• 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.

When the Gpu's prices will go down ?

Ah success - last time I looked (a while ago) I couldn't find timespy scores for these new GPUs.

It looks like some scores have changed since, so I haven't updated the previous scores yet. But have been able to add in the new series.

Untitled.png


Its a far straighter line than the last series, so actually in terms of price vs performance scaling, its fairly linear. You can see here actually it supports the 3080 being the outlier of the previous gen, its way off the curve of the rest of the series.

There is no clear optimum product in the latest gen, nothing stands out as being above or below the linear relationship of price to performance.
 
Last edited:
danlightbulb said:
Ah success - last time I looked (a while ago) I couldn't find timespy scores for these new GPUs.

It looks like some scores have changed since, so I haven't updated the previous scores yet. But have been able to add in the new series.

Untitled.png


Its a far straighter line than the last series, so actually in terms of price vs performance scaling, its fairly linear. You can see here actually it supports the 3080 being the outlier of the previous gen, its way off the curve of the rest of the series.

There is no clear optimum product in the latest gen, nothing stands out as being above or below the linear relationship of price to performance.
Click to expand...
The 3090 Ti is slower than a 3090?
 
danlightbulb said:
Yes as I said, I have tried to update it but can't locate comparable timespy scores to those I have for previous generations.

If you have them please post them and I'd be happy to update it.
Click to expand...
Jensen knows what you are trying to do :). And hiding the evidence...
 
Last edited:
Eddie99 said:
The 3090 Ti is slower than a 3090?
Click to expand...
Murphy said:
4090 = 36154
4080 = 28125
4070 Ti = 22753

The 4070 results seem to be giving a server error, maybe it works for you.
Click to expand...

Problem is get different scores depending on source. Originally I got the 30 series scores from Jay2Cents youtube reviews. The older series scores from a website somewhere. The new scores come from 3Dmark website and differ to the above.

So the 3090Ti score is indeed lower on 3Dmark than Jay2Cents had in his review for the 3090. That's probably not right so I will get around to updating all the scores from the same source.
 
Last edited:
Calin Banc said:
Probably will use 10-12gb just fine - depending how vram hungry future games will be. But, since they couldn't get just 12gb easy on it... 16gb it is. Same with 3090 and its 24gb or variants from AMD that have 16gb or more which will be too much for what power the GPU has.
Click to expand...

Yeah funny thing is in Steve's/hub video, the 3080 actually comes out on top for 4k gaming despite having less vram than the 6800xt :p
 
Calin Banc said:
Probably will use 10-12gb just fine - depending how vram hungry future games will be. But, since they couldn't get just 12gb easy on it... 16gb it is. Same with 3090 and its 24gb or variants from AMD that have 16gb or more which will be too much for what power the GPU has.
Click to expand...
Yeah I mean it's essentially like a 3070-ish thing in terms of performance, and a 3070 does get bottlenecked by VRAM in some situations. So it follows that going bigger than 8GB can bring some benefit. It's not like they can arbitrarily stick e.g. 11GB on it or something so 16 makes some sense. Or it would if it wasn't for the grand canyon of a margin they're putting on the 16GB part.
 
danlightbulb said:
Ah..any excuse to fish out my old chart. If you straight line between the 3070 and 3090, the 3080 is higher performance at lower cost than would be expected.


353f9gh-png.410856
Click to expand...

I disagree. Your chart seems to show that Turing was lower performance than expected and Ampere simply got "back on track".

Turing looks like the anomaly on this chart.
 
Well they have done the same as they did for the 4070.

The 4070 performs similar to a 3080 for $100 less RRP
The 4060Ti performs similar to a 3070 for $100 less RRP (same vram though so even worse)

Basically giving us a $100 saving after 2 and a half years and trying to justify it with frame generation...: /
 
Jono8 said:
Well they have done the same as they did for the 4070.

The 4070 performs similar to a 3080 for $100 less RRP
The 4060Ti performs similar to a 3070 for $100 less RRP (same vram though so even worse)

Basically giving us a $100 saving after 2 and a half years and trying to justify it with frame generation...: /
Click to expand...
The 4070 is slower at 4k. This is what stagnation looks like.
relative-performance-3840-2160.png
 
Last edited:
Only 5% less than the 3080 is within striking distance IMO. You don't even have to overclock it that much to match it. The price is the problem
 
Seems like if you want the best deal wait a few months for someone to regret purchasing their pc or gpu otherwise you'll be waiting way too long for retailers to get anywhere close to the same price range. Of course with the risk of not knowing how they've used the gpu...

Dilemma I'm thinking about at the moment.
 
You must log in or register to reply here.
Back
Top Bottom