Radeon RX Vega 64 vs. GeForce RTX 2080, Adrenalin 2019 Edition Driver Update Benchmark Test

[Poneros]

I think it’s also worth bearing in mind that vega at 100% fan speed is not indicative of true performance as it’s simply unplayable so I will take ‘suicide run’ performance figures with a pinch of salt too.

Who said anything about 100% fan speed? If you look at the comparable Vega 64 in UK it's the £420 Nitro, which most assuredly doesn't need anywhere near 100% fan speed for even high OCs. Here: https://www.tomshardware.com/reviews/sapphire-rx-vega-64-nitro,5388-8.html

If anything, AMD's cards always perform better when adjusted for temperature because they don't throttle as soon as Nvidia's. Even more so in Polaris' case.

While that's great for the Vega 64, this doesn't look great for the VII to be honest. It's supposedly ~30% better than the Vega 64. 83*1.3=107.9, that puts it level with the 1080 Ti and 10% behind the 2080 that it's supposed to compete with.

Like I said before, those aren't indicative numbers, as TPU is a shoddy site, and more importantly the numbers given by AMD for V64 are from the LC edition, which is above any 2070 anyway. So rest assured, Radeon 7 will at best be no worse than a 2080, though likely better especially if we look at the tweaks still possible for AMD cards which for Nvidia are locked away.

 

[Smrtka]

Do you think the vega 64 is better then 2070 for gaming ? I am considering an upgrade from 1070 to potentially drive 1440p UW screen soon, and this seems like vega might be better deal

 

[Brazo]

Do you think the vega 64 is better then 2070 for gaming ? I am considering an upgrade from 1070 to potentially drive 1440p UW screen soon, and this seems like vega might be better deal

Nope.

This should tell you all you need to know.
https://www.techpowerup.com/reviews/Palit/GeForce_RTX_2070_GameRock_Premium/27.html

Vega 64 spars with the 2060.

 

[Poneros]

Do you think the vega 64 is better then 2070 for gaming ? I am considering an upgrade from 1070 to potentially drive 1440p UW screen soon, and this seems like vega might be better deal

They trade blows depending on the game, and which 2070 we're talking about. From a price point of view, it's not a contest in the slightest - the Nitro Vega 64 wipes the floor with the 2070s.

Example: £420 Nitro vs £650 2070 Aorus.

You see, I explained to Brazo TPU is a shoddy site but he wants to remain ignorant and shill Nvidia instead. What you don't see in all these reviews sites is how misleading the comparisons are & how poor their methodology is (for one thing you can't even tell what configuration they set it at). When you compare real cards at real prices, it's obvious how much better the Vega 64 is atm. And you can tell because they have a very neat slight of hand trick: they use premium model reviews for performance and then they use the price of the bottom tier cards and pretend like they're all "2070s" even though there's very real performance deltas between them, starting at 10% for A vs non-A chips, to say nothing of all the other differences like cooler which is more important for Nvidia because they throttle earlier. If we look instead at the cheapest comparable card (on OCUK - no competitor talk) to the Nitro, it's the £500 Gaming Pro 2070 which isn't even in stock (the cheapest in-stock with an A chip & decent cooler is Zotac's AMP at £575...). To say nothing of the better value of the game bundle from AMD (because it's sellable and has 3 games not 2). And we haven't even touched OC'ing!

So how could anyone say with a straight face that 2070 is better than V64 when also considering price? The answer is clear, only shills and fanboys could.

 

[henson0115]

They trade blows depending on the game, and which 2070 we're talking about. From a price point of view, it's not a contest in the slightest - the Nitro Vega 64 wipes the floor with the 2070s.

Example: £420 Nitro vs £650 2070 Aorus.

You see, I explained to Brazo TPU is a shoddy site but he wants to remain ignorant and shill Nvidia instead. What you don't see in all these reviews sites is how misleading the comparisons are & how poor their methodology is (for one thing you can't even tell what configuration they set it at). When you compare real cards at real prices, it's obvious how much better the Vega 64 is atm. And you can tell because they have a very neat slight of hand trick: they use premium model reviews for performance and then they use the price of the bottom tier cards and pretend like they're all "2070s" even though there's very real performance deltas between them, starting at 10% for A vs non-A chips, to say nothing of all the other differences like cooler which is more important for Nvidia because they throttle earlier. If we look instead at the cheapest comparable card (on OCUK - no competitor talk) to the Nitro, it's the £500 Gaming Pro 2070 which isn't even in stock (the cheapest in-stock with an A chip & decent cooler is Zotac's AMP at £575...). To say nothing of the better value of the game bundle from AMD (because it's sellable and has 3 games not 2). And we haven't even touched OC'ing!

So how could anyone say with a straight face that 2070 is better than V64 when also considering price? The answer is clear, only shills and fanboys could.

so much this. i always lol when someone posts a tpu link. there methodology is shady at best.

 

[Poneros]

so much this. i always lol when someone posts a tpu link. there methodology is shady at best.

Exactly. And here's the sad part: In my own neck of the woods the situation is the exact opposite, and I have no issue saying it, where the 2070 & 2060 have way better prices & selection than the Vegas. I don't need to support either corporation, the individual should look out for themselves not these companies. No one's giving out free ****.

 

[Raz]

Maybe people feel as if they need to justify to themselves/others spending £500+ on a single component, and naturally fight for whatever they bought.

As a consumer it should be about value for money, paying the price for what you want in running, and playing games.

Like the look of a Vega 64? Check out reviews, feedback on the forums. Same for Nvidia.

 

[Brazo]

Exactly. And here's the sad part: In my own neck of the woods the situation is the exact opposite, and I have no issue saying it, where the 2070 & 2060 have way better prices & selection than the Vegas. I don't need to support either corporation, the individual should look out for themselves not these companies. No one's giving out free ****.

As a vega owner and soon to be Ryzen I struggle to see how you can call me a fanboy! I am however a realist. The 2070 can be bought for as little as £450 and I have posted up a link for you before that shows the performance of the cheaper cards being no different (overclocking aside) Whilst we can’t name competitors, OCUK isn’t competitive when it comes to Turing prices the cards can be had for a lot cheaper in other sites.

This ‘A’ chip stuff is just guff.

 

[Brazo]

Here is a £459 2070 I think an extra £39 is worth 15-20 FPS at 1440p?

Here is a different review site if you don’t like tpu. Their 12 game average again Tells the same story. Just say if you don’t like this one and I can find another with the same results.

https://www.pcgamer.com/uk/evga-geforce-rtx-2070-black-review/

 

[Poneros]

As a vega owner and soon to be Ryzen I struggle to see how you can call me a fanboy! I am however a realist. The 2070 can be bought for as little as £450 and I have posted up a link for you before that shows the performance of the cheaper cards being no different (overclocking aside) Whilst we can’t name competitors, OCUK isn’t competitive when it comes to Turing prices the cards can be had for a lot cheaper in other sites.

This ‘A’ chip stuff is just guff.

Which makes your comments even more bizzare.

Guff? Here's the reality:

https://www.gamersnexus.net/hwreviews/3386-nvidias-secret-gpu-tu106-400-vs-400a-2070-xc-ultra-review

Even with the golden sample of the black when OCing it's still at least 5-7% worse, to say nothing of the real scenarios of 4K where it's clearly 10%, esp. DX12 where the card actually gets utilized.

Here is a £459 2070 I think an extra £39 is worth 15-20 FPS at 1440p?

Here is a different review site if you don’t like tpu. Their 12 game average again Tells the same story. Just say if you don’t like this one and I can find another with the same results.

https://www.pcgamer.com/uk/evga-geforce-rtx-2070-black-review/

Show me a review where you have the methodology & configurations clearly shown, not these garbage sites. Like I did earlier, you can see the clocks in real time & know what's actually being benchmarked. Further more, it should be equivalent models, not reference Vegas that haven't been sold since Bitcoin was $20k and which have no additional OC/UC configuration. Stock clocks vs stock clocks, fine, but must also show OC results.

If the top end Vega 64 model is £420, then show me the 2070 that's near £420 and how it compares, or baring that go balls out with the premium 2070 models, but let's not ignore the >£100 difference. Btw, that EVGA isn't even £450, it's £520 in the UK at the cheapest, or 520 euros @ EVGA's own site. Not in stock anyway though.

 

[henson0115]

As a vega owner and soon to be Ryzen I struggle to see how you can call me a fanboy! I am however a realist. The 2070 can be bought for as little as £450 and I have posted up a link for you before that shows the performance of the cheaper cards being no different (overclocking aside) Whilst we can’t name competitors, OCUK isn’t competitive when it comes to Turing prices the cards can be had for a lot cheaper in other sites.

This ‘A’ chip stuff is just guff.

Whilst I don't care for the v64 Vs 2070 debate, the tpu benches are crud, look at the naming of the cards. You telling me that all of those cards tested are all reference? On a like for like driver, now go look at enthusiast benches you know the people who do it because they love it, not because they get paid to do it, you will notice the difference such an explanation of setup, methodology, Drivers, specific model numbers, including things like frame time, resolution, CPU utilisation, deltas.

 

[Smrtka]

Which makes your comments even more bizzare.

Guff? Here's the reality:

https://www.gamersnexus.net/hwreviews/3386-nvidias-secret-gpu-tu106-400-vs-400a-2070-xc-ultra-review

Even with the golden sample of the black when OCing it's still at least 5-7% worse, to say nothing of the real scenarios of 4K where it's clearly 10%, esp. DX12 where the card actually gets utilized.



Show me a review where you have the methodology & configurations clearly shown, not these garbage sites. Like I did earlier, you can see the clocks in real time & know what's actually being benchmarked. Further more, it should be equivalent models, not reference Vegas that haven't been sold since Bitcoin was $20k and which have no additional OC/UC configuration. Stock clocks vs stock clocks, fine, but must also show OC results.

If the top end Vega 64 model is £420, then show me the 2070 that's near £420 and how it compares, or baring that go balls out with the premium 2070 models, but let's not ignore the >£100 difference. Btw, that EVGA isn't even £450, it's £520 in the UK at the cheapest, or 520 euros @ EVGA's own site. Not in stock anyway though.

From GN it looks like the vega 64 is bit less powerful, but if its £100 less that seams like a better deal. I wonder if the OC vega 64 can match oc 2070 as stock to stock looks like 5% difference.

 

[Poneros]

From GN it looks like the vega 64 is bit less powerful, but if its £100 less that seams like a better deal. I wonder if the OC vega 64 can match oc 2070 as stock to stock looks like 5% difference.

Ofc it can, AMD allows for great OCing. GN's Vega 64 coverage is scattered and poor unfortunately so can't go by it. For one thing we don't even know if they modded it to avoid the throttling characteristic to these Asus garbage cards due to their shorter thermal pads, and for another they don't have any OC results for V64 nor do they test it with all their suite - which is just bizzare. Unfortunately the only decent reviews I have for them are older. Eg:

The reviews I trust the most are from overclockers club but unfortunately they didn't test V64 & they don't seem to update much so I use them to decide on GPU at launch. They do have this interesting piece on V64's efficiency though: https://www.overclockersclub.com/reviews/vega_efficiency/

 

[Brazo]

Whilst I don't care for the v64 Vs 2070 debate, the tpu benches are crud, look at the naming of the cards. You telling me that all of those cards tested are all reference? On a like for like driver, now go look at enthusiast benches you know the people who do it because they love it, not because they get paid to do it, you will notice the difference such an explanation of setup, methodology, Drivers, specific model numbers, including things like frame time, resolution, CPU utilisation, deltas.

There is no 2070/64 debate, Perhaps a 2060/64 debate. All professional benchmarks on the web are very clear in their methodology. Yes some are better than others but I’ll take them any day of the week over a youtube suicide run vid.

 

[Poneros]

There is no 2070/64 debate, Perhaps a 2060/64 debate. All professional benchmarks on the web are very clear in their methodology. Yes some are better than others but I’ll take them any day of the week over a youtube suicide run vid.

Why keep spreading fud? At best the 2070 is 10% better than a V64 overall - STOCK (with a throttling Asus). OC for OC they trade blows. I'm gonna call you out on your nonsense every time.

 

[henson0115]

There is no 2070/64 debate, Perhaps a 2060/64 debate. All professional benchmarks on the web are very clear in their methodology. Yes some are better than others but I’ll take them any day of the week over a youtube suicide run vid.

really i see a whole bunch of missing information from the supposedly professional tpu benchmarks, the most important part of course being the actual card names of the cards used, without this we have no idea of which card they are actually referring to. there can be massive variation from a single card made by two separate vendors or even the same vendor. i/e the gamesnexus benchmark posted by peneros shows the evga 2070 xc ultra beating the stock 2070 by some margin, until TPU fixes there shonky testing many people will ignore there results regardless of what you say is "professional" or "not".
p.s professional does not mean perfect, even unbiased or in anyway better than a non proffesional review, it simply means paid for by someone i/e an employee, doesn't exempt them from making mistakes etc - which is why people will always ask for methodology, as this can prove the same has been done across all cards etc.
by your logic we should trust facebook did not share our data as they are professionals, so we should just trust them and we all know how that turned out.

 

[tommybhoy]

'A 64 is faster than a 2070' because it runs well on Frostbite which was developed for GCN-my sides are killing me!

Lets get a comparison up of Fortnite!

 

[bru]

Why keep spreading fud? At best the 2070 is 10% better than a V64 overall - STOCK (with a throttling Asus). OC for OC they trade blows. I'm gonna call you out on your nonsense every time.

You seem to come across very biased, anyone who disagrees is a shill and websites who show NVIDIA doing well are using cruddy testing methodology.

May be this is doing you a disservice, but that is the way it comes across.

On the 2070 Vs Vega 64 debate, yes the 2070 is faster, but it is dearer. So it is upto the individual to decider, depending on the prices they can find them for.

 

[Poneros]

You seem to come across very biased, anyone who disagrees is a shill and websites who show NVIDIA doing well are using cruddy testing methodology.

May be this is doing you a disservice, but that is the way it comes across.

On the 2070 Vs Vega 64 debate, yes the 2070 is faster, but it is dearer. So it is upto the individual to decider, depending on the prices they can find them for.

Fair.

 

[Brazo]

Why keep spreading fud? At best the 2070 is 10% better than a V64 overall - STOCK (with a throttling Asus). OC for OC they trade blows. I'm gonna call you out on your nonsense every time.

I’m sorry sunshine but your posts come across very immature. You cherry picked one game (BF V) whereas I have shown benchmarks across whole suites of games.

You highlight the most expensive 2070 when one literally £200 cheaper performs very similar and still better than the vega 64.

I can read benchmarks the same as you can, if you choose not believe them I can’t help you but it’s discourteous to accuse someone of spreading fud when they are simply linking to professional reviews just because they are an inconvenient truth for you.

Edit. I see in your post above you have at least now acknowledged your own bias.