With Black Friday deals, some 32GB DDR5 RAM sticks are JUST a bit more expensive than a PS5 DE

LectureMaster

LectureMaster

Has Man Musk
IposJ39xHtPk2m6d.png

7XzA6REPiKXQzewe.png
 
AI demands so much RAM that AI boom skyrocketted the price. Im not much of a techy, but does AI really demand so much RAM?
 
Last edited:
Build now. Eat shit on the ram, save it back on some black friday deals. Get a machine built around a $500+ gpu now and you'll be able to use that for like 10 years because nobody's going to be able to afford to upgrade once the shit really hits the fan. The baseline is going to be stuck at PS5 for a long ass time. If you're on console only, maybe get that pro now.

If we have to halt somewhere for awhile, this is not a bad place to stop, though. I'm glad for that. Imagine getting stuck in the ps3 era when everything was choppy and raggedy.
 
You can still find some 32GB ram for under $200, but it used to be under $100. No idea how long this will last. If it doesn't normalise soon phones and consoles will start spiking in price as well.
 
yogaflame said:
AI demands so much RAM that AI boom skyrocketted the price. Im not much of a techy, but does AI really demand so much RAM?
Click to expand...
VRAM is required in enormous amounts. Even when you use a quantized version of AI (quantization reduces the precision of AI: for example, twice less size at the cost of 5 to 15% precision; models can be quantized even more aggressively, but the smaller AI is, the more its suffers from quantization - all the way down to becoming practically lobotomized), it is still ridiculous. Say, for example, you have a relatively dumb local AI model with 24 billion parameters. It can run on a single RTX 3090 (24GB VRAM) at 32k context length (it's about 20k characters). Some other model, let's say with 32 billion parameters, requires TWO 3090's to run at the same 32k context length (especially if the model is more complex and has more 'layers').

Now, things like ChatGPT have trillions of parameters and they probably run at full precision (no quantization). Just imagine how much memory on the server do they gobble up, all things considered - millions of users around the world making requests at the same time, and such giants also have a huge context window size (going from 32k to 1m would increase VRAM consumption shockingly high).

As for the actual RAM (not VRAM), there are models that have "total" parameters and "active" parameters: like 100B total, 20B active (MoE architecture: Mixture-of-Experts, where a few 'experts' activate and use parts of 'total' knowledge somehow). The former could be loaded into RAM while the latter (together with context) loads into VRAM - and it gives a rather slow (compared to full VRAM) but nonetheless usable experience.
Using ONLY RAM is never a good way, unless it's some kind of unified memory (think of iMac devices, or whatever they're called now - those with M1 - M5 chips).

Point is, it's going to get only worse, since those assholes developing AI can't find a way to make their models more efficient. Everything they release now aims to 'know everything' and they feed more and more training data into their AI. All enthusiasts who want to use AI locally, beg the developers to give them smaller and faster models. That's also where all the fears about "AI everywhere (like in Windows)" come from: people know that only gigantic, huge AI is somewhat smart. And you can't put such a thing in every PC around the world, not now in 2025 at least - so, if Microsoft ends up shoving some kind of agentic garbage down the people's throats, it's going to be erroneous, non-reliable trash that would cause more harm than good. Probably smarter than local 24B - 32B models, but still shit, since there's just not enough hardware and electricity to saturate the whole world with 'good' AI.
 
Last edited:
Autistic_Pancake said:
VRAM is required in enormous amounts. Even when you use a quantized version of AI (quantization reduces the precision of AI: twice less size at the cost of 5 to 15% precision), it is still ridiculous. Say, for example, you have a relatively dumb local AI model with 24 billion parameters. It can run on a single RTX 3090 (24GB VRAM) at 32k context length (it's about 20k characters). Some other model, let's say with 32 billion parameters, requires TWO 3090's to run at the same 32k context length (especially if the model is more complex and has more 'layers').

Now, things like ChatGPT have trillions of parameters and they probably run at full precision (no quantization). Just imagine how much memory on the server do they gobble up, all things considered - millions of users around the world making requests at the same time, and such giants also have a huge context window size (going from 32k to 1m would increase VRAM consumption shockingly high).

As for the actual RAM (not VRAM), there are models that have "total" parameters and "active" parameters: like 100B total, 20B active. The former could be loaded into RAM while the latter (together with context) loads into VRAM - and it gives a rather slow (compared to full VRAM) but nonetheless usable experience.
Using ONLY RAM is never a good way, unless it's some kind of unified memory (think of iMac devices, or whatever they're called now - those with M1 - M5 chips).

Point is, it's going to get only worse, since those assholes developing AI can't find a way to make their models more efficient. Everything they release now aims to 'know everything' and they feed more and more training data into their AI. All enthusiasts who want to use AI locally, beg the developers to give them smaller and faster models. That's also where all the fears about "AI everywhere (like in Windows)" come from: people know that only gigantic, huge AI is somewhat smart. And you can't put such a thing in every PC around the world, not now in 2025 at least - so, if Microsoft ends up shoving some kind of agentic garbage down the people's throats, it's going to be erroneous, non-reliable trash that would cause more harm than good. Probably smarter than local 24B - 32B models, but still shit, since there's just not enough hardware and electricity to saturate the whole world with 'good' AI.
Click to expand...
Very interesting what you wrote. I will research more about AI later. Got to sleep now and back to work later. Thanks.
 
Autistic_Pancake said:
As for the actual RAM (not VRAM), there are models that have "total" parameters and "active" parameters: like 100B total, 20B active. The former could be loaded into RAM while the latter (together with context) loads into VRAM - and it gives a rather slow (compared to full VRAM) but nonetheless usable experience.
Using ONLY RAM is never a good way, unless it's some kind of unified memory (think of iMac devices, or whatever they're called now - those with M1 - M5 chips).
Click to expand...
Good to know this. Makes sense cuz from what I've learned RAM+ CPU is way way inferior to VRAM+ GPU.
 
LUiTnxi4hxaeMAMk.jpg


My man Newegg sold me these for Black Friday. I think picking the highest priced 32 GB DDR5 kit you can see and comparing it instead of the lowest priced one is a little disingenuous, even if the lowest priced one is still almost double what they cost just a few months ago
 
So everything is being sucked up by AI servers, cloud servers whatever it is. Big corporations placing large orders and getting everything, first in line. The poor consumer who got you there is screwed, consoles probably get screwed eventually too.

Won't goverments think of the gamers for once? I can't wait to see the price of the Steambox now. This AI bubble can't burst soon enough.
 
The price of DDR5 is ridiculous right now.

The cheapest 32gb overclockers have in stock is £155. ($200)

Meanwhile I got 32gb of DDR4, pre-owned, for £40 ($50) for an AM4 build.

AM4 builds are still plenty good enough for gaming. Not worth paying so much more ATM.
 
Jinzo Prime said:
If you think these ram prices won't affect PS6 or PSN prices, I don't know what to tell you.
Click to expand...
They will but clearly it's not an imminent issue. If Sony thought it was going to be an issue for PS5 very soon they wouldn't be currently having aggressive discounts. They'd be trying to keep inventory so they don't have to manufacture as many PS5s with ludicrous prices.

Companies are trying to ramp up production, a couple of years from now it probably won't be quite as big a problem as it will be over the next few months.

 
Last edited:
the price is still too low for PCMR standard, should be like 600-700 to justify the PCMR status.
 
Jinzo Prime said:
If you think these ram prices won't affect PS6 or PSN prices, I don't know what to tell you.
Click to expand...
They porbably will next year around spring time, but in the meantime they managed to stockpile some ram before the craziness to last them months.
 
Last edited:
deriks said:
Uh... This price hyke is until next month or so, right?

Right?!
Click to expand...
As I was saying in the other thread, if you're thinking of building or upgrading soon you need to accelerate your timeframe and do it now

Prices aren't getting better in the short run on memory and they are about to get a lot worse on GPU and SSD soon
 
You must log in or register to reply here.

Similar threads

The NeoGAF Black Friday 2025 Deals Thread
Social Deals  2
89
870
Jinzo Prime replied
Nintendo reveals Black Friday and Cyber Monday 2025 sales and deals
News 
29
1K
MidGenRefresh replied
Nintendo Announces Holiday Offers for Black Friday and Cyber Monday 2025
Deals Game Dev 
40
544
coffinbirth replied
[The Verge] Sony is taking $100 off its PS5 consoles for Black Friday
2 3 4 5
209
562
Baemono replied
[IGN] PlayStation Is Winning the Black Friday Console War Simply By Discounting PS5
Opinion  2 3
105
206
RetroGamingUK replied
Top Bottom