• 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.

VEGA 64 or RTX 2070

Soldato
Joined
26 May 2014
Posts
2,952
I own the same monitor. Tried the lesser Vega 56 nitro. Sold it after a month. To Tolivery good fps, the card after many, many adjustments, was consuming between 200/215W and boiling the whole case. Went for the RTX2080. An overkill to the Vega 56, I know, but manage to get the same performance, capping the fps to the same level as the Vega was delivering, and as the GPU wasn't being pushed at all, was running really cool, and consuming between 80/100W. Doesn't seems a lot, price wise, as I don't play for long hours, not a deal breaker, but the heating and consequently the noise of the fans at higher speed, yes, it's day and night.
The RTX should be fully compatible with your monitor, the fact that is a fast card, will make use of the monitor's resources, and for VR, should be a great card.
One friend bought the 2070 and is really happy with it. At least one hour a day trying to convincing me how great it is.
Buying a card that's twice the price (and then some) just so that you can not use it is a bananas strategy. You must have some pretty terrible case airflow too, as I've had many 250W+ cards, used them to their full potential and never had them "boiling the whole case". Even when paired with extremely thirsty and hot CPUs like an overclocked 3970X. Not really a problem if you actually set your fans up properly. You certainly don't need to buy a ridiculous overkill card and then use it to only 25% of its potential to have relative silence. I just sold a Palit GTX 1080 Jetstream that was completely inaudible even under full load for hours at a time, and I have my case a couple of feet from my ears. The fans barely spun on the thing thanks to the massive cooler. I have no doubt that you can buy RTX 2060s and 2070s that are similarly functionally silent and would be far more suitable. Of course, you suggest that you have money to burn and don't care, which is your perogative I suppose. You'd think you could sort yourself a few NF-A14s out with that in mind though.
 
Associate
Joined
5 Mar 2017
Posts
2,248
Location
Cambridge
Buying a card that's twice the price (and then some) just so that you can not use it is a bananas strategy. You must have some pretty terrible case airflow too, as I've had many 250W+ cards, used them to their full potential and never had them "boiling the whole case". Even when paired with extremely thirsty and hot CPUs like an overclocked 3970X. Not really a problem if you actually set your fans up properly. You certainly don't need to buy a ridiculous overkill card and then use it to only 25% of its potential to have relative silence. I just sold a Palit GTX 1080 Jetstream that was completely inaudible even under full load for hours at a time, and I have my case a couple of feet from my ears. The fans barely spun on the thing thanks to the massive cooler. I have no doubt that you can buy RTX 2060s and 2070s that are similarly functionally silent and would be far more suitable. Of course, you suggest that you have money to burn and don't care, which is your perogative I suppose. You'd think you could sort yourself a few NF-A14s out with that in mind though.
When running RE 2 remake to achieve 75, or even the 144 supported by the monitor, didn't put stress on the GPU.
In games like RE 2 or DMC5, I can't honestly tell the difference if running over 100fps. Or even 75. But to do that, the 2080 does without any trouble, even saving energy and pulling less heat. The Vega 56 had to be full-blown to delivery such performance, some times dropping, and at 200-215w, the heat generated is a lot.
I'm using a small case, true, the Phanteks P400S, but considering that the PSU is fully isolated from the system, I'm using a 12cm back, supplied with the case, 2 X 14cm exhausting at the top and 3 X 14 cm pulling at the front. The CPU using the Dark Rock Pro 3.
The main problem with the Vega 56 Nitro was the cooling, I assume, as the hot air was being blown straight on top of the M.2 and the side window, not outside the case, as any other GPU I ever owned.
When you touch the glass window, it was uncomfortable, unable to hold more than a few seconds.
Not the case to burn money. Got my HD7970 on release and just changed last January.
I think that buying a more expensive GPU, and using it for few years is better than changing GPU every year or year and a half.
In other scenarios, I'm running everything I'm playing at 2k, max, no issues, and the GPU should keep like that for few years, just the case to go easy on a few settings eventually or in the near future.
Just as comparison, same small case, same setup, running benchmarks, which certainly will push any system available: the Asus HD7970DCU was acceptable warm, just noticeable fan noise, but not too much. The TheM.2 temperature, around 40C. Sapphire Vega 56 Nitro, hot and the fan noise slightly more than the HD7970. The temperature of the M.2, was nearly 75C. I would guess the side windows slightly cooler than the M.2. the XFX2080, slightly cooler than the HD7970 and slightly quieter. M.2 temperature about 35C.
Considering the performance and the power drain, and especially the TDP, yes, the Vega 56 is poor. Specially those non-blower models where unless water cooled the heat will stay inside the case, and unless you build a case of fans, will always be a nuisance.
 
Associate
Joined
5 Mar 2017
Posts
2,248
Location
Cambridge
If the case is hot to touch then it's not dispersing the heat properly. Glass probably isn't the best material to use.
The main issue is probably the design of the card, not blowing the hot air outside, as one would expect but to the sides of the card. That happening, considering that it's a very hot card, there isn't much to be done, as the M.2 next to it, and the side panel will get it's 70, 80C air and will be a while until the cold air being pulled in and the hot air being pushed out, creating a tunnel would be able to cool the case, but will never be too efficient, as none of the hot air from the GPU is being blown out by the card's cooling system.
The glass panel isn't ideal, true, but even without the window, the motherboard will still receive the blast of hot air, like a hairdryer pointed to it. The fans will do what they're designed to, but won't be able to cool the whole case to the point to undo the effects of such heat being generated and blown against other components.
My recommendation against the Vega is because, despite the many adjustments, it still draining a lot of power and dissipating too much heat. That for the Vega 56. The Vega 64 is said to be worst. Then as the OP needs a GPU, has a monitor that can work flawless with any of the mentioned GPUs, my recommendation would be for the cooler, more power efficient card.
 
Soldato
Joined
22 Nov 2006
Posts
23,361
No cards blow the air outside, except for reference ones with blowers on them. As we all know they aren't as good.

You use case fans to keep air flowing through the whole case.
 
Associate
Joined
5 Mar 2017
Posts
2,248
Location
Cambridge
No cards blow the air outside, except for reference ones with blowers on them. As we all know they aren't as good.

You use case fans to keep air flowing through the whole case.
My previous Asus hd7970DCII did, and almost no hot air was left inside. You could feel the top of the card warmer, and the area around it slightly warmer, but as efficient as possible blowing the hot air out. My actual 2080 does too, but a small quantity still inside the case. The top of the card doesn't feel even as warmer as the hd7970, but the area around it feels warmer despite the overall smaller heat dissipation.
The issue is that the hot air being blown straight on the M.2 area doesn't allow for any fan setup to cool it off quickly enough.
Tried to relocate the card to the second slot but there's a drop in performance, as the second slot runs at 4x.
 
Soldato
Joined
6 Feb 2010
Posts
14,594
My previous Asus hd7970DCII did, and almost no hot air was left inside. You could feel the top of the card warmer, and the area around it slightly warmer, but as efficient as possible blowing the hot air out. My actual 2080 does too, but a small quantity still inside the case. The top of the card doesn't feel even as warmer as the hd7970, but the area around it feels warmer despite the overall smaller heat dissipation.
The issue is that the hot air being blown straight on the M.2 area doesn't allow for any fan setup to cool it off quickly enough.
Tried to relocate the card to the second slot but there's a drop in performance, as the second slot runs at 4x.
Cards with 2-2.5 slot custom cooler with 2-3 fans on top of the headsink that doesn't have a blower cooler like the HIS IceQ generally wouldn't push much heat out the back, as it is simply not designed to do that.

If there's large amount of heat coming out the back of the graphic card despite being a non-blower cooler, then it just mean the case have very poor airflow as there's so much heat built-up inside of the case, to the point there enough of them to get out of the back of the graphic card. If a case had good airflow- with decent amount of cool air intake and exhaust fan(s) that properly remove the heat from the case there should be minimal amount of heat coming out from the rear of the graphic card.

The scenario you describe with the Asus 7970 DCII is an exception to the rule, as it has an very unique designed cooler with it being a 3 slots card which the shroud actually cover up the side of the card (rather than with the heatsink openly exposed like most other custom cooler cards), which is why it managed to direct the heat from the card to the rear vent out the back to an extent.
 
Associate
Joined
5 Mar 2017
Posts
2,248
Location
Cambridge
That's correct. Even my actual 2080 doesn't have the whole heatsink "wrapped" as the HD7970, reason why some hot air scapes from its sides.
The vantage is that it runs much cooler than the Vega 56, and particularly I didn't like the design of the Sapphire Pulse. Maybe would fare better at my previous case, the Fractal Design Define XL, but after many years and between upgrades, few panels started vibrating, reason why I decided for a smaller case, to avoid being tempted to build water cooled systems.
 
Soldato
Joined
10 Oct 2012
Posts
4,421
Location
Denmark
Okay i'm having some issues with those benchmarks shown in the beginning of this entire thread. We have seen so many where the rtx 2070 is being compared to the gtx 1080 and they are pretty neck and neck with perhaps 1-3% difference going to the rtx 2070. Vega 64 in recent titles just about edges out the gtx 1080 as well so in theory should be pretty neck and neck with the rtx 2070 and now according to the videos in the beginning it cant even match a rtx 2060? like wtf? did i miss something?
 
Soldato
Joined
18 Feb 2015
Posts
6,484
Okay i'm having some issues with those benchmarks shown in the beginning of this entire thread. We have seen so many where the rtx 2070 is being compared to the gtx 1080 and they are pretty neck and neck with perhaps 1-3% difference going to the rtx 2070. Vega 64 in recent titles just about edges out the gtx 1080 as well so in theory should be pretty neck and neck with the rtx 2070 and now according to the videos in the beginning it cant even match a rtx 2060? like wtf? did i miss something?

It comes down to this: There's clock variation, there's peculiar situations with OC due to the way these boost algorithms work, there's game variation, there's drivers variations, there's different platforms (CPU etc), there's different settings used, there's different windows/bios settings which affect FPS esp for Vega (fast boot et al) etc.

In effect all these factors end up obscuring the picture of what actual performance will look like. This why I say always avoid any Techpowerup benchmarks (as well as some other publications) because they're always wrong.

This is also why I added the caveat in the first post.

Let me be very clear: The real performance is like this 2070 > V64 > 2060. The margin of difference will vary depending on resolution (e.g. 4K where V64 = 2070) & game (e.g. Unreal games = +10% performance boost for Nvidia almost always), as well as settings (e.g. MSAA hits AMD cards harder than Nvidia; 4K HDR used to MURDER Pascal GPUs due to puny bandwidth).

Remember when 1080 was being praised to high heavens & everyone thought it bested the Vega 64 because some ****** reviewers (read: 99.9% of all of them) didn't know how to test nor did they bother to experiment? How many of them complained about how poor Pascal did with HDR when that's something so basic & essential to modern gaming? Unfortunately, there's no one publication with a stellar record you can rely on. There's no easy way except a lot of research (and ideally personally done as well).

Some people did do the work, but they didn't get advertisers so now they're gone, just the way of the world:

 
Soldato
Joined
10 Oct 2012
Posts
4,421
Location
Denmark
It comes down to this: There's clock variation, there's peculiar situations with OC due to the way these boost algorithms work, there's game variation, there's drivers variations, there's different platforms (CPU etc), there's different settings used, there's different windows/bios settings which affect FPS esp for Vega (fast boot et al) etc.

In effect all these factors end up obscuring the picture of what actual performance will look like. This why I say always avoid any Techpowerup benchmarks (as well as some other publications) because they're always wrong.

This is also why I added the caveat in the first post.

Let me be very clear: The real performance is like this 2070 > V64 > 2060. The margin of difference will vary depending on resolution (e.g. 4K where V64 = 2070) & game (e.g. Unreal games = +10% performance boost for Nvidia almost always), as well as settings (e.g. MSAA hits AMD cards harder than Nvidia; 4K HDR used to MURDER Pascal GPUs due to puny bandwidth).

Remember when 1080 was being praised to high heavens & everyone thought it bested the Vega 64 because some ****** reviewers (read: 99.9% of all of them) didn't know how to test nor did they bother to experiment? How many of them complained about how poor Pascal did with HDR when that's something so basic & essential to modern gaming? Unfortunately, there's no one publication with a stellar record you can rely on. There's no easy way except a lot of research (and ideally personally done as well).

Some people did do the work, but they didn't get advertisers so now they're gone, just the way of the world:


Yeah i really miss Mindblank Tech, learned a lot about Ryzen tuning from him.
 
Associate
Joined
29 Aug 2013
Posts
1,176
Vega 56 as I dont think either are worth the price premium over it for such little performance gain espcially when the 56 has undergone the undervolt/oc treatment.
2070 at near £500 you might aswell just get a 2080 imo.
 
Soldato
Joined
6 Feb 2010
Posts
14,594
2070 at near £500 you might aswell just get a 2080 imo.
That's actually a good point.

I mean the 2080 is not at its initial crazy launch price of £800+ anymore with it now at just around £600, while the 2070 with the B-chip are still averaging at around £430, and A-chip 2070 is around £480-£500; though it is also true that the bargain basement 2080 are most likely with B-chip.

Nvidia has stop the production of the B-chip though, the question is how long become old stocks are gone from the channels and shops, for the A-chip to make it into even the bargain basement models cards as well.
 
Back
Top Bottom