Heh well, forget DDR4, DDR3 rides again <a href="https://www.youtube.com/watch?v=e4c2FSSaCsw" rel="nofollow">https://www.youtube.com/watch?v=e4c2FSSaCsw</a>
My AI model queried itself just this morning to verify cluster status, and it says although DDR5 has been available since 2020, that 2+2 is still equal to 4 for those of us still running DDR4. V<p>In all seriousness, anyone remember when every library, business, government agency, and institution, etc was buying $20,000 cisco routers to "get on the internet" during the dot com boom/bust in the late 90's ?<p>Only to see them all end up on the surplus market, unused, for a couple hundred ?<p>We're in the AI boom/bust cycle, and its my understanding DRAM fab is specialized enough, that the supply is getting squeezed.<p>As for "doing it on purpose" ? It's being milked for all its worth. That's what companies do. They milk the market for what it will pay.<p>Also, for everyone who is trying to do local AI, on Xeon or EPYC platforms, memory bandwidth scales with module count, making them that much more in demand in terms of processing capacity / wafer area.
> As for "doing it on purpose" ? It's being milked for all its worth. That's what companies do. They milk the market for what it will pay.<p>It’s what everyone does, from company to individual. When you sell your used car or list your house, you price it according to what the market will pay. When companies sell a product, they list it for what the market will pay.<p>The conspiracy theories and accusations of collusion were flying everywhere until CXMT came online. It felt like every comment section and headline was predicting that CXMT’s arrival was going to crush the cartel and smash the “price fixing” when a new competitor arrived.<p>Then CXMT memory arrived at nearly the same price as everything else because, to the surprise of nobody with basic economic understanding, they were selling their products at market rate too.<p>The high price is being driven by demand combined with a supply chain that takes years to increase output. Just prior to this demand explosion, RAM and SSDs were so cheap that producers couldn’t justify building out more capacity. It would have been a bad choice at the time unless they had a crystal ball that could have predicted the future.
The important question isn’t so much “are people trying to get the best deal possible for themselves?” but rather “how well is this market functioning?” Or phrased another way, “how competitive is this market?”<p>This particular market is pretty clearly not functioning well, and there are textbook Econ 101 reasons why a market may not function well. Of those textbook reasons, I suspect the most obviously applicable one here is large barriers to entry.
> and there are textbook Econ 101 reasons why a market may not function well. Of those textbook reasons, I suspect the most obviously applicable one here is large barriers to entry.<p>Econ 101 would be understanding that you can’t snap your fingers and have a new DRAM factory online overnight, and that this isn’t an indicator that the market is not functioning.<p>You’re defining a market “functioning well” as meaning the price is low for you and all demand is satisfied immediately without raising prices.<p>That’s not really what is meant by functioning markets. A functioning market <i>does allow the price to increase</i> as this is what spurs investment into new capacity and new developments.<p>It also allows the DRAM to go to the most profitable activities. This is actually what the whole Georgism and Land Value Tax people advocate for with property taxes, which bizarrely many people think they like: The idea is to raise the price of something (in that case, your property tax) to ensure it’s not being “wasted” on an economic activity that isn’t optimal. The most optimal economic activity is the one that can afford to pay the most, which is the one that captures the resource.<p>It’s not fun when your use case isn’t the one that can pay the most, but that’s literally how markets work.
For what it’s worth, I’m willing to buy as much RAM from everyone else at 50% of what these companies are selling it at. If everyone works together, we can cut the market price by selling it to me. I will provide the much needed capital in dollars to shift the price and others can provide the RAM to me.<p>I’m willing to spend a lot to drive down the price. As a show of good faith, I promise not to actually pay any more than 50% of what you can get on the used market.
there's a big difference between now and then. we built up capacity. dark fiber. and couldn't use until we got broadband. In general. ai capacity is being used right now today.
This does not seem to be the case. The very first VeraRubin NVL72 systems are barely installed and won't be running real workloads until beginning of '27.
Colossus II appears to be the largest user of Blackwell Ultra NVL72s and those are also only recently on-line in September of this year with the remaining 3/4 of them not until end of the year. The delays in datacenter deployment are very long. Even their GB200s didn't come on-line until January of '26. Hopper H100/200s still dominate.<p>The other hyperscalers are even further delayed. CoreWeave has some VeraRubin, but not in volume (as I understand).<p><a href="https://en.wikipedia.org/wiki/Colossus_(data_center)" rel="nofollow">https://en.wikipedia.org/wiki/Colossus_(data_center)</a><p><a href="https://www.moduledge.com/blog/nvidia-hopper" rel="nofollow">https://www.moduledge.com/blog/nvidia-hopper</a>
That’s not exactly true, a lot of GPUs are sitting in warehouses because they are lacking prerequisites to being installed, in some cases this is even just power.<p><a href="https://www.tomshardware.com/tech-industry/data-centers-in-nvidias-hometown-sit-idle-as-grid-struggles-to-keep-up" rel="nofollow">https://www.tomshardware.com/tech-industry/data-centers-in-n...</a>
I don’t think the article you cite supports your point with confidence you convey. Yes other dimensions can complicate deployment, but incentives to work around those issues are high, given the cost of GPUs
That’s not the same. Those GPUs are allocated to real projects with real demand. If they could turn them on today they would do it and they would have customers.<p>It’s not the same as dark fiber capacity being overbuilt for demand that wasn’t there yet.
There appear to be a lot of uninstalled GB200s (looking at you Azure), but the power efficiency of those is ~10x worse than a Vera Rubin. Does it really make sense to install them, when your data center is power/cooling limited, and it's not built yet?<p>Maybe somebody will use them, but not the original owners.
Yes. I have 3 GPUs in the other room, sitting in boxes, waiting to be powered on, because I dont have the electrical circuit installed in the room yet.<p>I can directly confirm this fact, and I'm just some random dude out here in suburbia, who can only imagine what the Indiana Jones warehouse of glowing GPUs looks like in a data center.<p>We're all salivating over this imagined future where an AI toilet seat greets us every morning.<p>Who here has already asked AI how to make your own DRAM ? I know you have. Thats a sign of peak interest in a subject or market.
That’s one big difference. The other big difference is that capital back then was mainly spent by startups with no or little traction.<p>Today, the "old" corporations like google, meta, amazon are also spending a lot of money.
We’re doing a similar thing now too, building tons of extra power capacity and links to data centers.<p>After the crash no one will need to build a new data center for 20 years, and there will be plentiful power connections
> ai capacity is being used right now today.<p>A lot of its current use is just basic questions that could be handled by local llms just as well, yet they are being processed on GB200s on the other side of the world.
Plus it's the same ram cartel players since the early 2000s.
> As for "doing it on purpose" ? It's being milked for all its worth. That's what companies do. They milk the market for what it will pay.<p>In supply constrained markets the bidders set the price floor, not the manufacturer. A manufacturer produces a supply; they tender the supply for bids. The highest bidder gets the opportunity to purchase the supply. Everyone else has to wait for the next bid, the waiting is their opportunity cost, and is priced into their bid.
You have described the market milking process very accurately.<p>Theres machine learning algorithms in the mix, and it all works perfectly smoothly, until it doesn't.<p>Smoothly here, being defined as the markets emotional feelings about it all, not, the price went up.<p>The price went up because it could, as you pointed out.
Routers don't do productive work.<p>This boom is only superficially like the dot com bubble.<p>There are absolutely parallels.<p>But, a GPU that can help me write emails, code, articles, etc. has intrinsic value that a Cisco router simply didn't provide.<p>I don't believe GPUs are a new asset as Jensen Huang claims. (His take seems very bubble-ish.) But, I feel these elevated prices are going to last until production catches up.
If you're growing onions in the backyard, you don't need a router.<p>If you're writing emails, you're gonna need to send them through a router or two before someone reads them.<p>> I don't believe GPUs are a new asset as Jensen Huang claims.<p>He sells GPU's, so taken in context, he's promoting the financing of GPU's, by anyone who is listening to his sales pitch. He travels in CEO circles, so, he's talking to other CEO's, and the banking industry.<p>The phrasing of it also directly pushes back against the articles accusing the AI industry of circularly financing itself to inflate everyones balance sheet.<p>And the other thing this idea of it being an asset that you invest in, like a house presumably, is that it primes all of us for 2 future realities: price increases and minimal performance gains ( "its useful for longer so finance it" ).<p>As to economic value proposition to any individual or group, it boils down to how they are going to use it to recoup the expense.<p>Don't just take the sales guys word for it, ever.<p>There have been many videos lately, on whats happening in the automotive market. The industry has been adding more and more tech, to the point where its now financially extractive via mandatory dealer interaction aka fees, and people are becomming fed up with it all, wanting something simpler for less money.<p>Translate this to the big fat GPU card industry, and how long is it before AMD/Intel add just enough AI ops into every CPU that noone cares anymore ?<p>May you all live in interesting times.
Adjadent to the article, there are some interesting builds with pulled parts from decommissioned servers in places where hardware is expensive or heavily taxed.<p>There's an entire parallel ecosystem around Chinese X79 and X99 boards [1](brands like Huananzhi and Machinist). They exist almost entirely to soak up scrap-tier Xeons and decommissioned server DDR3 and DDR4 ECC RAM that get pulled from datacenters for next to nothing. You see them online in Russian/Chinese/SE Asian online communities as budget builds, so nearly all the BIOS mods and documentation are in Russian or Chinese.<p>I think in NA we'd tend to write this hardware off as e-waste unless you're a homelabber (though I know Kimsufi still includes DDR3 builds in their catalogue). But in emerging markets, people have been squeezing real utility out of discarded enterprise silicon for a long time out of pure economic necessity. I've been tempted to do my own build but I already seek to have collected an overwhelmingly unnecessary amount of 2012-era workstations that I'm not sure my partner would appreciate an addition to.<p>[1] <a href="https://post.m.smzdm.com/talk/p/am9d3zv4/" rel="nofollow">https://post.m.smzdm.com/talk/p/am9d3zv4/</a>
I have 2 Huananzhi EPYC boards.<p>At about $350 each, I had to try them. Everything else was double.<p>Only thing I can say so far is, they boot linux on second hand EPYC 7002 series CPU's.<p>Haven't put them through their paces yet. V<p>Oh, I remember quirk #1 ... apparently, only 1 of the case fan headers is speed controllable, the others run at full speed all the time, with no bios setting.<p>You can fix it with a fan hub that supports pwm so not a show stopper by any means, for those wanting to try these motherboards.
Until about 6 years ago I also mostly used ex-server "ewaste" hardware, because it's always been cheaper, except for electricity use which I wasn't paying for myself. My desktop in ~2015 was built from 2006-era hardware, with 32GB of ECC DDR2, which felt like a lot at the time.<p>My "6 years ago" build was with more typical current-gen consumer-grade hardware, because at that point I was on the hook for the electricity bill. I haven't thought about what my next build will look like because it's still going strong (aside from a flakey DIMM... so my next system will have ECC).
I've been extremely pleased with my R320 in a NAS role. Idles at 40 watts, is quiet, has ECC, and the drives stay cool.<p>Maybe that's still too much power for some people, but I'm happy with the tradeoffs.
I imagine those countries have sane energy prices compared to Western Europe.
Regardless if the market allows consumers access to negative solar prices or not, consumers can attach solar panels to their servers (independent of the grid), so the LLM is there when the sun is shining.
Energy is a commodity so the price will be very similar for the same goods. What they tack on as taxes might be different.
Energy means electricity in this case, and the EU prices are about 2-3x those paid in China AFAIK, for both residential and commercial.
Electricity cannot be shipped from one place to another in bulk quantities.
Me too.<p>EU average is just under 29c/kWh (non-contract price) [0]<p>In the UK I got 27.5c/kWh (+ 58c/day fixed charge) on an 18-month contract a month ago<p>[0] <a href="https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Electricity_price_statistics#Source_data_for_tables_and_graphs" rel="nofollow">https://ec.europa.eu/eurostat/statistics-explained/index.php...</a>
That is expensive. If it makes you feel better, U.S. utilities' published kWh rates don't include a whole bunch of other things that make bills much higher than you'd imagine. For example, I used 280 kWh last month for my apartment, at a nominal rate of either 11 or 13 c/kWh (the rate changed from 13 to 11 partway through the month as it got cooler).<p>But after all the "sales true ups" and "affordability charges" and "resource adjustments" and "fuel cost charges" and "interim rate adjustments", not to mention the city fees and sales tax, the final bill divided by my usage gave me an effective rate of 23 c/kWh.
My thought was that the EU average can't be that high, but apparently Finland is just the outlier here.
I think in Quebec they've been paying me to use power.
The E5-2670v1 was goated for this back in 2016-2017. You could get them for like 25-50 bucks. I still have one of the supermicro boards I got from work in hardware decommissioning and one of the chinese franken-boards with a modded bios.
I just bought a Xeon E5-2690 v4 for $16 and a Sparkle A770 for like $350. Going to try some local models on there. I think it will be plenty good for my use cases.
Typing this from an Asus X99 WS 3.1 with a E5-2696v4 . I got 256GB of RAM, sees only 128GB but I am sure with some overvoltage and lowering the ram speed it could see all 256GB.<p>if anybody got a tip, i'd be more than glad to listen, please.
Thanks for this! Your post inspired me to look at my X99-UD4 system that was put in a closet some years ago. The board currently has a 5960x CPU, and I did not know I could put an E5-2697v4 CPUs in it. I have a couple of these v4 CPUs and tons of DDR4 modules from a recent data center decomm event. Time to hack again - it may revive my old PC!
> Typing this from an Asus X99 WS 3.1 with a E5-2696v4.<p>Yeah the Xeon E5 v4 to me are really an amazing spot. These things, today, are basically free. If you look a bit you can buy stuff like four 20 cores E5 v4 for $60, making them effectively $15 each. Basically free.<p>And here's the trick: I don't need my Xeon to be up 24/7. 8/7 is sufficient, so I save 2/3rd of the electricity cost. I set up my network so that when I sleep, my main server doesn't need to be on. The services that really need be on 24/7 (say my DNS for when I'm sleeping and my wife/kid need Internet) are running on a Raspberry Pi (which consumes nothing).<p>And when I need the shitloads of cores / ZFS etc.: I turn the Xeon on.<p>Now it's still a server in my case: my main machine is a recent AMD.<p>> if anybody got a tip, i'd be more than glad to listen, please.<p>Slap 128 GB of the 256 in a $50 used Xeon server and use that has a backup / to run VMs / ZFS / etc.
There are significant performance/energy ratio problems with older hardware, particularly with the rise of ARM.
How are all these related news regarding silicon shortages surprising at all? It was well predicted that while 2023 chip turmoil[1] was happening, big AI aimed to secure almost all of the demand[2] intended for consumer markets regarding RAM chips so as to win the AI race going forward from that date. They cannibalized almost the entirety of the supply chain, so that they could sell you the incredibly subsidized AIaaS[3] with still considerably higher inference and astronomically higher training process without as of late 2026, failing to achieve any meaningful productivity at all[4]. The serfs now will be happy to own their decade old DDR4 architecture.<p>[1]: <a href="https://www.electronicdesign.com/technologies/embedded/article/21257375/objective-analysis-understanding-the-2023-chip-market-collapse" rel="nofollow">https://www.electronicdesign.com/technologies/embedded/artic...</a><p>[2]: <a href="https://www.reuters.com/technology/nvidia-supplier-sk-hynix-says-hbm-chips-almost-sold-out-2025-2024-05-02/" rel="nofollow">https://www.reuters.com/technology/nvidia-supplier-sk-hynix-...</a><p>[3]: <a href="https://arxiv.org/abs/2606.24616" rel="nofollow">https://arxiv.org/abs/2606.24616</a><p>[4]: <a href="https://www.apollo.com/insights-news/insights/daily-spark/No-signs-of-AI-in-the-productivity-data" rel="nofollow">https://www.apollo.com/insights-news/insights/daily-spark/No...</a>
No productivity from AI is a crazy conclusion to make given the impact we already know it's having on software development. Just because you can't see it in the data yet doesn't mean it's not there, which is what the article you cited says.<p>Remember too that these models only really became useful less than a year ago with Opus 4.6 being an inflection point. It takes a long time for the economy to reorganize itself so it's not surprising we're not immediately seeing the payoff in economic data.
Of course, when we talk about utilizing AI as an efficient tooling, it is well expected to see the very outcomes you desire. But the point here is the disconnect between the perceived productivity gains and their corresponding measuring metrics. Remember the Goodhart's Law[1]? It becomes fairly unreliable when you try to measure productivity by a set of standards instead of actually producing productivity. When it comes to future about this issue, I'd hope to see the progress actually happening in open source, decentralized AI instead of the current way of using it.<p>[1]: <a href="https://en.wikipedia.org/wiki/Goodhart%27s_law" rel="nofollow">https://en.wikipedia.org/wiki/Goodhart%27s_law</a>
Yep. To the point it makes you wonder if the "AI bubble" is totally misunderstood. It's less "AI bubble" and much more "Monopolize computing". Buy up computers until most companies have no choice but to use cloud, then buy out chip fabs entirely, then never give anyone a working free-programmable computer again. All software goes through approval from Google/Apple/Microsoft.<p>You know, what Microsoft has been trying with Palladium (can't risk general purpose computers! People might commit, oh the horror, copyright infringement, with those!)<p>Imho, as hardware keeps getting better, it seems to me this is the only thing that can make cloud investments worth it in the long term (which is only 2-3 years away) for hyperscalers. It's the only explanation. If OpenAI/SpaceX/Musk/... really thought AI would take over all jobs, they wouldn't be investing in compute, they'd be buying up farms, land, harbors, mines, bankrupt factories, fucking tax prep businesses ...<p>Make compute only possible within Goog/Amazon/Microsoft/... walled gardens ... that's the play. And this AI is a cute technology that conveniently needs lots of compute, but it's not even the actual core of the play. A very simple very old style monopoly grab, you know, the business plan people got convicted for trying in ancient Persia, 5000 years ago.
> the business plan people got convicted for trying in ancient Persia, 5000 years ago.<p>Excuse my ignorance, do you have more details on this?
> Yep. To the point it makes you wonder if the "AI bubble" is totally misunderstood. It's less "AI bubble" and much more "Monopolize computing". Buy up computers until most companies have no choice but to use cloud.<p>Unless prices get <i>categorically</i> higher, I don't think that's likely. Like 10x more than it is now.<p>Companies were buying computers for all their employees in the 90's, where a mediocre laptop could pretty easily cost two grand in <i>1990s money</i>. My mom just bought a middle-of-the-road laptop for about a thousand dollars in today-money, and that was a consumer buying it; companies can generally spend more. "No choice but to use the cloud" feels like it's way too strong of a conclusion to draw.
I think they'll keep cheap consumer electronics for the most part. Not as cheap as this last decade, but still cheap. They just want to control the entire software chain from boot to network-accessed resources.<p>Are you human? Then you must verify it from a "trusted computer", preferably with your driver's license. Gets governments and corporations as much information as they want, along with a guarantee that you can't "steal" their content and can't avoid their ads.
A "monopoly" grab from a conspiracy spanning all of the major tech firms? You really think that's the most likely explanation? Not just that the firms are competing in AI and all realize they need massive compute?<p>Prices are up a bit but there are plenty of "free-programmable computers" from a dozen or more manufacturers on the market. That's not going away.
The word conspiracy is misused there. That is not what a conspiracy is. This would be all those companies acting in their own best interest, separate from each other. The word for that isn't conspiracy but collusion, and it doesn't even require those companies to actually cooperate with one another as implied by the word conspiracy. They would all be acting in their common interest, with no or limited agreements between them.
I don't think you have your definitions straight since collusion does mean conspiring to work together. Your original post pointed towards AI just being cover for a plan to create monopoly power amongst the tech firms. I don't see how you can make that argument with a straight face. Apple, HP, Dell, Asus, etc are just going to stop selling computers?<p>I take your point that AI is pushing up the price of computers, at least temporarily, but you're leaning way into conspiracy and cynicism in your original post.
I firmly believe this shortage is artificially produced and I really hope chinese manufacturers flood the market and make memory as cheap as toilet paper
It is not artificial constraint. The build out is titanic; the supply of AI compute is doubling every 6.8 months. To feed that much growth hyperscalers have to price everyone else out of the market.
It's only "artificial" in the sense that getting this much money on the AI companies that don't even make a profit is not normal.
It's not artificial as in "RAM vendors lie about real demand and refuse to invest in more capacity".<p>There is definitely <i>some</i> exaggeration around timelines (2030 instead of 2027) and the industry has a cartel but most of it is real demand. From AI companies that don't have the money to pay for it and probably won't ever receive actual shipments but they locked down contracts ahead of time.
I don't think that an abundance of any computing resource is a good thing. Cloud, storage, memory: this idea that you can always get lots of it leads to all sorts of undesirable outcomes.
Was expecting the article to go deep into DDR3 land by tone of the title...<p>Something like 3200mhz DDR4 is still perfectly fine for 90% of consumer use. You really don't need more to look at cat pictures on facebook.
The article doesn't go into detail about it, but if you go to eBay and find a really old Dell rack server that that does use DDR3 ECC Registered (such as an R910 or R920 with 512GB in it), they have <i>significantly</i> gained in value over the past 12-18 months. Or like an R720 with 256 or 512GB in it. Also major price jump in resale value. That's what we have come to now.
Yeah but TBH an old server with DDR3 ECC is still way sufficient to serve something like 99.9% of all the servers out there's jobs.<p>That's the plain truth: most sites/servers simply serve very very very little and have not anywhere near the need for 256 GB of RAM, let alone 512 GB.<p>These old servers aren't sufficient for the Amazon/Apple/Microsoft/Meta/Google/Oracle of this world, but for 99.9% of the rest of the world, they're more than plenty.<p>And the world and technology shall keep on marching: we'll have DDR6 and as soon as that happens, suddenly all the hyperscalers shall consider DDR5 to be oh-so-lame and dumped by the truckloads in landfills.<p>We may have two more painful years ahead for RAM but we'll then have plenty of used enterprise/datacenters gear on the market.
Actually, because it is absolute fine DDR4 prices are also rising. The only thing I found I could find for a decent price is unbuffered server ram or laptop ram.<p>I think the reason why those processors are produced are because it rather makes sense to upgrade your old rig with them. Seems to be the only way actually to increase CPU sales beyond laptops ans smart phones.<p>I don't see more DDR4 ram being produced. Chinese brands are already starting to produce DDR5 and soon HBM. 'Bringing back DDR4' is IMHO a too bold claim.
I actually recently bought DDR3 memory: $150 for 32GB for an older computer. The price is so much cheaper than what I expected. (My expectation of current RAM price is probably pretty out of whack due to market conditions.)
My old laptop has this exact RAM. Funny thing is that in terms of throughput it's on par with my current one, where I put just one 32GB stick, with one slot left for upgrades.<p>I still use it occasionally and it's really the old CPU and worn out SSD which slow it down.
That 90% probably also includes computer games. I'm not sure whether those need fast main memory. They certainly need fast VRAM.
DDR4 have similar latency as DDR5 and that's what matters in all games, while raw bandwidth only matters in some games
Most games do not, unless you want the best graphics. But then you’re out of the 90% use cases
I was eyeing a 96GB ddr5 so-dimm upgrade for my laptop last year, it was about 200-250 euro at the time. The only offerings I find for it now are in the range of 2000 euro. Regretting that I didn't max out ssd/ram at the time, but I'm optimistic the situation will get better in a year or so.
This mentions DDR4, but I used to be in electronic recycling and I have still so much 1GB and 2GB DDR2 and DDR3 RAM. Happy to ship a couple of sticks if you guys could find a use. It’s going to scrap otherwise.
A friend and I bought RAM for our aging servers at eBay in 2024. That was 512GB (16x32GB -- we split it equally) of DDR4 ECC 2133MHz, registered. For… $536.86 plus shipping.<p>Somehow it feels like we got a pretty good deal, in retrospect. (My server has been since replaced, and I can tell you, 256GB of DDR5 ECC was more than that! There's been a further doubling or two of prices since I bought, though. I think I could maybe sell half the RAM and finance the entire server…)
Currently 128GB of normal non-ECC DDR5 as a 2x64 set is around 2000 euro and increasing...
I know the only producer of such 64GB modules: Crucial (Micron), but Micron doesn't sell new chips to Crucial, thus almost no chance to get them anymore (though few months ago there were some). Kingston (using Micron, Hynix and Samsung) still does not have 64GB.<p>Soon Zen 6 should become available (well, or not so soon). With 7200 MTs memory controller? I haven't seen such JEDEC modules at all. Are they possible without the additional clock driver (like on 6400 CUDIMMs)?
One would think just some classes and forms of RAM increased in price. I have 128 GB (2 x 64GB DDR5 SO-DIMM) in my home NAS. Home NAS. Those nice ugreen DXP8800 Plus thingamabobs. It'd be over 2k euros for RAM alone now. Ridiculous. Add to that two SN850X 4TB SSDs, and 8 (+1 unpacked spare) IronWolves Pro and it's now a serious worth.<p>No end in sight.
2x16GB costs around $190 right now, or closer to $375 if you want ECC.<p>Not much cheaper than DDR5.
The biggest advantage is that the older technology can also produced by Chinese manufacturers like CXMT. They're ramping up production to claim a sizeable portion of the market.
They are already able to produce state-of-the-art DRAM, not only old DDR4 or slower DDR5, but they have difficulties to ramp up production for such DRAM, for not being able to buy adequate equipment due to the US "sanctions".
Not even DDR4... Things are this bad:<p>It's quite preposterous to think that an old 256GB RAM DDR3-using dual xeon server (128GB per NUMA node) that I had sitting around on a shelf gathering dust significantly appreciated in value over the past 12 months. But it's factually correct, looking at what comparables sell for for on eBay.
It’s making ddr3 relevant again as well. Bought 256GB DDR3 ECC for ~200 usd a couple weeks ago for my ai server (wip).<p>These old workstation platforms seem to be perfectly fine for hosting GPUs for local llms.
The big companies owe us money here. I call it a heist.<p>Also, the article makes no sense. Why would I want to pay
more for DDR4 now? They should tax those AI companies
that are responsible for this.
I'm posting from a machine with 8GiB of DDR3-1333. The modern web is slow, but I avoid the modern web as much as possible anyway. The software I actually enjoy using is mostly fast enough. Swap on zram (LZ4 compression, vm.page-cluster = 0, earlyoom userspace OOM killer, no disk swap) makes it feel like 9GiB with no perceptible downside.
Fact is that performance intensive code, games or whatever, has always optimised locality and so RAM speed shouldn't usually be a 1:1 impact on performance. Using older generations of slower RAM will be acceptable for many games and workloads.
Don’t call it a comeback, I’ve been here for years!
<a href="https://archive.is/yrGeb" rel="nofollow">https://archive.is/yrGeb</a>
Hmm. That might push me further towards just decomissioning this machine and using the laptop with a dock instead of rebuilding with Linux. The 64Gb of DDR4 in it might be worth a fair bit second-hand more than I assumed. I'll have to scan my box of old parts too, I'm sure there is some DDR4 in there (or maybe that is old enough to be DDR3…).
I recently reactivated a 10 years old Kindle as an AI-Story-Generator (<a href="https://github.com/donmb1/ai-story-generator" rel="nofollow">https://github.com/donmb1/ai-story-generator</a>). Old hardware is valueable and with AI you can do great things with it.
So LLMs are simultaneously making new hardware unaffordable, while also making old hardware usable again.<p>That too sounds like an actually good development all things considered. For the planet, anyway. And with the big caveat of that cloud tether, though just for scaffolding.
Old e-readers were already usable. Generating extruded story-like product on them when they're already perfectly capable of displaying the vast majority of human literature doesn't make them any more usable.
Also making old hardware unaffordable too.
> That too sounds like an actually good development all things considered<p>It would be good if we would use less resources.<p>But we are not.<p>It's just that some people don't get the newer resources.
Thanks for sharing, maybe it's time to buy some more old Kindles... I use bunch as smart paintings (with changing pictures generated locally with stable diffusion), and of course Home Assistant dashboard. They are awesome little machines.
Look at CrowPanel, you can get nice e-ink devices made for development, at a reasonable price. I have a small 2.13in esp32 board, I spent some time yesterday making it into a pocket flashcard device for my German courses. The energy consumption is minimal, small form factor, easy to flash via esptool or arduino utils
Thanks for the tip, I've seen those small e-ink display on sale in Aliexpress, but wasn't sure about their quality. Maybe I'll get one for my next project, I can see many use cases.
This is interesting. Maybe add a couple examples of generated stories.
Please. There are so many human stories out there—read those.
unlesss it's for children, just read books.<p>and if it is for your child, read them books.
Please don't raise your kids on AI slop. They'll grow up without knowing anything else.
I think in upcoming days everything gonna be making comeback previously like one or two years ago I was thinking the mobile price are so dropped that mobile phones gonna be like as cheap as possible but now after the AI their price are going as high as possible and same with the computers and laptops and all.
Maybe more efficient code will make a comeback.
How many prompts include "make it as efficient as possible"? Or will efficiency clash with shipping as many features as possible(not even humanly anymore).
I have bridge to sell you if you think that is likely.
We can get refurbished phones with 8 GB LPDDR5 for less than regular DDR5 modules.<p>LPDDR5 is slower but maybe a small cluster could be useful?
The wages of the West's deliberate deindustrialization and deliberate dependence on Asian manufacturing.<p>30 refineries were closed in Europe in the past 15 years about, and that is just a small example of a pathological situation.
I was reading that oil from the Middle East had to go to South Korea to be refined before being shipped as diesel to Australia.<p>Cost of hiring an oil tanker has ballooned massively (like 40 fold) since America decided to bomb its supply chain, so these long supply chains are massively expensive.<p>The Republican Party has decided that running a world of global free trade was no longer desired. America wont pay the price for another decade or two, but it will.
Well capitalism is responsible for that.<p>I think in a few hundred years we'll have learnt what a disgusting, poisonous system it is.
I think in a few hundred years the very small number of survivors of humanity will be living in caves and huts in the high Arctic and in a greening slowly-reafforesting Antarctic, they'll be hunter-gatherers with maybe some simple small-scale unmechanised agriculture, and they won't have money. Maybe a barter system for a few.<p>Among the many other things they won't have is an economic model.
Europeans and Americans choose to make things unsustainable and unprofitable to do in Europe, and chose to make it possible to import things cheaply from places like China and India who use slave labour and destroy the environment … and then blame Capitalism.<p>As you like … but blaming capitalism is not going to bring industry back, and it won't stop China from using slave labour or India from using their rivers as landfills.
I mean arguably, this is good.<p>The amount of waste produced due to phones not receiving new batteries and software updates was utterly insane, given what goes into making them.
100% agree. This is generally a good news story. The e-waste issue is one of the darker aspects of tech which is mostly ignored.
We should promote brands like Fairphone and Framework that are trying to make a difference here. Reuse more of your kit!
It's good if you're a tech guy in silicon valley with a good salary who used to upgrade every year to the latest pro max.<p>It's terrible if you're a labourer in India who buys phones under 10k INR (~100 USD) when your salary is 20k INR per month.<p>Unfortunately there are way more in the latter category than the former, and most of them won't know what ram is. They'll just be paying more and getting less going forward for a few years.<p>(Source, the salaries and phones I've seen of the staff at office. And we're in a major region you've likely heard of, not some tiny rural village.)
I have a Dell R740xd that is the core of my home lab, with 384 GB of DDR4 ECC. That was by far the most expensive part of the server. I get a lot of use out of it, but it certainly contributes to a higher-than-average power bill. I'm working on solar at some point next year.
I’m trying to find a use for DDR3 RAM which is like $50 or less for 64GB but I found out most things are incompatible
I just bought a ~10 years old SuperMicro 1U server with 2x 24 cores of Intel's E5. For ... 250 Euros. Really cheap, hehe.
Then I had to buy second-hand 128GB DDR4 ECC for 400 Euro and 4x 3,84 TB Datacenter SSDs for 2000 Euros, second-hand of course.<p>I mean it somehow still feels cheep, if I watch the prices for new hardware.<p>But it took me days to create this upper mentioned system from all the different sellers and different sources on eBay. There is no "cheap" one-click solution with included guarantee anymore.
It was obvious AI would further stratify the haves and have-nots, but I was too naive to realize the tech of the have-nots could literally just regress. Has such a thing ever happened before in tech?
Sorta, kinda. For a long time, a new PC tended to be built with new parts. New motherboard, new storage, new CPU, new graphics card. New.<p>During every mining crunch that made GPUs very expensive, we sometimes went backwards on parts of new builds. A used GPU from years ago was often a more-viable choice than a new model in any kind of condition.<p>edit to add: A lot of folks in 3D printing like to have a small computer driving the machine, especially with tinkerer-friendly hobbyist printers of yore. That often meant some manner of Raspberry Pi because it was inexpensive-<i>enough</i>, small, and a known quantity that was simple to deal with in an appliance-like way.<p>During the covid-era chip shortages, Pi hardware was expensive, hard to find, or often both. So folks started driving their printers with things like old cell phones or cheap Android TV boxes instead.
I remember getting a Radeon card on sale for about $180AUD in 2012ish. And 4 years later looking for an upgrade and discovering the cheapest higher performing card was at least $300 Aud.<p>I eventually upgraded to an RX 570 in 2018, and hit the same problem 5 years later.<p>So as far as loe to mid range cards are concerned you can wait 5 years and still have to pay more to replace it for an upgrade.
Yes, the previous fad (crypto) killed consumer GPU prices for good. I don't believe anything has compared on price-performance to a 1080 in the 10 years since.<p>Also the enshittification of software might count? You used to own Photoshop for life, now you're expected to pay a subscription for life. And legal onslaught, too. Encryption and consumer VPNs are on the chopping block in much of the world at the rate things are going. The end goal of governments and corporations alike is to remove the ability of consumers to own hardware they control. We are rapidly moving towards a future where you will only be able to access the internet and government services with a "trusted device", which will be a thin client with a TPM, biometric verification, telemetry of everything you do on it, and which will require paying an ongoing subscription to use.
As far as I can tell, in terms of MSRP, 980ti was $649 at launch. 1080 and 1080ti were both $699 when they hit the market.<p>$699 in 2015 (when the 1080 was new) is ~$975 today, adjusted for inflation.<p>$975 is pretty close to the launch MSRP of a 5080, which is "just" $999. (We can't actually buy them for $999, of course, but we couldn't usually get 1080 cards at MSRP during the mining frenzy a decade ago either.)
The big difference here is that by the time the 2080 had saturated the market, you could regularly find a 1080 for ~$300 (I paid $275 for an EVGA B-stock model).<p>Whereas today, with the 5000 series in the market for quite a while already, a 4080 is still well over MSRP
Back in the day 1080 was top of the line. For couple of generations now it's xx90 lineup.<p>You could say that there was Titan but in practice they were not seriously considered. At least for gaming.
I don’t see that at all obvious.<p>Businesses are competing like crazy to make algorithms that are in high demand. AI could theoretically let people and create new things that they would never be able to otherwise.<p>It is entitled to thing that people wanting GPUs and new PCs for gaming and hobbies is a more valid use. And if someone values it, all they have to do is pay the market price.
<a href="https://archive.ph/yrGeb" rel="nofollow">https://archive.ph/yrGeb</a>
Only a couple of my work machines use DDR4. Most of my own kit, both for work and for play, uses DDR3 which is cheap and plentiful.<p>I avoid dedicated GPUs which have always been a money-pit. I find modern video games dull and boring, and this is not offset by them being very <i>very</i> pretty. Gaming for me peaked with Quake II and there's been nothing of real interest since Portal and Untitled Goose Game.
I have a laptop with soldered RAM.
The motherboard upgrade to a 16GB variant was 120 EUR. Same price as the RAM soldered on it.
It came with a CPU upgrade as well. I'm happy.
It is not just DDR4; I know of at least one supplier sampling units with DDR3 for their ONTs.
As an owner of a 1700 socket mainboard and 64GB of DDR4 (recently acquired a second 32GB kit on eBay), I am very pleased to see this.
Love the link preview tagline. Thanks TheVerge<p>"Somehow, DDR4 has returned."<p>__<p>Besides that, wasn't DDR4 good enough for everything we did, anyway?<p>At least I did not notice all that much of a difference when migrating to my current DDR5 system other than the CPU being massively more performant. Though I am sure that the memory did play some role there.<p>Clanker claims that the felt absolute latency between decent DDR4 and DDR5 should be similar, while bandwidth might be lower, apparently? Something something CAS latency.
For daily office use there is much more benefit in having more memory than having faster memory. Because apps all started using more memory. Especially anything Google Chrome related.<p>If you have some unused extra memory it will be used as disk cache, which also helps in the responsiveness of a desktop. Waiting for the disk is orders of magnitude slower than reading from memory cache.
I would be interested to see a comparison of the inflation caused by the AI rollout versus the actual value generated by the use of AI. I really wouldn’t be surprised if the later is dwarfed by the former (that’s assuming integrating AI in a company generates significant economic value in the first place, which is still speculative as far as I’m aware)
I would like to see a Raspberry PI with DDR3 RAM. The price is insane right now.
This will drive linux adoption like nothing else.
I just sold 4x8GB of DDR4 that I wasn't using for $200 on Ebay so this article is quite timely
<a href="https://archive.is/yrGeb" rel="nofollow">https://archive.is/yrGeb</a>
>But at least it’s a good time to hang onto your DDR4 RAM.<p>A year and a half ago I bought 2 new 8G DDR4 for 47 USD. I was going to use it to upgrade my W541 from 16G to 32G. The W541 has 4 slots. But seems the one I have maxs out at 16.<p>Now I have another reason t hang on to the RAM :)
My idea was taking 20+ year old designs that are patent-free, implementing them on any modern nodes that work (esp MOSIS-supported), their performance per watt will probably be better, and just use those. Some of the old designs, like NonStop architecture or MPP clusters, still have amazing benefits.
row, row, row your hammer, gently down the bit flip stream...
My Home PC is from 2019, it has DDR4, it gets most things done, and I have never felt it was sluggish or anything. Maybe not the best gaming rig, but I get by.
I built my current home pc on the cusp of the transition of ddr4 to ddr5 and ddr5 was at the time out of my budget so I went with ddr4 and I can still play modern games fine with 32gb even at 2k.
I might not max out everything but that was never a priority for me. Usually besides resolution I seldom touch the graphical settings if the game is running ok.
In fact the only thing I upgraded one it until now was my GPU since that was a holdover from my previous one.<p>Home server also runs ddr4 and does ok, but that has been hitting more cpu bottlenecks now rather than memory.<p>That said its kinda sad personal users are getting tossed back a generation due to current generation ram being almost all gobbled up for AI.
We are getting tossed back a generation, but we are also becoming as rare as chicken teeth. Most people do not own a Desktop PC, most people don't even own a Laptop any longer. To do the most good for the most people requires datacentres at the moment, and it will probably remain like that for the forseeable future.
The crazy thing is knowing so many of the produced GPUs are just sitting in storage collecting dust because they don't have anywhere to plug them in and even where they have space do not have the power to run them. So basically the whole RAM situation is driven by things that are sitting dark in storage.
I am still using the machine I built in 2017.. Prices for GPUs have been consistently insane since shortly after then, (NFT-idiots, crypto-bros, now AI-twats) and now ram.. My work setup uses 100% of my ram now though. I'd like to double my ram by getting 2 more of the same sticks I have in it.. But the <i>exact same</i> sticks of DDR4 now cost <i>more</i> than they did in 2017.
[flagged]
[flagged]
Please consider being more civil when you post, you’ll probably find it more enjoyable in the long run.<p>The archive link was posted [1] long before you posted. Also from the HN FAQ [2]:<p>> Are paywalls ok?<p>It's ok to post stories from sites with paywalls that have workarounds.<p>In comments, it's ok to ask how to read an article and to help other users do so. But please don't post complaints about paywalls. Those are off topic.<p>[1] <a href="https://news.ycombinator.com/item?id=50040452">https://news.ycombinator.com/item?id=50040452</a><p>[2] <a href="https://news.ycombinator.com/newsfaq.html">https://news.ycombinator.com/newsfaq.html</a>