, because you’re only protecting one platform.
So that’s where you’d see our roadmap continue to focus: It’s on the cost, it’s on security, and then obviously, how do you enable AI into the platform so that your private cloud essentially becomes more and more automated over time.
