Epic Games CEO Tim Sweeney says RAM pricing crisis isn't going away anytime soon — Fortnite boss says crunch 'will be a real problem for high-end gaming for several years'

Trident Z5 RGB
(Image credit: G.Skill)

RAM pricing is quickly developing into an existential problem for high-end PC gaming new-builds and upgrades. If you didn’t already have an uneasy feeling about a 64GB RAM kit costing more than an entire PS5, Epic Games CEO Tim Sweeney’s sage observations might begin to stir your gizzard.

On the massive inflation observed by the Amazon shopper in the U.S., Sweeney remarked, “RAM price increases will be a real problem for high-end gaming for several years.” That’s a worryingly far-reaching prediction. Many users might be able to fend off new build or upgrade plans for a few months, but if we are talking several years, plans for next-gen systems could be adversely affected, or have to be shelved long-term.

Don't blame the resellers (yet)

Sweeney explained to his social media followers that “Factories are diverting leading-edge DRAM capacity to meet AI needs, where data centers are bidding far higher than consumer device makers.” That’s simple economics, where DRAM makers shift production to more profitable lines. It is hard to blame them for making hay while the sun shines, with the boom/bust cycles the industry is prone to. However, the Epic boss’s statement avoids pointing fingers at consumer-facing companies like Amazon or Crucial for engaging in any pricing shenanigans.

We have also recently highlighted the perils of buying PC RAM kits, even as we approach the peak of Black Friday. Great deals on great RAM kits like the G.Skill Trident Z5 Neo RGB Series 64GB (2 x 32GB) 288-Pin PC RAM DDR5 6000 kit on Newegg at $599 ($40 off), are definitely not that great if we step back and look at prices for just a few weeks prior. This same G.Skill Trident Z5 RAM kit was sold for as low as $220 as recently as September 20. That shift will give plenty of folks non-buyer's remorse.

Of course, the AI server building frenzy isn’t just affecting DRAM markets. In recent months, it has also been sucking bargains from consumer SSD-land, and effects are even now being felt on storage solutions like HDDs and high-capacity microSD cards. Where will it end? Some financial soothsayers have gotten more vocal about an AI bubble popping in recent weeks. However, Sweeney doesn’t seem to feel like an AI bubble is anywhere near popping pressure, going by his “several years” prediction.

Google Preferred Source

Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

Mark Tyson
News Editor

Mark Tyson is a news editor at Tom's Hardware. He enjoys covering the full breadth of PC tech; from business and semiconductor design to products approaching the edge of reason.

  • gamerk316
    A lot of these discussions have been focused on the highest-end RAM kits; I'm wondering if some of the more discount ones aren't seeing as much of a price premium put on them.

    But yeah, until the AI bubble pops, RAM could well become more expensive then most mid-range GPUs out there, which is shocking given how cheap RAM has been the past decade or so.
    Reply
  • thestryker
    gamerk316 said:
    I'm wondering if some of the more discount ones aren't seeing as much of a price premium put on them.
    Nope even 32GB JEDEC kits (such as DDR5-4800 CL40) aren't safe and are over $200 (these would have been around $80 a couple of months ago).
    Reply
  • DougMcC
    AI bubble pop hopers are deluded. AI spending is baked in for at least 5+ years. The pop is not in sight. E.g. my company has committed spending to at least that far in the future because there is a long list of stuff for us to do with AI that will advance the business.
    Reply
  • ravewulf
    DougMcC said:
    AI bubble pop hopers are deluded. AI spending is baked in for at least 5+ years. The pop is not in sight. E.g. my company has committed spending to at least that far in the future because there is a long list of stuff for us to do with AI that will advance the business.
    That kind of thinking is no different than in any other bubble
    Reply
  • DiegoSynth
    DougMcC said:
    AI bubble pop hopers are deluded. AI spending is baked in for at least 5+ years. The pop is not in sight. E.g. my company has committed spending to at least that far in the future because there is a long list of stuff for us to do with AI that will advance the business.
    Let's see: what has AI given so far? How much money and resources have been invested for that? How many people were fired? Are companies being run by AI? Why are managers needed but employees fired?
    Does this all add up?

    Is AI drawing better than a human? No.
    Is AI programming better than a human? No.
    Is AI creating anything? No.
    Can AI be trusted with decisions? No.
    Are AI "accelerated" GPUs (with fake frames and blurry scaling) better than real horsepower (hardware)? No.
    Is AI finite? Yes, very much.

    Not saying it's useless: it speeds up things such as translations, summaries, searches, data gathering, etc. It helps with calculations (hence GPU stuff, but this is a workaround covering for greed on hardware production). It helps with sketches, concepts and prototypes, but it's no more than a drafting tool.

    Investing billions and trillions on it, messing up the environment and firing everyone because of it is total idiocracy, regardless of the outcome.

    And everything that goes up too high ends up having a quite big fall.
    Reply
  • LordVile
    DougMcC said:
    AI bubble pop hopers are deluded. AI spending is baked in for at least 5+ years. The pop is not in sight. E.g. my company has committed spending to at least that far in the future because there is a long list of stuff for us to do with AI that will advance the business.
    I hope you're keeping your CV up to date
    Reply
  • DougMcC
    LordVile said:
    I hope you're keeping your CV up to date
    No need. My company has about a decade of revenue growth already committed. The worst case for me at this point is that an AI selloff causes a stock price drop, affording me a massive upside buy opportunity.
    Reply
  • DougMcC
    DiegoSynth said:
    Let's see: what has AI given so far? How much money and resources have been invested for that? How many people were fired? Are companies being run by AI? Why are managers needed but employees fired?
    Does this all add up?

    Is AI drawing better than a human? No.
    Is AI programming better than a human? No.
    Is AI creating anything? No.
    Can AI be trusted with decisions? No.
    Are AI "accelerated" GPUs (with fake frames and blurry scaling) better than real horsepower (hardware)? No.
    Is AI finite? Yes, very much.

    Not saying it's useless: it speeds up things such as translations, summaries, searches, data gathering, etc. It helps with calculations (hence GPU stuff, but this is a workaround covering for greed on hardware production). It helps with sketches, concepts and prototypes, but it's no more than a drafting tool.

    Investing billions and trillions on it, messing up the environment and firing everyone because of it is total idiocracy, regardless of the outcome.

    And everything that goes up too high ends up having a quite big fall.
    AI drawing is better than _most_ humans.
    Ai programming is better than _most_ humans.
    AI is creating lots of stuff, but maybe your definition of 'creating' is weird.
    Can AI be trusted with decisions. Yes. And it is.
    GPU frame fuzziness: probably you have a point here, but I have no experience with it.
    Is AI finite? Yeah, obviously.
    Reply
  • DougMcC
    ravewulf said:
    That kind of thinking is no different than in any other bubble
    Well, i'd contrast it with the dotcom bubble where revenues were nowhere in sight. AI is actually bringing in loads of money, just not enough to be profitable, yet. VS dotcom where pets.com made no money and had no pathway to make money yet was valued crazy high.
    Reply
  • thestryker
    DougMcC said:
    Well, i'd contrast it with the dotcom bubble where revenues were nowhere in sight. AI is actually bringing in loads of money, just not enough to be profitable, yet. VS dotcom where pets.com made no money and had no pathway to make money yet was valued crazy high.
    AI is not making loads of money when compared to the expenditures. There is also no pathway to actually pay for it either as it's not getting cheaper to scale out. There's a reason why all of the big established money is starting to protect itself and back away from AI. This is not to say that there aren't any positives or business models where it slots in simply that the industry taken in entirety is a house of cards.
    Reply