Folks,

We are not yet in a world where cloud > local for the top 45% of Dell's most profitable customers (think big, very big). And if we ever get there, I don't think they will be looking to Dell/VM for a solution.

Those "whales" are going to want a more integrated solution, so Amazon, Google, and Msft will take their cloud services one step further--further away than Dell/VM will ever be able to take it: dedicated fiber for their largest customers. How? Any of these three can lour their existing client base as bait and cut a deal with the Verizon and ATTs of the world for bulk (dedicated) bandwidth purchases. And if they don't have enough existing clients (Azure) they can post a cash-bond guaranteeing revenues to the telecom (just like msft did with the record companies to sell music in Win8).

At any rate, when your borrowing cost is exponential, and you have no cloud business to speak of, you won't be able to take your services to that next level. I may end up eating my words here, but this is too little too late for Dell--the beginning of the end if you will.

This is not financial engineering, but rather Silverlake and EMC denying reality, or at least postponing it until they can each cash out.

> We are not yet in a world where cloud > local for the top 45% of Dell's most profitable customers (think big, very big). And if we ever get there, I don't think they will be looking to Dell/VM for a solution.

> Those "whales" are going to want a more integrated solution, so Amazon, Google, and Msft will take their cloud services one step further--further away than Dell/VM will ever be able to take it: dedicated fiber for their largest customers. How? Any of these three can lour their existing client base as bait and cut a deal with the Verizon and ATTs of the world for bulk (dedicated) bandwidth purchases. And if they don't have enough existing clients (Azure) they can post a cash-bond guaranteeing revenues to the telecom (just like msft did with the record companies to sell music in Win8).

You're kidding, right? Dell has provided the solutions for two of those three companies. The datacenter segment of their business is both successful and demolishing its only real competitor: a crumbling HP.

Dell's able to build and deliver ridiculously energy efficient datacenters at scale, very quickly. And they do, for Microsoft, eBay, and a number of other companies.

Their competitive advantage on the hardware side is substantial. Controlling VMWare would catapult the software side of things very, very far forward. If they want to build a new cloud platform, they've got extremely strong fundamentals to launch it from. If not, well, they'll probably power whoever does build it in some way (and profit handsomely).

Amazon provides Direct Connect which is in the same neighborhood of functionality as taking out dedicated fiber for their cash cow customers that happen to need a lot of bandwidth to their internal DCs.

I work for an entity that is evaluating Dell's cloud solution in theory and my previous comments on the viability of hardware makers going into cloud services are still unchanged.

The common theme with these hardware makers trying to move to cloud infrastructure services is that they all fail primarily because these companies are so poor at software in comparison. People like to say that Apple is bad at services, but they are far, far better at delivering mass-scale services like iMessages and iCloud than HP or surprisingly even Cisco could (forget Webex, they were acquired in and service availability has dropped like a rock when they were industry-leading after Velchamy left for VMware).

The fundamentals of cloud services of any kind is that it is a service and that means well-defined continuous operational models with tons of software-defined and enforced methods. This is just not in the DNA of most of the large hardware vendors.

This analysis is spot on. I think the fundamental idea here is to use VMWare's excellent products as a gateway into the cloud. In theory, it should work out really well. But like you said, I'm rather skeptical that Dell can actually deliver great software.

I get the sense that everyone in the big-metal hardware market (Dell, HP, IBM) is scared of something like Open Compute driving them to irrelevancy. They're all teching up to own some key piece of the IaaS space instead. VMWare is a nice keystone for that.

VMware is also scared of SDNs and public cloud infrastructure. They try to remain relevant, and they may have an easier path forward than Dell and HP, but their business model is built on virtualizing old Windows boxes. It's the stepping stone for businesses as they virtualize their infrastructure. Just move off physical boxes and into the VMware cloud.

But that's quite different from native-cloud applications. Right now you may need to incorporate VMware in your deployment toolchain, but it won't stay that way forever. None of the big public clouds (Google, Amazon etc.) run on VMware, and I wouldn't want to bet on internal virtualization infrastructure going the other way.

Furthermore, integrating anything VMWare without paying extra for their (clunky, limited) solutions is painful.

I say this as someone that's writing a deployment toolkit for internal use. Their API is an ill-thought-out, illogical, and painful mess, and the more cynical side of me believes it to be that way on purpose. It's harder to get commoditized if you're harder to interop with.

Then again, it could just be cultural. It took years of people complaining about the awfulness of their web based admin tool before they decided to take those complaints onboard and deal with them; and those were the admins who have to use the thing every day, a much louder and larger group than annoyed developers.

Have you checked out tools such as Ansible and Salt? They've done a lot of the hard work for many common operations in vsphere.

http://docs.ansible.com/ansible/vsphere_guest_module.html

https://docs.saltstack.com/en/latest/ref/clouds/all/salt.clo...

Believe it or not, Ansible is what I'm using right now. Except wrapping a friendly web-app around the deployment bits.

My hat is off to their developers, about an hour of dealing directly with VMWare SOAP API was a prelude to a migraine.

They've actually started their own public cloud claiming to directly compete with Azure and AWS: http://vcloud.vmware.com/

EMC/VCE tried that already. Didn't work so great outside of niche markets.