• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Why does VRAM die so often?...

Turbonium

Turbonium

Platinum Member
I'm getting artifacting on my 6800 GT, which I'm assuming is due to the VRAM (GDDR3) going bad. This led me to ask the question: why does VRAM die so often, while system RAM almost never does?

- Is VRAM of lower quality than system RAM?
- Does VRAM operate at higher speeds, and thus at higher temperatures, than system RAM?
- Does GPU heat 'cross-over' to VRAM, due to 'single-piece' reference cooler designs?
 
Originally posted by: MarcVenice
Aren't you jumping to conclusions? Who sais vram dies more often?
Click to expand...
Not really. VRAM seems to die quite commonly after a few years, at least with this generation of cards.

On the other hand, I've virtually never heard of system RAM (run at stock) going bad.
 
Same here, the video ram on my X800XT started going bad after about two years. Underclocking it by 10 MHz to 490 MHz removed the artifacts for the remaining period I used the card. I've *never* had system RAM go bad on me even when overclocking and overvolting it.

That's not enough to draw any conclusions, but there tends to be a few posts from time to time about video cards going bad after a year or two.

I'm thinking it has to be the temperature... system RAM sits relatively isolated, usually directly in the airflow going from the front of the case to the rear, and barely gets warm. The RAM chips on my 4850 on the other hand sit almost right up against a GPU that runs at 72C even when idle.
 
Temperature is rarely the problem unless your talking about 125C+ where the material itself will breakdown. Most ICs can operate in ridiculous ambient heat levels. Also memory modules dont dissipate much heat.
 
Probably because GDDR is stressed more than system ram while being essentially the same product in architecture?

DDR3/GDDR3 for example: The fastest DDR3 system ram is 2200mhz, the GTX 285 uses 2500mhz DDR3 on a package that will almost always be hotter in the first place. Additional heat from the PWM + GPU in a smaller, cramped area, working at higher speeds, of course it's going to fail sooner, makes sense to me, not to mention, most peopel using system ram are not using the top spec 2200mhz ram, but rather 1333 or maybe 1600mhz ram, so there's a even bigger discrepancy between stresses.
 
[kidding]VRAM (dual ported ram) dies because they stopped producing it in the late 90s and the chance of failure after 15years is great.[/kidding]

vram is video ram but video ram is not necessarily vram.

 
Hogwash.

If your RAM on your video cards are more error prone than your system RAM, then it's because you are buying quality system RAM that has underwent extensive testing while the brands of video cards you are buying are getting their RAM chips from the bargain bin.

Never had a video card's ram go bad.
 
Originally posted by: reallyscrued
If your RAM on your video cards are more error prone than your system RAM, then it's because you are buying quality system RAM that has underwent extensive testing while the brands of video cards you are buying are getting their RAM chips from the bargain bin.
Click to expand...
Since when is eVGA bad? 🙁

btw, I should mention that this same card has artifacted before, about 4 years ago, during a very warm day while gaming. So I suppose it's possible that the memory was faulty to start with, though it's been working flawlessly ever since, up to last week at least.
 
It's the fault of these stupid cooling designers. They use same heatsink to cool the vram and GPU core that reaches 70-80C. When memory doesn't even reach that high of temperature.
 
The 6800gt, although revolutionary in a way, was released well before the fabrication process for that type of card was ready. I have a dead one right here. They were hot for their performance level.
 
Stop overclocking and your VRAM might live forever.

The only card I've had artifact (which is at least as likely to be GPU failure as VRAM) was a 6800GT, but it was using a custom passive / fanless cooling design by Gainward that wasn't good enough for this very hot GPU.
 
Originally posted by: Schmide
The 6800gt, although revolutionary in a way, was released well before the fabrication process for that type of card was ready. I have a dead one right here. They were hot for their performance level.
Click to expand...
Can you elaborate? I'm interested.
 
The 6800gt was a 130nm process 288mm^2 square using 55-65w comparable to the Athlon XP/Early Athlon64/Pentium4(Northwood). They typically ran into the low 80s to mid 90s under load. As thermal compounds break down, the heat goes up, and there was no throttling built into it. I have no hard numbers only a dead card and a few shared experiences with others.
 
Originally posted by: Schmide
The 6800gt was a 130nm process 288mm^2 square using 55-65w comparable to the Athlon XP/Early Athlon64/Pentium4(Northwood). They typically ran into the low 80s to mid 90s under load. As thermal compounds break down, the heat goes up, and there was no throttling built into it. I have no hard numbers only a dead card and a few shared experiences with others.
Click to expand...
So you're saying the same GPU produced today would be made using a smaller fab process, and preferably with throttling built-in, avoiding most, if not all, its heat-related problems? Or in other words, high heat over a long period of time will damage an IC, even though it's supposed to be able to tolerate high temps?

I remember an article on AT a long time ago about the expected lifetime of a CPU, based on its heat rating (TDP? not sure what this is called exactly), and cooling. It had lots of graphs and stuff. I didn't really understand most of it, but it was a fun read. I think it concluded that a chip will last longer the cooler it is, even if it is still within design limits should it run warmer; cooler is always better. If I'm wrong, please correct me on this.
 
VRAM is pushed to the limits; system memory much less so. Just looks at the speeds they are released at - when Video memory was going past 1 GHz, system memory was between 200 and 400 MHz.
-----

And as for the 6800GT being released before the thermals were under control - it's been that way pretty much since Transform & Lighting was introduced, and particularly since the FX 5xxx series of video cards. It became an arms race between Nvidia and ATI for the performance crown, and the definition of "acceptable" thermals and lifespan for most modern graphics cards would make an Intel or AMD (CPU division) engineer cringe.

GPU makers have the added constraint that, unlike CPU's, their GPU's face upside down (in most tower cases), with much inferior cooling to CPU's, and not the most ideal location to dump heat (below the GPU or outside a tiny hole in the back of the graphics card).
 
Originally posted by: Turbonium

This led me to ask the question: why does VRAM die so often, while system RAM almost never does?
Click to expand...
VRAM runs at much higher speeds and people tend to overclock it more. In particular, many often overclock to dangerous levels but it still works fine for a few months, so it seems like everything is okay. Until it craps out.
 
Originally posted by: Turbonium
On the other hand, I've virtually never heard of system RAM (run at stock) going bad.
Click to expand...

You've obviously never owned any Crucial Ballistix DDR2 from the past few years. (and bryanW1995 beat me to it)

I've also had some Corsair and A-Data RAM DDR RAM die on me.

Note that this doesn't mean I've stopped buying those brands.
 
Originally posted by: Schmide
Have you thought about cleaning the card and applying new thermal grease? (white)
Click to expand...
Too late now... already ordered a replacement card.

I might try that out though, should I find the time (not likely anytime soon).
 
System ram does fail, even the good stuff. I had a highend stick of corsair XMS die on me at stock speeds. In any case, you had a good run with that 6800. I think you can say you got your moneys worth out of it. This is as good an excuse for an upgrade as anything.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top