If you've looked at the price of computer parts lately, you may have noticed something strange.

RAM isn't as cheap as it used to be. SSDs aren't as cheap as they used to be. And even traditional hard drives are being affected.

So what's going on?

The answer has a lot to do with something you probably hear about every day: artificial intelligence.

The AI boom isn't just affecting chatbots and graphics cards. Building and operating today's massive AI systems requires enormous amounts of memory and storage. Data centers around the world are buying hardware on a scale that is difficult for the traditional PC market to compete with.

And that demand eventually reaches your local computer store.

At Milwaukee PC, we're already seeing the effects when we're pricing upgrades, replacing storage drives, and helping customers decide whether repairing an older computer still makes financial sense.

So why is this happening?

Let's break it down.

AI Needs an Incredible Amount of Memory

When most people hear "AI," they think about a GPU.

And that's understandable. The graphics processors powering AI systems have received most of the attention.

But GPUs aren't the only thing these systems need.

AI systems also require enormous amounts of memory to operate.

Large AI models can contain billions or even trillions of parameters. Those models have to be loaded, processed, moved around, and accessed extremely quickly.

That requires specialized high-speed memory, including High Bandwidth Memory (HBM), along with traditional DRAM.

HBM isn't the same thing as the DDR4 or DDR5 RAM you put in your desktop computer, but they're all part of the broader memory manufacturing ecosystem.

And manufacturers have limited production capacity.

When massive data centers start purchasing huge quantities of high-performance memory, manufacturers have a financial incentive to prioritize those products.

That can affect the supply and pricing of other types of memory.

In other words, the computer sitting on your desk isn't competing with another person's gaming PC.

It's competing, indirectly, with some of the largest technology companies in the world.

AI Doesn't Just Need RAM. It Needs Storage.

This is where things get even more interesting.

AI systems don't simply process information and forget about it.

They generate and store enormous amounts of data.

Training datasets.

Model versions.

Logs.

User interactions.

Generated images.

Generated video.

Audio.

Documents.

Backups.

And everything else required to operate these systems.

Western Digital has highlighted this distinction in its own research: AI creates an ongoing storage demand because the data generated by AI continues accumulating even after the initial AI infrastructure is built.

Think about how much data one person can generate.

Now multiply that by millions of people using AI services.

Then multiply it again by businesses using AI internally.

Then consider the enormous datasets used to train those systems in the first place.

Suddenly, the storage requirements become difficult to comprehend.

Where Do SSDs Come Into This?

Modern AI data centers use extremely fast storage for workloads that need rapid access to huge amounts of information.

That's where enterprise SSDs come in.

These aren't necessarily the same inexpensive 1TB SSD you might buy for a home computer. Enterprise systems can use enormous numbers of high-performance drives, sometimes working together as part of massive storage systems.

The underlying technology, however, is related.

SSDs use NAND flash memory to store data.

And NAND flash is also used in phones, laptops, desktops, game consoles, USB drives, and countless other devices.

The problem is that there is only so much NAND manufacturing capacity available at any given time.

As AI data centers consume more of that capacity, the supply available to the consumer market can become tighter.

TrendForce reported that NAND flash remained undersupplied through 2026 as AI-related demand surged and manufacturing capacity remained constrained.

That's one reason we're seeing pressure on SSD pricing.

And Yes, Even Hard Drives Are Getting Caught Up in This

Here's the part that surprises people.

AI doesn't only need SSDs. It needs massive amounts of bulk storage, too.

Hard disk drives are still extremely useful when you need to store enormous quantities of information at a relatively low cost per terabyte.

A data center might use extremely fast SSDs for information that needs to be accessed immediately while using massive hard-drive systems for information that needs to be retained long-term.

Western Digital describes this as part of a larger shift toward continuous, large-scale data systems. Its research indicates that HDDs remain an important part of AI infrastructure because of their economics and ability to provide huge amounts of capacity.

That's important because it means AI is putting pressure on multiple parts of the storage market at once.

It's not simply an "SSD shortage."

It's a much larger demand problem.

Why Does AI Get Priority?

This comes down to simple economics.

Imagine you manufacture computer memory.

You have a limited amount of production capacity.

You can sell that memory to someone building a $700 desktop computer...

Or you can sell enormous quantities to a company building a multi-billion-dollar data center.

The second customer is obviously extremely attractive.

Major cloud providers and technology companies are making enormous investments in AI infrastructure. WD has cited hundreds of billions of dollars in projected hyperscaler capital expenditures in 2026, with further growth expected.

Those companies aren't buying one SSD.

They're buying storage by the rack, by the data center, and eventually by the exabyte.

That kind of purchasing power can have a significant effect on the entire supply chain.

Why Can't Manufacturers Just Make More?

This is the million-dollar question.

Unfortunately, semiconductor manufacturing isn't like making more sandwiches at a restaurant.

Building and expanding semiconductor manufacturing capacity takes enormous amounts of money, specialized equipment, engineering, and time.

Manufacturers also have to decide what types of memory they should produce.

If the market suddenly needs dramatically more HBM for AI systems, shifting production toward HBM isn't something that happens overnight.

The same basic problem applies to NAND flash and other memory technologies.

TrendForce expects NAND supply conditions to gradually improve as production expands and manufacturing processes advance, but its current outlook puts meaningful easing of the supply tightness later in 2027 rather than immediately.

So unfortunately, we shouldn't expect every memory price to suddenly fall back to what we remember from a few years ago.

What Does This Mean for Your Computer?

This is where the AI boom becomes a Milwaukee PC problem.

If you're looking at upgrading an older computer, RAM and storage may represent a larger portion of the repair cost than they did previously.

For years, adding RAM was one of the easiest and cheapest ways to improve an older computer.

Likewise, replacing a small hard drive with a larger SSD was an incredibly affordable upgrade.

Those upgrades are still worthwhile.

But pricing matters.

If your five-year-old laptop needs a new SSD, additional RAM, and a battery, the total repair cost may be significantly different than it would have been a few years ago.

That's why we recommend looking at the entire computer before deciding what to do.

Sometimes spending money on an upgrade is absolutely worthwhile.

Other times, replacing the computer makes more financial sense.

And that's exactly where a professional evaluation can help.

Should You Buy RAM or an SSD Now?

If your computer needs an upgrade, don't necessarily wait around hoping prices will immediately return to what they were several years ago.

Nobody can predict exactly where memory pricing will go.

The current market is being affected by enormous AI infrastructure investments, manufacturing constraints, and rapidly changing demand.

If your computer is already running out of storage or struggling because it doesn't have enough memory, waiting may not save you money.

On the other hand, don't upgrade simply because you're afraid prices will increase.

The right upgrade depends on your actual needs.

That's where we come in.

Milwaukee PC Can Help You Decide

At Milwaukee PC, we work with customers every day who aren't sure whether they should repair, upgrade, or replace their computer.

That's completely understandable.

Computer hardware changes constantly, and it can be difficult to know whether spending $100, $200, or $400 on an older machine is actually worthwhile.

Our technicians can evaluate your computer and help you determine what makes the most sense.

Maybe you simply need an SSD.

Maybe adding RAM will give your computer several more years of useful life.

Maybe your current system is too old to justify putting expensive new hardware into it.

Or maybe you're perfectly fine with the computer you have and don't need to spend anything at all.

We'd rather help you make the right decision than sell you something you don't need.

The AI Boom Is Changing More Than AI

Artificial intelligence is changing the computer industry in ways that go far beyond ChatGPT and image generators.

It is affecting processors.

Graphics cards.

Networking equipment.

Electricity consumption.

Data centers.

RAM.

SSDs.

Hard drives.

And eventually, the prices consumers pay for computer hardware.

The next time you see a headline about another massive AI data center being built, remember that it's not just another building full of computers.

It's another enormous consumer of the same underlying technology that goes into the computer sitting on your desk.

And right now, that demand is helping reshape the entire memory and storage market.

Need More Storage or RAM?

If your computer is running out of storage, slowing down, or struggling with newer software, don't guess at what it needs.

Bring it to Milwaukee PC.

We'll take a look at your system, explain your options, and help you decide whether an upgrade, repair, or replacement makes the most sense for you.

Because even in the middle of an AI hardware boom, you shouldn't have to spend money on hardware you don't actually need.