2025 has certainly been another wild year, and 2026 will probably be no different.  Looking out at the future is always challenging.  My crystal ball is no better than yours.  If it was, I’d be a multi-billionaire.

So on the spectrum from speculation to prediction to guarantee, these are definitely towards the left end.  The hosting industry has a special talent for making fools of us all (usually right after you prepay for three years).

Here are five bets for what 2026 will look like.

Continued IPv4 Availability and Gradual IPv6 Adoption

We’ve been hearing for years that we’re at a cliff for IPv4 addresses.  Prices are going to skyrocket because they just ain’t making them any more.

I long ago predicted what would happen:

• As prices climb, organizations that hold large blocks would reach a point where the work of re-IPing is outweighed by the possible revenue from selling those blocks.
• These organizations would sell their blocks, increasing supply, and prices would plateau.
• Then as those blocks were used up, the cycle would repeat.

And this is what has happened.  There’s been a slow trend upwards for prices, not a sudden hockey stick graph change.  Also, the widespread of IPv6 in non-hosting applications (e.g., mobile phones) has helped.

Today you still see people offering VPSes for $1/month.  If IPv4 was truly scarce, that wouldn’t be possible.

I think you’ll see more of the same in 2026.  IPv4 will not be the driving factor in VM pricing.

But two other things will.

RAM Pricing Continues to Be a Pain

There’s a huge shortage at the moment in RAM, and unfortunately it’s not the sort of thing where you can just crank up the factories and make more quickly.  It’ll take time for capacity to come on board.  And that’s assuming it will come on board.  Manufacturers don’t want to have a massive build-out and then find themselves with a glut.

AI is driving this, and even if the bubble pops, RAM will be constrained in the short term.  We may see some interesting VPS offers – more disk, more network, little increase in RAM resources versus 2025.

Power Becomes Scarce

One watt in 25 in the US is used by datacenters.  AI is predicted to double datacenter electricity use by 2030, globally, and in the US it could go from 4% to 12% of all electricity generated.

I think this will manifest in a price spread by location.  Where power is cheap, you may see lower prices/more resources in 2026.  Where it’s not, prices will be flat or resources constrained.

This trend may lead to…

Increased Mainstreaming of ARM

ARM has moved beyond Raspberry Pis into datacenters and some providers have offered ARM-based systems for several years.

In 2026, you may see more providers bringing this option to market.  If power is more expensive, it’s logical that power-sipping gear will be more attractive.  Relatively few subscribers actually drive their x86-64 systems to 100% CPU.  Getting an ARM-based dedicated server (or VM) may be a good deal if you get enough power and don’t get an increased price.

Security Continues to Be a Pain Point

There will be more breaches, more hacks, more port scans, more DDoS attacks, and more social engineering.

That probably isn’t a controversial take.  What I’d like to see is:

• All providers adopting MFA for their panels.
• All providers stop emailing credentials in the clear on email, and instead allow installation of SSH keys at subscription time.
• Templates default to SSH-key-only access.
• DDoS protection becomes part of the standard package.

Will we get there?  Probably not, but we should steer towards more secure setups.

Honorable Mention: AI

Who knows.

Some things seem obvious:

• Generative AI video creation will improve.  Today you can generate high quality 10-second videos easily.  By end of 2026, these will be a lot longer.
• AI services will continue incremental improvements.  ChatGPT 6 will be better than today’s 5.2, etc.
• Nvidia is moving to annual GPU release cycles, which will have interesting effects.  I think a lot of the depreciation schedules companies use for their GPU gear are going to come under pressure, with potentially disastrous accounting results.  Some companies say their GPUs are going to be productive for 5 years, which is hard to accept.
• The “Agentic AI” buzzword will be replaced with some other buzzword.
• Prices for using AI services will climb.  OpenAI loses money on every query today.  Venture capital is not unlimited.  You may see more ads in AI services.

Will the AI bubble pop?  Will we reach ASI?  Your guess is as good as mine.

raindog308

Raindog308 is a longtime LowEndTalk community administrator, technical writer, and self-described techno polymath. With deep roots in the *nix world, he has a passion for systems both modern and vintage, ranging from Unix, Perl, Python, and Golang to shell scripting and mainframe-era operating systems like MVS. He’s equally comfortable with relational database systems, having spent years working with Oracle, PostgreSQL, and MySQL.

As an avid user of LowEndBox providers, Raindog runs an empire of LEBs, from tiny boxes for VPNs, to mid-sized instances for application hosting, and heavyweight servers for data storage and complex databases. He brings both technical rigor and real-world experience to every piece he writes.

Beyond the command line, Raindog is a lover of German Shepherds, high-quality knives, target shooting, theology, tabletop RPGs, and hiking in deep, quiet forests.

His goal with every article is to help users, from beginners to seasoned sysadmins, get more value, performance, and enjoyment out of their infrastructure.

You can find him daily in the forums at LowEndTalk under the handle @raindog308.