• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

NVIDIA GeForce RTX 3080 Founders Edition

Can you please post information that has more credibility? Some random user on Reddit does not qualify.
the Zotac rep says it was intentional to make their more expensive cards look faster
 
Oh really? Maybe your math skills are really worse than your drawing one but you don't need more, just extend the line like you did earlier... even ignoring the leveling and drop that happen when you keep increasing the power are you seriously suggesting that Turing line on the graph would be like that or "even higher" to quote you?
Maybe in your fantasy world where actual physical laws don't apply...
View attachment 169101
but what are you saying?
simply the graph says that for the same watt, at 240watt Turing generates 60fps, Ampere 90fps, + 50% performance for the same watt and this does not happen in reality, they are lying...
wake up
 
As far as I know none are dual slot they're all triple slot solutions

Saw a 3090 blower the other day... so its possible, if you want to abandon all quality of life.

Oh really? Maybe your math skills are really worse than your drawing one but you don't need more, just extend the line like you did earlier... even ignoring the leveling and drop that happen when you keep increasing the power are you seriously suggesting that Turing line on the graph would be like that or "even higher" to quote you?
Maybe in your fantasy world where actual physical laws don't apply...
View attachment 169101

I'm not sure its wise to discuss this graph at all. It puts 'FPS on the Y and 'W' on the X without any further info telling us it massively improved.... yeah. Best case, RT enabled scenario since the first iteration was utterly shite. I'm sure in some weird niche the info is sound. But useful it is most certainly not.

I already have a 1080 SLI 3 years ago and it's great for 4k experience minimum 70 - 80 fps in Ultra in most games...

Not any games going forward as Nvidia is dropping SLI profile support entirely within 4 months (see headlines). Maybe they'll push some sort of community effort...?


I gotta say that overall the more I'm seeing this 3080 'mature' into the market and with the recent higher cap VRAM versions announced... I'm not very convinced with Ampere just yet. Let's pray for some real competition and movement. There's no real rush, zero content is out that actually truly needs the RT yet...
 
Fair enough. Not very surprising though. We're talking about base clocks. 1% - 3% is not much to whine about let alone make a mountain out of molehill over, especially when you can use a utility to OC and get that back plus more.
I don’t know — at the point when a sales rep admits that it’s a lower-end design, I think that means it’s a lower end design. Not really news, although bottom-barrel nvidia cards usually perform on par with FE cards. Still, for once I (partially) agree with @lexluthermiester — 1-3% is practically a margin of error. I’d be behooved not to correct you, though — even when OC’d the Zotac card performs worse than an FE overclocked, as some reputable reviewers have shown. Unlike the FE's 370W power limit, the Zotac card is limited to 336W.

In any case, you're right -- the performance difference is mostly negligible in the real world (unless you overclock). It's not at all unusual for AIBs to release different tiers of cards, and it's not like ASUS or Gigabyte released their high-end cards at launch (and MSI if there's a lightning). *shrug*

Relatively, the Zotac doesn’t overclock well and is slower than the FE OOB. It also seems to be missing some components that the FE has. It’s Zotac news for Zotac fans, and some frustrations for buyers who just wanted a "reference" PCB. Maybe the delay means they’ll go the extra mile with Amp! this round, but given the 3080's limitations I doubt it.
Hmm, should I listen to people with technical credibility, or ask a sales rep? :roll:
No random user anywhere qualifies. You need uh... qualifications. Paper or otherwise. Hard to find but very useful in debates to use people that have those. 10/10, would recommend.

Depends, are you looking at getting a job in sales, or asking technical questions about computer parts?

Depending on your answer, one may be more suitable than the other.
If you'll read the source in question before making assumptions, it was reviews like this one that got people upset:

Couple that with "issues" like these and you might see why early adapters are a bit miffed:


 
Last edited:
Saw a 3090 blower the other day... so its possible, if you want to abandon all quality of life.


yeah I saw that too unfortunately I think the only thing "Turbo" on that card is going to be the fan and the amount of noise it's going to make when you hit 70+ degrees C
 
yeah I saw that too unfortunately I think the only thing "Turbo" on that card is going to be the fan and the amount of noise it's going to make when you hit 70+ degrees C
at least you never have to worry about the freezing cold as with the 3090 you get an automatic free heater :)
 
yeah I saw that too unfortunately I think the only thing "Turbo" on that card is going to be the fan and the amount of noise it's going to make when you hit 70+ degrees C
@Vayra86 - I heard the blower is good for about .5 bar / 7 psi... will take 6th gen Honda Civic B16a2 motor to 300whp (intercooler required). :roll: :pimp:
 
at least you never have to worry about the freezing cold as with the 3090 you get an automatic free heater :)

HA LOL have you seen the absurd prices here in New Zealand $1,798.99 NZD for a 3080 you can bet your top dollar the 3090 will be twice that price @ $3,000+NZD
 
@W1zzard Are all the 3080 reviews done now?
The Temp and noise table in the review here needs updating still.
 
HA LOL have you seen the absurd prices here in New Zealand $1,798.99 NZD for a 3080 you can bet your top dollar the 3090 will be twice that price @ $3,000+NZD

Well I was right $3500+NZD for the 3090
 
Well I was right $3500+NZD for the 3090
$2,749 is the cheapest across the ditch in AU, how does that compare with the current exchange rate?
 
$2,749 is the cheapest across the ditch in AU, how does that compare with the current exchange rate?

Poorly very poorly @ $ 2,821.22NZD for one from Oz vs $3554.99NZD to buy here still waaaaaaaaaaay to expensive
 
Hell, we just need 3080 SE, Stock Edition.
 
Poorly very poorly @ $ 2,821.22NZD for one from Oz vs $3554.99NZD to buy here still waaaaaaaaaaay to expensive
If you're not in a hurry give it a few months. November is going to be a good time to buy. Drivers that fix a few issues that have been discover will be out and supplies will be restocked.

Updated, thanks for reminding
Is someone at TPU going to do a review on the 3090 or are you focusing on "other" reviews ATM?
 
You mean stock 3090 out of stock?
 
I have a problem with my nvidia 3080, while it runs perfectly fine if you don't overclock it tends to live in the 1800-1900mhz range. My issue is the Asus RTX 2080 Ti I also have that I am replacing seems to get much higher scores than what the benchmarks in the review have it as. I'm confused how with a simple power increase it runs between 2070-2130mhz core and 8000mhz mem while never exceeding the 60s.
I play in 4k with HDR (when available) and the FPS difference so far has a maximum of 9% Port Royal.. while all the other 3Dmark tests have the 2080 Ti slower by 0.04%-9%. Is this because it is more of an 'apples to apples' comparison due to the 2080 Ti running a similar power limit? I guess most reviews must use cards with low power limits and no power limit increase from afterburner or bios update.
Just find it weird as the 3080 has many more CUDA cores but I suppose it generally runs at a slower speed and memory speed isn't much different after overclock. Yet the 2080 Ti even with the overclock is 100% stable and runs cooler where minor overclocks have caused me issues with the 3080..
Guess I fell for the hype. Might sell the 3080 as it would be a significant upgrade from a non 2080 ti user
 
I have a problem with my nvidia 3080, while it runs perfectly fine if you don't overclock it tends to live in the 1800-1900mhz range. My issue is the Asus RTX 2080 Ti I also have that I am replacing seems to get much higher scores than what the benchmarks in the review have it as. I'm confused how with a simple power increase it runs between 2070-2130mhz core and 8000mhz mem while never exceeding the 60s.
I play in 4k with HDR (when available) and the FPS difference so far has a maximum of 9% Port Royal.. while all the other 3Dmark tests have the 2080 Ti slower by 0.04%-9%. Is this because it is more of an 'apples to apples' comparison due to the 2080 Ti running a similar power limit? I guess most reviews must use cards with low power limits and no power limit increase from afterburner or bios update.
Just find it weird as the 3080 has many more CUDA cores but I suppose it generally runs at a slower speed and memory speed isn't much different after overclock. Yet the 2080 Ti even with the overclock is 100% stable and runs cooler where minor overclocks have caused me issues with the 3080..
Guess I fell for the hype. Might sell the 3080 as it would be a significant upgrade from a non 2080 ti user
There are many instances of people saying that buying a 3080 to replace a 2080ti might yield results that you may be less that happy with. There is no doubt that the 3080 beats out the 2080ti. Might not be by much for some programs, but the 2080ti is not the victor in that competition. What you should have bought was the 3090. The 3090 is the replacement for the 2080ti and would have granted the general leap in performance you were expecting. As both were in the same price class as each other(about $1500), the 3090 was the logical choice.
 
I have a problem with my nvidia 3080, while it runs perfectly fine if you don't overclock it tends to live in the 1800-1900mhz range. My issue is the Asus RTX 2080 Ti I also have that I am replacing seems to get much higher scores than what the benchmarks in the review have it as. I'm confused how with a simple power increase it runs between 2070-2130mhz core and 8000mhz mem while never exceeding the 60s.
I play in 4k with HDR (when available) and the FPS difference so far has a maximum of 9% Port Royal.. while all the other 3Dmark tests have the 2080 Ti slower by 0.04%-9%. Is this because it is more of an 'apples to apples' comparison due to the 2080 Ti running a similar power limit? I guess most reviews must use cards with low power limits and no power limit increase from afterburner or bios update.
Just find it weird as the 3080 has many more CUDA cores but I suppose it generally runs at a slower speed and memory speed isn't much different after overclock. Yet the 2080 Ti even with the overclock is 100% stable and runs cooler where minor overclocks have caused me issues with the 3080..
Guess I fell for the hype. Might sell the 3080 as it would be a significant upgrade from a non 2080 ti user
At 4K is where you should see the most improvement. At an average of 26% (TPU review of Asus TUF), that is a generational leap. Yes, a 2080Ti overclocks more than these 3080's. A 3090 would have given you another 10-15% performance, but @ double the price, the value proposition (I use that term loosely) isn't there. Note according to NV, the 3090 is essentially a Titan replacement, as well as the 3080 being labeled the "flagship" in the same video). That said, I surely wouldn't return/sell the 3080. You can sell the 2080Ti for $500 and have a $3080 for ~$400. I'd call that an upgrade. :)

Depending on what benchmarks, the CPU in play may have something to do with the results too. 3DMark likes cores and threads, for example...
 
Last edited:
A 3090 would have given you another 10-15% performance, but @ double the price, the value proposition (I use that term loosely) isn't there.
While that a good point, the price paid for the 2080ti is the same as one would pay for a 3090. They are effectively in the same class for their respective generation.
Note according to NV, the 3090 is essentially a Titan replacement
Sort of. Jensen did call it "Titan class performance", but in reality the real Titan replacement will be a Quadro card. The 3090 is effectively the "ti" of this generation of GPU's, both when gauged by performance and by price.
 
While that a good point, the price paid for the 2080ti is the same as one would pay for a 3090. They are effectively in the same class for their respective generation.

Sort of. Jensen did call it "Titan class performance", but in reality the real Titan replacement will be a Quadro card. The 3090 is effectively the "ti" of this generation of GPU's, both when gauged by performance and by price.
What someone paid a couple of years back has little bearing (in my mind) on what is what today. Price and class are different things as both are moving targets. Clearly NV has dropped prices significantly so it's tough to compare like that.

A 3090 is a Titan replacement. It has (most of) the makings of a Titan, including inflated price, too much ram for gaming, and Jensen said so "Titan class". It just isn't called one (for marketing reasons). If you forced me to guess, I'd say it is to get better market segmentation. The 3090 today is more gaming than the Titan RTX, but it is still a crossover card, surely. If you look at price and performance, you'll see that a 3090 more closely resembles a Titan than it does 2080 to 2080Ti, for example. We see meager increase in performance (even at 4K), and shed loads of RAM. The difference between the 2080 and 2080 Ti was almost 25% whereas now 3080 to 3090 is ~15%. Barely a SKU difference. The titan to 2080Ti was a couple of % IIRC.

I'm also thinking if the 3090 was the Ti, they'd call it the Ti. The Ti of this gen is going to be a 3080 Ti down the pike. Similar to the GTX 780 and then the 780 Ti that came out later (with Titan on top of that). They just dropped the Titan name. :)
 
Last edited:
Back
Top