• 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.

*** The AMD RDNA 4 Rumour Mill ***

Agreed, first attempt at AI Upscaling and it's already generally ahead of DLSS3 and not far behind DLSS4.

Given we were generally happy, except for bad implementations, with DLSS before the Transformer variant, then it says a lot.

Don't get me wrong, I'm still of the native + good AA is better camp, but both FSR4 and DLSS4 are good at what they do it seems!

The fact it automatically via a switch in the control panel, upgrades FSR3.1 titles (DLSS swapper style), then all they need to do is get FSR in more stuff and they're onto a winner. Make it easy to implement and they'll get results.
 
Last edited:
well the temps are sat at 85 degrees so I'm assuming that messing with the fan curve to reduce would just raise the temps? I'm curious because someone Techavidian (same card) suggesting his is nice and quiet under full load
85C just sitting there? If so that isn't normal. Almost like they forgot to install the thermal pad.
 
it may well raise the temp but worth checking - it might not need to run the fans at max to hold the temp at 85c for example.

edit: ah just saw your post above about trying it on silent mode and getting the same results. yea sounds like you might have a lemon there unfortunately. i'd still have a quick play with the fan curve and an undervolt though just to be sure. be nothing worse than returning it and ocuk or whoever you bought it from find no issue with it at their end (unlikely i know but stranger things have happened)
ok...

Load:
clock speed 2800 (+- 50hz)
fan speed 3157rpm (holds here, assuming this is 100%)
gpu temp 85 (+- 2degrees)
gpu hotspot 113 (moves between 110 and 117)

Idle:
clock speed 100
fan speed 0
gpu temp 48
gpu hotspot 49

I've taken a video, listen to those fans. lol

 
ok...

Load:
clock speed 2800 (+- 50hz)
fan speed 3157rpm (holds here, assuming this is 100%)
gpu temp 85 (+- 2degrees)
gpu hotspot 113 (moves between 110 and 117)

Idle:
clock speed 100
fan speed 0
gpu temp 48
gpu hotspot 49

I've taken a video, listen to those fans. lol

Yeah, I would be sending that back, day 1.
 
Agreed, first attempt at AI Upscaling and it's already generally ahead of DLSS3 and not far behind DLSS4.

Given we were generally happy, except for bad implementations, with DLSS before the Transformer variant, then it says a lot.

Don't get me wrong, I'm still of the native + good AA is better camp, but both FSR4 and DLSS4 are good at what they do it seems!

The fact it automatically via a switch in the control panel, upgrades FSR3.1 titles (DLSS swapper style), then all they need to do is get FSR in more stuff and they're onto a winner. Make it easy to implement and they'll get results.

It should be injectable into any DLSS game with Optiscaler. Waiting for the Optiscaler dev to add it and for 9070XT owners to test.
 
Honestly, if it works well, AMD should come up with an inhouse tool to do the same thing (or hire the guy who makes Optiscaler to do an official version) and then automate it in the drivers. Would be a big win for them if suddenly every FSR3.1 and every DLSS title could be converted automatically via a simple tick in the control panel to use FSR4 instead of DLSS; set your preset and then just it uses FSR4 instead of DLSS.

Have to wonder if they've already had that idea, unless there's a legal reason not to do it.
 
Last edited:
hmm a bit toasty.

My case has decent cooling
hmm a bit toasty.

My case has decent cooling

Apparently I was misinforming people that the reaper runs too hot 90C is toasty even if it is within spec- will hurt the vram long term- imagine putting your finger next to a boiled kettle at 90C for hours- eventually you get burnt? :D:D:D:D:D:D:D
 
Last edited:
So what? Like I said in the post in response to yours, it will depend on the game suite tested. If you focus on heavy RT games it won’t look great for either GPU. I mean I looked at tech spot and in Alan Wake 2 with quality upscaling at 1440p

(Min/avg)
5070 - 30/39 (identical to the older 4070)
9070 XT - 28/36

Neither are remotely playable. These kind of benchmarks in this level of GPU proves absolutely nothing.

Oh… wait, are you comparing a 5070Ti to a considerably cheaper GPU and using that for your RT comparison?

In tech spot at 1440p the 9070 XT and 5070 are essentially identical in RT. In raster the 9070XT is about 23% ahead for about 8% extra cost and 4GB more VRAM.

Yet other reviews have the 9070XT ahead of the 5070 by about 10% on RT and 25% ahead on raster.

*cough*, go check both yours and my original posts. We were talking about the 9070. I don't know why you are now going on about the XT.
 
It should be injectable into any DLSS game with Optiscaler. Waiting for the Optiscaler dev to add it and for 9070XT owners to test.
Does it work in reverse too, so you could use DLSS on any FSR3.1 game?
It would be really nice if we got to a point where they're both interchangeable so devs can just pick one rather than having to implement, well 3 if it can do the same with XeSS (or whatever the Intel one is).
 
Does it work in reverse too, so you could use DLSS on any FSR3.1 game?
It would be really nice if we got to a point where they're both interchangeable so devs can just pick one rather than having to implement, well 3 if it can do the same with XeSS (or whatever the Intel one is).

The latest Optiscaler (nightly build) does allow input from XeSS, FSR3.1 or DLSS so will work in any way you want. Which game is FSR 3.1 only though?
 
Last edited:
*cough*, go check both yours and my original posts. We were talking about the 9070. I don't know why you are now going on about the XT.

Because the 9070 XT is not much more expensive than a 9070 and unless you need the power efficiency, it is a better choice. You literally can’t really separate the two on price and as such both are actually directly comparable to the 5070. The prices of the 5070 are far more in line with the 9070 XT than the 9070.

If we are comparing GPUs in the £550 to £650 range then the 9070 XT qualifies and in general can be had cheaper than a 5070. It’s literally ignoring the elephant in the room to only consider the 9070 vs the 5070 as if they both exist in a bubble where the 9070 XT doesn’t look better on most metrics.
 
Last edited:
Because the 9070 XT is not much more expensive than a 9070 and unless you need the power efficiency, it is a better choice. You literally can’t really separate the two on price and as such both are actually directly comparable to the 5070. The prices of the 5070 are far more in line with the 9070 XT than the 9070.

Yup there's what £30-£50 difference? Pointless going for the non XT

If the non XT was £100 cheaper yeah that's ok
 
Thoughts on a 2080ti to 9070XT jump for 1440p? In all fairness the 2080ti does fine for what i want but im thinking i can prob get a £200 for the 2080ti? and throw £350-400 at a 9070xt for a cheapish upgrade? Just not too sure on how much of an uplift i'll see. I'm not too fussed about RT
 
Back
Top Bottom