Home / Trends / Hardware-as-a-service is not the future of on-prem IT – Gigaom

Hardware-as-a-service is not the future of on-prem IT – Gigaom

Every hardware vendor in the world is now pushing for Hardware as a Service (HaaS). While I agree that Software-as-a-Service (SaaS) subscriptions are interesting hardware consumption models, I do not think this will make any difference to the overall performance of enterprise IT and, in general, one of the many options … not even the prevailing one.


Hardware-as-a-Service is one of many ways to describe a new subscription-based model for purchasing hardware, software, and services. An attempt is being made to mimic the flexibility of the cloud and to shift more investment from CAPEX (capital expenditure) to OPEX (operating expenses). All providers now have similar models, and the differentiation is minimal.

When you talk to vendors, they all love this sales model and talk about increasing their business. But let's be honest, if you look at the numbers that are largely transforming the existing business and not really increasing revenue. Some of them are talking about "double-digit" growth in 201

9. In my view, they are overly excited. Just to give you an example: Last week, I heard a CEO of a primary hardware vendor talking about double-digit revenue growth from subscriptions. Last year, it accounted for 0 or just a few percentage points of total revenue. I'm not a financier, but for me even + 79% of almost nothing is still almost nothing.

It's a great model (in theory)

But let's be positive here. Theoretically, HaaS is very important to the user. The private data center will not disappear so quickly, and trying to move from traditional shopping to something more flexible is not a bad idea.

Pay-as-you-go is what everyone wants. No oversupply, no commitment, fast allocation and release of resources, flexibility. Who does not want that? Unfortunately, this is not the case and financial wizards can not provide you with the cloud. Cloud and on-prem are completely different beasts, and trying to make them similar does not bring benefits to the user.

A quick comparison (in practice)

The big advantage of the public cloud is its extreme flexibility. You pay for everything to the minute, with unlimited resources spread all over the world, with all the services you can imagine and ready to go. You pay for it, actually for everything. Cloud is expensive.

The main reason why you go to the cloud is that you want to be agile regardless of the cost. Fast results, the freedom to quickly fail out and restart to avoid any kind of complex and risky capacity planning, upfront investment, and so on draw users to the public cloud. Can you understand this in your local data center? Can HaaS change that? The quick answer is simply "no".

Your local data center is rigid and there is no point paying much more for hardware to achieve a very limited improvement in flexibility. Here's why:

  • Minimum order: As far as I know, sellers still need to apply this model to small installations or clients. In most cases, it is not clear how high the stake is at the table, but to play, you have to put real money on the table. According to an informal survey I conducted last week, it's difficult to get a quote for hardware that has no equivalent of $ 100,000 or more. This means that small projects, SMEs and IT departments are in many cases out of the game.
  • Minimum Time: Another problem arises from the minimum time required for this type of subscription, which in most cases amounts to one year. This means that for at least a year, you have to stick to what you have bought, even if you made a mistake or your project fails. Of course you can extend it, but you can not change what will be the core of your infrastructure. By the way, you can not downsize it even in the first year.
  • Delivery time: We are talking about hardware! Even in the most positive scenarios, it takes days for new devices to be delivered to your data center. However, weeks are more likely. It's not quite as fast as in the cloud, right?
  • Data Center Time: Have you recently made any changes to the management of your data center? Because the same inefficiencies as in the past affect the installation and configuration of new hardware. Including power and network cables and everything that lies between and the moment you can turn on the new hardware.
  • Decommissioning: Nobody talks about it, but what happens if you want to return whole pieces of hardware as a service purchased? Good luck with it.

As I wrote above, all the alleged flexibility described above has a price … because at the end of the day, it's clear that someone has to pay for it and it's not usually the provider.

Snack Stalls (and Common Sense)

Subscription-based models make multiple sense. However, I doubt that they will be the predominant method for buying hardware. In some cases it might be better than traditional leasing, renting or other financial magazines, but I'm sure that was not the goal. There are also some niche application cases, such as the case of MSPs, where in my opinion such a sales model makes more sense (but this could be a story for another day).

All this HaaS craze will fade away time. The beauty of the Hybrid Cloud is its ability to move applications and data where they best fit their business, financial, and technical needs. Why should we aim for more expensive IT in the field without actually achieving a level of flexibility that is not close to the cloud? Is not it better to have rigid IT at low cost and flexible IT at higher prices? Is not it better to have multiple levels for our IT services?

Containers, Kubernetes, SD networks, and new storage technologies redefine application and data mobility concepts. Data and applications can now move freely and (almost) seamlessly. Users can start in the cloud and benefit from flexibility, agility, and infinite scalability. Then, they can move applications and data to lower-cost local IT infrastructures, providing stable behavior that makes capacity planning simpler and less expensive becomes a key aspect.

Source link