• 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.

Sanity check vefore I return my Radeon VII

Status
Not open for further replies.
Man of Honour
OP
Joined
19 Oct 2002
Posts
29,516
Location
Surrey
I think this has gone off topic somewhat. Thanks for the feedback so far. The general feedback seems to be to return the VII and get either of those two cards.
 
Associate
Joined
8 Jul 2013
Posts
2,089
Location
Middle age travellers site
Definitely, they'll be the best cards to come out of this NV/AMD release.


Agree looking forward to seeing using one of customs ...i have high hopes for this new architecture...

Just bought a 5700 for a friends pc upgrade ..i did tell him to wait until the customs are out ..but he wants now so he gets .... but i am sure with little bit of loving care undervolting and fan curve it can be tamed a little.. From what i read with one review the fan curve is a little borked and there using the same curve for the 5700 as the XT which does not make any sense if its lower volts frequency... i will know more for sure tomorrow
 
Soldato
Joined
18 Oct 2012
Posts
4,146
Location
Oxfordshire
Can you find a review praising the blower?

It's not about praising it. It's about showing the performance relative to the 2070S. That is AMD direct to NVidia not AIB on either.

Going back to AIB to AIB could also be done but currently both are just stock options so performance as is and what has been shown is relevant.

People here are suggesting they just don't work which isn't true cause even with what is thought to be a poor blower it is trading blows with the 2070s.

Surely that is what matters at moment. With that though, I'd you read the reviews I've already linked although the blower are not of course excellent or anything you are still able to get a reasonable overclock if you look at how to achieve best performance, just like the 56/64/VII previous cards.

A little fan management, a little power change, a little undervolting etc can go a long way for some good gains. What the replies were too though is people suggesting you need an AIB XT to really compete to the stock 2070S which isn't true as shown in the results posted.

Just means if you look at AIB for one you are likely doing same for other. Semi moot as in 4 weeks there will be AIB for the 5700 & 5700XT.
 
Soldato
Joined
22 Apr 2016
Posts
3,427
It's not about praising it. It's about showing the performance relative to the 2070S. That is AMD direct to NVidia not AIB on either.

Going back to AIB to AIB could also be done but currently both are just stock options so performance as is and what has been shown is relevant.

People here are suggesting they just don't work which isn't true cause even with what is thought to be a poor blower it is trading blows with the 2070s.

Surely that is what matters at moment. With that though, I'd you read the reviews I've already linked although the blower are not of course excellent or anything you are still able to get a reasonable overclock if you look at how to achieve best performance, just like the 56/64/VII previous cards.

A little fan management, a little power change, a little undervolting etc can go a long way for some good gains. What the replies were too though is people suggesting you need an AIB XT to really compete to the stock 2070S which isn't true as shown in the results posted.

Just means if you look at AIB for one you are likely doing same for other. Semi moot as in 4 weeks there will be AIB for the 5700 & 5700XT.
https://www.techpowerup.com/review/amd-radeon-rx-5700-xt/28.html
It’s 10% behind a 2070s it’s on par with a stock 2070. TBH I wouldn’t fancy it’s chances against a well overclocked non super 2070, the ones that can beat an R7 in timespy!

But I would be very interested to see an AIB board claw back that deficit and of course it will happen, a properly cooled XT may be worth waiting for a blower version is simply a bad choice, period.
 
Soldato
Joined
7 Feb 2015
Posts
2,864
Location
South West
Going back to AIB to AIB could also be done but currently both are just stock options so performance as is and what has been shown is relevant.

People here are suggesting they just don't work which isn't true cause even with what is thought to be a poor blower it is trading blows with the 2070s.

Pretty much this, you have *Blower Noises" competing with a card sporting what people consider a "Superior" twin axial fan cooling solution.
 
Soldato
Joined
18 Oct 2012
Posts
4,146
Location
Oxfordshire
https://www.techpowerup.com/review/amd-radeon-rx-5700-xt/28.html
It’s 10% behind a 2070s it’s on par with a stock 2070. TBH I wouldn’t fancy it’s chances against a well overclocked non super 2070, the ones that can beat an R7 in timespy!

But I would be very interested to see an AIB board claw back that deficit and of course it will happen, a properly cooled XT may be worth waiting for a blower version is simply a bad choice, period.

As above its average 4% overall and actually ahead of a 2070S in some titles so not as clear cut as saying it's just 10% slower.

Both can be clocked well even the blower tbh to get close if not a little over 2000mhz with some tweaking. That seems pretty decent. Meaning an AIB should be pushing 2100/2200mhz depending on cooling.

Then add in driver updates to mature over next few months (which super already has got since same drives as their non supers) and you could see the 5700XT clawering a little more back and so at least sorting out the outliers in current tests.
 
Associate
Joined
31 Jan 2012
Posts
1,976
Location
Droitwich, UK
I think this has gone off topic somewhat. Thanks for the feedback so far. The general feedback seems to be to return the VII and get either of those two cards.

Absolutely. And if you can wait see what 3rd party cards bring to the the 5700 XT. Being a new architecture I expect it to match/beat the more expensive 2070S in future but it depends how long you'll be keeping the next purchase for.
 
Soldato
Joined
22 Apr 2016
Posts
3,427
As above its average 4% overall and actually ahead of a 2070S in some titles so not as clear cut as saying it's just 10% slower.

Both can be clocked well even the blower tbh to get close if not a little over 2000mhz with some tweaking. That seems pretty decent. Meaning an AIB should be pushing 2100/2200mhz depending on cooling.

Then add in driver updates to mature over next few months (which super already has got since same drives as their non supers) and you could see the 5700XT clawering a little more back and so at least sorting out the outliers in current tests.
Comparing to the 2070s doesn’t change the fact it’s a useless cooler.

Compare the 5700xt to a non super 2070 and they are identical at 4K and within tiny margins of error at other resolutions.

My point is the same as many in this thread which is to ignore this ‘offering’ and wait for the AIB models which will have less compromises.
 
Soldato
Joined
18 Oct 2012
Posts
4,146
Location
Oxfordshire
Comparing to the 2070s doesn’t change the fact it’s a useless cooler.

Compare the 5700xt to a non super 2070 and they are identical at 4K and within tiny margins of error at other resolutions.

My point is the same as many in this thread which is to ignore this ‘offering’ and wait for the AIB models which will have less compromises.

Okay but this all started as performance comparison to the 2070S which with the poor blower of the 5700XT is still only 4% averaged off and tbh depending on game beats out the 2070s which almost everyone ignored and claimed the XT was getting destroyed which isn't factually true at all.

Edit so discussion went to is that extra 4% average performance worth £100 more. I'd say not likely so be it ref or AIB later that to me would be the better bang for buck card
 
Soldato
Joined
4 Feb 2006
Posts
3,202
Okay but this all started as performance comparison to the 2070S which with the poor blower of the 5700XT is still only 4% averaged off and tbh depending on game beats out the 2070s which almost everyone ignored and claimed the XT was getting destroyed which isn't factually true at all.

Edit so discussion went to is that extra 4% average performance worth £100 more. I'd say not likely so be it ref or AIB later that to me would be the better bang for buck card

Only the resident Nvidia shills are claiming that the 2070S destroys the 5700XT. The reality is that the 5700XT blower version is trading blows with the 2070S which has a better cooler so boosts higher. Once AIB cards are out then the performance will be closer but I expect pricing to move up a bit for both. The 5700XT is new architecture so drivers will mature over the next few months and the difference will become negligible or it may even take the lead.

Also the new consoles will have AMD NAvi architecture so devs will be optimizing for the 5700XT architecture.
 
Soldato
Joined
22 Apr 2016
Posts
3,427
Okay but this all started as performance comparison to the 2070S which with the poor blower of the 5700XT is still only 4% averaged off and tbh depending on game beats out the 2070s which almost everyone ignored and claimed the XT was getting destroyed which isn't factually true at all.

Edit so discussion went to is that extra 4% average performance worth £100 more. I'd say not likely so be it ref or AIB later that to me would be the better bang for buck card
I don’t think we are looking at the same chart mate as the tech power up average summary shows an XT at 100% and a 2070S at 114% for 4K?
 
Soldato
Joined
18 Oct 2012
Posts
4,146
Location
Oxfordshire
I don’t think we are looking at the same chart mate as the tech power up average summary shows an XT at 100% and a 2070S at 114% for 4K?

Nope because I was using the tech spot I linked above. It comes down to what games you sample though. The point is you could do it all one way or other.

If you calculate the totals thus far from the 3 I linked for 4k the average actually works out around 8% down on the 2070S which I am still going to say makes it much better bang for buck.

And we can still likely see the gap close when the outliers are sorted as there appears to be a few that are just not working well again with windows scheduler.

Fortunately I don't play those games apart from R6S which seems to be one those outliers at moment.
 
Soldato
Joined
4 Feb 2006
Posts
3,202
I don’t think we are looking at the same chart mate as the tech power up average summary shows an XT at 100% and a 2070S at 114% for 4K?

This video comparison shows that there is negligible difference between the two cards and the 5700XT actually is the best bang for buck. It's running at 200MHz slower than the 2070S but keeping up and sometimes beating it.

 
Soldato
Joined
22 Nov 2009
Posts
13,252
Location
Under the hot sun.
This video comparison shows that there is negligible difference between the two cards and the 5700XT actually is the best bang for buck. It's running at 200MHz slower than the 2070S but keeping up and sometimes beating it.


Interesting that the 5700XT is stuck at 1700 -_-
And given the prices, someone might wonder how can justify 25% higher price of the 2070S or the 60% higher price of the 2080!!!!!! (that applies for RVII also which is terrible value).
It gets beaten in some games, by a AMD blower card!!!!!!!! And several are Nvidia titles and flagships for RT!!!!

For p sake it costs <10p to put 4 (0.75-1mm) plastic washers on the retention plate screws to improve cooling. Doesn't even need to undervolt.
 
Last edited:
Soldato
Joined
22 Apr 2016
Posts
3,427
This video comparison shows that there is negligible difference between the two cards and the 5700XT actually is the best bang for buck. It's running at 200MHz slower than the 2070S but keeping up and sometimes beating it.

I can’t take these videos seriously. They only show a handful of games, often with dodgy methodology to invariably make a vendor biased point. The professional reviews always use a much wider variety of games and show different results for a reason. (Credibility).
 
Soldato
Joined
18 Oct 2012
Posts
4,146
Location
Oxfordshire
I can’t take these videos seriously. They only show a handful of games, often with dodgy methodology to invariably make a vendor biased point. The professional reviews always use a much wider variety of games and show different results for a reason. (Credibility).

Whilst agree more is better, some of these videos cover stuff more mainstream don't.

Not sure what was particularly wrong with this though. There are 9 games in that video which is about the same as most reviews seem to cover in the written reviews whilst showing all the correct graphics and videos playing behind to show what we would expect as a gamer which is much more than any written review.

The video seems pretty well put together with the graphs easy to read and plenty of data points to select.
 
Associate
Joined
29 Aug 2013
Posts
1,176
I can’t take these videos seriously. They only show a handful of games, often with dodgy methodology to invariably make a vendor biased point. The professional reviews always use a much wider variety of games and show different results for a reason. (Credibility).

I trust them more than Techpowerup. They always have the Nvidia cards ahead of where they actually are compared to every other review, shame so many quote their relative performance charts.

Hardware unboxed has the 5700XT 2% slower than the 2070S (removing Forza is pointless imo, might aswell remove Nvidia favored Fortnite too then) and 9% slower than the 2080, big difference compared to Techpowerup. Guru3d also has them neck and neck.
 
Associate
Joined
18 Aug 2014
Posts
366
I don’t think we are looking at the same chart mate as the tech power up average summary shows an XT at 100% and a 2070S at 114% for 4K?
Those relative performance graphs are always a load of bull anyway. It has the Vega 64 at 84% and the RTX 2060 at 85% when in reality the Vega 64 is quicker, even in games favored by Nvidia
 
Status
Not open for further replies.
Back
Top Bottom