• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

NVIDIA ‘Ampere’ 8nm Graphics Cards

What i'm saying is the perf/Watt didn't increase at all..and that's with going to 8nm process.
And the cost is not lower the cards are coming at the same price as the "overpriced" Turings..just now they bring better performance uplift.

These are good and powerful cards but not world changing, and not even close to the NV hype slides.

Not hugely surprising. The 12nm TSMC node used for Turing was really mature, and the Samsung 8nm node was a fallback and was rumoured to be unrefined. It was used to increase the transistor budget and not the power consumption.
 
Bitwit noticed a potential issue with the FE 3080:

https://youtu.be/UV8WfnevkiI?t=533

Your AIO tubing may end up interfering with the fan on the 3080. It looks like my power cables will interfere with the fan:

bHH2IS3.jpg
 
You seem confident with no proper reviews out yet. You say prices will drop below 3070 prices? but what happens when they is a massive shortage of stock with the 3070 / 3080 you saying the 2080ti won't rise in price? I like your theories on everything
With a stock shortage in Q4 then you are right the prices of a used 2080 Ti may well rise, but it will only be temporary until stock comes back again. The time to sell was really a few weeks before the announcement.

From what we have seen so far, even without reviews, it is pretty clear there is a significant enough performance uplift with the 3080 to make a 2080 Ti feel like a last generation card.
 
35.6% faster than a 2080ti in Firestrike Extreme
36.2% faster than a 2080ti in Timespy Extreme

Looks good to me and what you'd hope for given advancements in technology and a smaller process node. The games show a little less gains but it's a very small sample size at this stage.

33.3% faster Tomb Raider 4K
22.6% taster Tomb Raider 4K DLSS
22.7% faster FarCry 4K

12.5% faster than 2060S in Farcry 1080p :rolleyes:
 
12.5% faster than 2060S in Farcry 1080p :rolleyes:

That's got to be CPU limited, there isn't much difference between all the cards at 1080p.

To test a GPU these days, especially the top end, you need to be testing at a minimum of 1440p, otherwise you're going to start to show CPU bottlenecks because the GPU won't be running at full utilisation
 
That's got to be CPU limited, there isn't much difference between all the cards at 1080p.

To test a GPU these days, especially the top end, you need to be testing at a minimum of 1440p, otherwise you're going to start to show CPU bottlenecks because the GPU won't be running at full utilisation

So Nvidia want everyone to upgrade to 4k monitors to justify a GPU upgrade, on top of maybe needing a new PSU to deal with the 30-40% higher TDP :P
 
It’s amazing how quickly it went from “twice the performance” to “30% is not bad”. Hopefully Monday will bring better results.

I know right, but then I expected that when they presented it with the big "UP TO" quote in the presentation, they had to get that in there so the pitchfork people are not hunting them down come Monday.
 
That's got to be CPU limited, there isn't much difference between all the cards at 1080p.

To test a GPU these days, especially the top end, you need to be testing at a minimum of 1440p, otherwise you're going to start to show CPU bottlenecks because the GPU won't be running at full utilisation

Exactly, I don't game at 1080p anymore, at 1440p so I can't justify getting a card that will just do 1440p again with less GPU usage because sure as hell it ain't doing 4k 60 in triple AAA titles next year or even this year for that matter when cyberpunk decides to show itself.
 
Back
Top Bottom