Nvidia GeForce RTX 4080 Reviews

[LordEC911]

I'll end the offtopic on my part now too, but looks like the power brick is optional in that, it just leads to 6-pin PCIe-plug, meaning the card would work just fine without it as long as you have 6-pin PCIe from your PSU

--
More on topic, this is either genius or someone getting fired

https://twitter.com/i/web/status/1593307035768336384

I mean, seems like a pretty genius way to hit their target market of 16-45 year old males because if you get it, you get it and if you are outraged about it... you still get it.
I would have to guess that there might be some backlash/pushback once the media sees it and figures it out... which then might make some higher-ups upset.
On the other hand, it is a pretty well known meme. I think if they would have stayed with the first 10 secs and then just showed off the card by itself, it would have been fine.

 

[techuse]

I’m not getting it .

 

[itsmydamnation]

I’m not getting it .

this explains so much /s

so the first 5 seconds should give it away.

 

[digitalwanderer]

I’m not getting it .

Yeah me neither. Now I feel either old or stupid just not sure which yet.

 

[techuse]

this explains so much /s

so the first 5 seconds should give it away.

What is there to explain? I dont understand the video.

 

[Kaotik]

It's a porn reference.

 

[digitalwanderer]

It's a porn reference.

Thank you! That's what I was sort of thinking, but I couldn't quite figure the connection.

 

[hurleybird]

possibly nsfw explanation

 

[Cyan]

 

[DegustatoR]

NVIDIA has sold 160K GeForce RTX 40 graphics cards already, but stores are still full of RTX 4080s - VideoCardz.com

NVIDIA RTX 40 sales reports The RTX 4080 is not selling particularly well on the first week. First estimates on RTX 40 series sales have been shared by a few sources, but even despite impressive numbers retailers are worried it may not as popular choice as RTX 4090. According to Wccftech and...

videocardz.com

 

[Newguy]

The 4080's raster perf/SM improvement is very encouraging if it holds true for the rest of the stack. Rough range of 1.25-1.4x perf/SM using techpowerup, pcgh and computerbase's numbers with 3080, 3080Ti and 3090Ti comparisons at 1440p and 4k. So using middle of the road numbers I'd expect full AD107 (24 SM) to be roughly 1.15x 3060/0.9x 3060Ti, full AD106 (36 SM) roughly 3070Ti level +/- a little bit, or >1.5x perf vs their Ampere counterparts

 

[trinibwoy]

Is the 4080 in stock due to lower than expected demand or is it priced right? We’ve been conditioned to expect good products to sell out. It’ll be interesting to see the price of this thing in 6 months.

 

[pjbliverpool]

Is the 4080 in stock due to lower than expected demand or is it priced right? We’ve been conditioned to expect good products to sell out. It’ll be interesting to see the price of this thing in 6 months.

I'm in full agreement with HUB here. The 4080 is just a horribly priced part. It offers lower value than the halo product while likely being terrible value vs the 4070Ti assuming NV don't bollock up the pricing on that too. And that's without considering the AMD parts.

I can see myself potentially holding my nose and getting a 4070Ti in Jan provided it's $799 or lower and we don't get ripped off on the pricing in the UK. It'll still feel dirty spending so much on a x070 class product but at this stage I don't think things are going to get any better than that. AMD is out for me unfortunately as DLSS and RT performance are major selling points for me. FSR2 is actually great in the games I'm currently using it on, but with DLSS being even better, ad more importantly, more common, that's the way I have to go.

 

[homerdog]

I can see myself potentially holding my nose and getting a 4070Ti in Jan provided it's $799 or lower and we don't get ripped off on the pricing in the UK. It'll still feel dirty spending so much on a x070 class product but at this stage I don't think things are going to get any better than that. AMD is out for me unfortunately as DLSS and RT performance are major selling points for me. FSR2 is actually great in the games I'm currently using it on, but with DLSS being even better, ad more importantly, more common, that's the way I have to go.

$800 on a 192bit card. I guess this is where we are. And of course things aren't going to get better if AMD is off the table on account of DLSS.

 

[Below2D]

Around here, the 4090 disappears as soon as it's in stock. I went to my local retailer and there were more 4080s than any other card.

 

[DegustatoR]

$800 on a 192bit card. I guess this is where we are. And of course things aren't going to get better if AMD is off the table on account of DLSS.

Vega 64 should've been what $4000 with this logic?

Not saying that the prices are good or anything but it's about time to stop connecting pricing to irrelevant things like die sizes and memory bus widths.

 

[PSman1700]

GTX960 was 128bit, the 256bit gtx670 wasnt the better gpu for that.

 

[homerdog]

Vega 64 should've been what $4000 with this logic?

Not saying that the prices are good or anything but it's about time to stop connecting pricing to irrelevant things like die sizes and memory bus widths.

I understand that there are outliers (GT200, Hawaii, all the HBM stuff etc.), but this is not one of them. Looking at the memory bus is a useful way to determine where a product belongs in the stack.

 

[homerdog]

GTX960 was 128bit, the 256bit gtx670 wasnt the better gpu for that.

Also the RTX3050 is faster than the 7800GT.

 

[pjbliverpool]

On the one hand we could argue that it's a maxed out XX104 chip, the third gaming chip down in the stack which makes it the equivalent of the 3070Ti last gen.

On the other hand we could point out that the difference between AD102 and AD104 (140% more SM's) is far bigger than GA102 and GA104 (75% more SM's) and so in comparison to the top of the stack, the 4070Ti will actually be no-where near a 3070Ti.

But then assuming the 4070Ti comes in at around 3090 performance or a bit better - a 40% uplift roughly speaking over the 3070Ti, that's a pretty decent gen on gen uplift. However it's no-where near the 80%+ of the 4090 over the 3090 and so clearly Ada, being as spectacular an architecture as it is, is capable of so much more than Nvidia are giving us with it outside of the 4090 itself.

If that were all they were doing I think I could accept it, afterall, they are the ones that made such a good chip and if they want to take advantage of that engineering prowess to call a chip that should sit further down the stack a 3070Ti, then fair enough... I guess. But to then also charge $200 more than its namesake last gen, which was relatively much further up the stack feels like they're just screwing us over. This is all based on it being $799 of course. Release it for $599 and I'll be singing their praises despite my arguments above. Of course that will never happen.