I work in e-waste recycling. Ever since the TurboQuant paper in March, I haven't been able to sell any DDR3. I'm guessing that the DDR2 and 3 this article is referring to is the actual memory chips, not modules/sticks that servers, desktops, laptops, etc. use, because the latter aren't moving.
Yep. Don't expect to sell those sticks on ebay at great price. Those new chips will be likely soldered to appliances like low end routers/APs, set top boxes, various adapters, low end systems, PLCs, IPBX, NVRs and various embedded devices.
I sold 7,2 Kg of DDR1/2/3 sticks two month ago, for gold recovery. As well as expansion cards, hdd PCBs and a few other things. Got about $600 from this.
Wild guess, but maybe China has something to do with that? They've got a huge "recover->break down/strip->recondition->sell refurbs to manufacturers" industry pipeline that doesn't seem to much exist outside of China.
Indeed, I was lucky that my PC is old enough to use DDR3 sticks, when I decided to upgrade a couple months ago. I think it's still cheaper to max out your RAM than to buy a new PC.
I thought the same until I calculated the cost of running DDR3. We’re at 0.35€/kWh so it adds up fast. Upgrading the motherboard and getting LPDDR4 sticks would be a net positive before the RAM prices soared.
Maybe you have priced it wrong? I just checked Ebay, a 16GB 12800 Registered ECC module goes for 40-50USD ea. That is crazy! Last year they were like 5 USD each.
Except almost nobody buys them, even last years for 10 bucks each. That's almost useless ECC Reg memory for HPE Gen 8 servers and workstations (from before late 2015 / start of 2016 with the introduction of the Gen9 using DDR4).
ECC unbuffered DIMMs (9 memory chips per side, no reg buffer/controller) is less available, quite widely used on level entry systems and thus costs a lot more even second hand.
Yeah, I only bought single sticks at a time as I saw good deals on them. I only needed six sticks for three laptops. Just stuck a bunch of low bids and grabbed whatever I could. At the time there were tons going for $5-10 every day. Trying to stick to ones with free postage otherwise the postage would cost as much as the RAM back then.
Worldwide chip pricing and it's effects on the overall economy are fascinating. I expected by now somebody with a gen -2 class VLSI plant would swing into action and make <thing> for 2/3 the cost of the majors, and clean up in volume as the market absorbs the price shock. But no, instead it suits everyone in the pipeline to whine about it, but mark up prices instead.
I am guessing the other side of this, the price drops will happen but slowly, and just like gas pricing, the profit is in rapid reaction to shortage and slow reaction to competition returning.
Chip pricing a sawtooth would make a LOT of money for somebody.
Well, you are working from all sorts of misguided assumptions about how you convert, the yield, the efficiency, the shared capital, and history.
Old gen logic barely exists, and won't convert. The tools are wrong so it wouldn't yield. And there isnt the expertise (on tool recipes and integration) to do it.
Historically, memory is a sawtooth business. But history isn't a great guide here: 10 years ago, a new plant added meaningful capacity, as it came with a shrink of about 30%. Today....it doesn't. So it take huge capital to add a very small amount of capacity.
You can dream that things like 3d dram and 4f cells will help, but they are unlikely to offer enough with demand.
And finally, everyone running a dram plant has lived through these capacity boom bust cycles and the consequently layoffs and pay cuts. I suspect they are happy to take money this time. I would be in their place.
I do hope we see a new dram or dram-like designs. And that there are more efficient dedicated ai processors. And that models themselves become more memory efficient. But I don't really believe any of these come soon.
I think a lot of what you say is true, but we're facing shortages for DDR3/DDR4 RAM and this isn't about increasing density or speed, it's about the pricing of what is a high yield functional commodity in todays manufacturing.
So the questions about 3D and 4D ram, they just aren't applicable.
Somebody should be making Samsung/Hynix ram chips and assembling them onto carriers 24/7 to sell to ordinary people. Instead, the entire production capacity is selling to Hyperscalers and going to AI.
Unfortunately from what I understood a production line that makes "normal" digital chips generally can't make memory cells with concrete physical differences in the process pipeline
Thats the kind of process issue I hadn't understood. I guess the dynamics here don't make operating a "2nd gen behind" system to make memory economic in the good times, and so nobody is ready to ramp up to meet market shortages in the short term.
EU is paying to build out "lo fi" chips for car and other needs: They decided the impact on domestic industry of supply chain logistics to TSMC was bad. A lot of people are shouting "why aren't they doing extreme UV 1nm chip design" when the decision was pretty simple: its possible to source the machines, it's probably faster to get to high yield, the exposure to supply chain risk is real, the return is hopefully high in strategic terms, you can improve density as a follow-on.
The memory crunch started October 2025. DDR6 will come out in 2027. At that point you're investing new capacity into an obsolete technology.
If I was a memory company, I would try to bring DDR6 to the market as soon as possible. DDR6 allows high end CPU based server platforms to reach the memory bandwidth of an A100 or half a H100 without the costly HBM.
By 2006 when the PS3 came 'round, the NES was definitely retro. It wasn't a livingroom staple anymore, and examples that survived were well-loved (either in the wear-and-tear way, in the appreciation-of-objects way, or both).
The NES was only ~21 years old at the time that the PS3 came out.
I was about to say. DDR 2/3 I can still see many use cases for that stuff in modern-ish tech.
I am currently looking for some EDO sticks for a really well preserved IBM Aptiva I found on the side of the road. That stuff is expensive for a very different reason.
There might be some Aptiva EDO RAM sitting in my parents basement... Ill have to look next time im out there. Im not sure if I ripped it out and stored it before we sent the computer off to recycling eons ago.
Okay, but... Are the new RAM players coming to the game? Or we'll be unable to afford a smartphone next year?
There were news that China has set up a new line, but tbh it's really bad that only a few companies are buying the ram at low low prices while others suffer. Economy and the invisible hand of the free market are failing their purpose.
Recently I had been struggling with a computer that kept crashing randomly. I finally figured out it was a bad stick of DDR4. I jokingly said to some friends I should put it on Facebook marketplace for free for repair. Maybe I should, who knows maybe someone can reflow the chips?
Sell it outright on eBay. Someone will pull off the good chips and put them in a product.
It really is that bad out there for small businesses trying to make consumer goods. At work we had to recycle prototypes to salvage memory for a production run
Does anybody have an understanding of when realistically there will be supply again? I've heard some positive sentiment about new developments in China, but that's not a market I am familiar with, and I haven't witnessed any effects of it thus far.
People act like RAM is thousands of dollars now. I checked prices recently and while it’s pretty high, it’s not that bad for a component you just buy once and use for a long time. I’m still more concerned about the prices and availability of decent GPUs.
RAM prices are so bad that my business couldn't launch our product this year. RAM increases completely obliterated our profit margin, and would have required raising prices 30%. For a small business, that just doesn't work.
A lot of small businesses are getting absolutely wrecked by this.
I sold 7,2 Kg of DDR1/2/3 sticks two month ago, for gold recovery. As well as expansion cards, hdd PCBs and a few other things. Got about $600 from this.
ECC unbuffered DIMMs (9 memory chips per side, no reg buffer/controller) is less available, quite widely used on level entry systems and thus costs a lot more even second hand.
Here's one I found in my email:
https://imgur.com/a/YWYpuzp
I am guessing the other side of this, the price drops will happen but slowly, and just like gas pricing, the profit is in rapid reaction to shortage and slow reaction to competition returning.
Chip pricing a sawtooth would make a LOT of money for somebody.
Old gen logic barely exists, and won't convert. The tools are wrong so it wouldn't yield. And there isnt the expertise (on tool recipes and integration) to do it.
Historically, memory is a sawtooth business. But history isn't a great guide here: 10 years ago, a new plant added meaningful capacity, as it came with a shrink of about 30%. Today....it doesn't. So it take huge capital to add a very small amount of capacity.
You can dream that things like 3d dram and 4f cells will help, but they are unlikely to offer enough with demand.
And finally, everyone running a dram plant has lived through these capacity boom bust cycles and the consequently layoffs and pay cuts. I suspect they are happy to take money this time. I would be in their place.
I do hope we see a new dram or dram-like designs. And that there are more efficient dedicated ai processors. And that models themselves become more memory efficient. But I don't really believe any of these come soon.
So the questions about 3D and 4D ram, they just aren't applicable.
Somebody should be making Samsung/Hynix ram chips and assembling them onto carriers 24/7 to sell to ordinary people. Instead, the entire production capacity is selling to Hyperscalers and going to AI.
(I could of course be wrong about even this)
Demand is up, prices are up, production is also up. The only "problem" is that you and I don't have disgusting amounts of money to buy in.
EU is paying to build out "lo fi" chips for car and other needs: They decided the impact on domestic industry of supply chain logistics to TSMC was bad. A lot of people are shouting "why aren't they doing extreme UV 1nm chip design" when the decision was pretty simple: its possible to source the machines, it's probably faster to get to high yield, the exposure to supply chain risk is real, the return is hopefully high in strategic terms, you can improve density as a follow-on.
No matter what ASML is doing well.
If I was a memory company, I would try to bring DDR6 to the market as soon as possible. DDR6 allows high end CPU based server platforms to reach the memory bandwidth of an A100 or half a H100 without the costly HBM.
DDR3 is not "retro", for chrissakes.
By 2006 when the PS3 came 'round, the NES was definitely retro. It wasn't a livingroom staple anymore, and examples that survived were well-loved (either in the wear-and-tear way, in the appreciation-of-objects way, or both).
The NES was only ~21 years old at the time that the PS3 came out.
It was 17 years old, or 14 years since wide distribution in the USA.
What counts as "retro" basically comes down to when the person you ask was born.
The final new international games were released in 1994, but Europe still got new games into 1995.
Japanese Famicoms were still being sold in Japan when the PS2 released. They sold more in Japan in 2001 than in 2000.
So no, calling the NES "retro" in 2000 wouldn't have made all too much sense.
I am currently looking for some EDO sticks for a really well preserved IBM Aptiva I found on the side of the road. That stuff is expensive for a very different reason.
There were news that China has set up a new line, but tbh it's really bad that only a few companies are buying the ram at low low prices while others suffer. Economy and the invisible hand of the free market are failing their purpose.
It really is that bad out there for small businesses trying to make consumer goods. At work we had to recycle prototypes to salvage memory for a production run
That is, unless the hyperscalers all drop dead overnight. Then we'll be swimming in more chips than anyone would care to buy
A lot of small businesses are getting absolutely wrecked by this.