NVIDIA GeForce RTX 2060 Founders Edition 6 GB

[bug]

Turing is nearly twice as efficient per watt. Vega 64 (4096 core, 10.2-12.7 Tflop) is beaten by RTX 2060 (1920 core, 5.2-6.5 Tflop). It should be obvious to anyone how inefficient GCN really is at this point. Even the advantage of 7nm will not make up for this.

Fwiw, Vega is still ahead in compute. But still, these aren't compute cards, at least not primarily. And, of course, even there Turing wins when it can flex its tensor muscles.

I know all of those things, my point is that even if AMD is competetive on 7nm against turing, they are not gonna win anything.
I'm pretty sure nvidia likes to move to 7nm as soon as possible, rumors say they will use Samsung's EUV solution which should reduce manufacturing costs.

AMD wouldn't need to win in one round. Matching Nvidia will allow them to reap some cash that can go towards a future better iteration of their architecture. At the same time it will prevent Nvidia from charging inflated prices that will go towards their war chest, to be used against AMD in the future.
Theory is easy.

 

[Ruru]

Awesome price/performance ratio. Damn!

 

[M2B]

Fwiw, Vega is still ahead in compute. But still, these aren't compute cards, at least not primarily. And, of course, even there Turing wins when it can flex its tensor muscles.

what do you mean by "compute"?
Gamers Nexus has a very good review related to professional workloads and even there nvidia is ahead most of the time:

https://www.gamersnexus.net/guides/...2018-for-adobe-premiere-autocad-vray-and-more

 

[GoldenX]

GCN is like the Core arch now, old and repetitive. The only thing saving it is the lack of good low and mid end offerings from Nvidia.
Remember people, when real prices launch, the price/performance charts is going to be different.

 

[Mistral]

So, in practice this is a 1070Ti @ $349?

 

[bug]

what do you mean by "compute"?
Gamers Nexus has a very good review related to professional workloads and even there nvidia is ahead most of the time:

https://www.gamersnexus.net/guides/...2018-for-adobe-premiere-autocad-vray-and-more

I meant, in compute Vega is still ahead of this 2060, that's all. See: https://www.anandtech.com/show/13762/nvidia-geforce-rtx-2060-founders-edition-6gb-review/13

So, in practice this is a 1070Ti @ $349?

A 1070Ti which can do DXR, DLSS and has a bunch of tensor cores to go with it. Yes.

 

[moproblems99]

AMD needs to come to the realization that they can't be a jack of all trades and master of none in the GPU space anymore. Leave GCN for compute and design a gaming arch from the ground up. If the product is as good or better than NV's and it is not bought, then return the mid range and compute segments and don't return to high end. Give consumers one last chance.

Edited for better words.

 

[bug]

AMD needs to come to the realization that they can't be a jack of all trades and master of none in the GPU space anymore. Leave GCN for compute and design a gaming arch from the ground up. If the product is as good or better than NV's and it is not bought, then return the mid range and compute segments and don't return to high end. Give consumers one last chance.

Edited for better words.

I think they know very well what they have to do. It's just that they didn't have the money to actually do it. Their graphics division held its own for a while and they used that time to revive their CPU business. Now that that's covered, they'll probably have resources to throw the way of their graphics division. I'm not sure what to make of the late high profile figures abandoning that boat though.

 

[Deleted member 50521]

Solid performing card. Definitely not for any demanding RTRT applications but good 1440p card nonetheless.


Navi better be good. xx60 matching previous gen flagship this is just sad for RTG. Not to mention the perf per watt.

 

[Vya Domus]

Give consumers one last chance.

They did, multiple times actually. It's what brought them to where they are today.

If there is one thing AMD can learn, and it's something that has nothing to do with technology, it's the fact that no matter how much you cater to the consumer in the most popular segments, someone will always steal the show with halo products ,sweet talk and unorthodox practices . Admittedly , they were hamstrung by many things in the past that were mostly out of their control but they did had their shots and they failed.

 

[Deleted member 50521]

They did, multiple times actually. It's what brought them to where they are today.

If there is one thing AMD can learn, and it's something that has nothing to do with technology, it's the fact that no matter how much you cater to the consumer in the most popular segments, someone will always steal the show with halo products ,sweet talk and unorthodox practices . Admittedly , they were hamstrung by many things in the past that were mostly out of their control but they did had their shots and they failed.

AMD had some questionable management at that time. The new team seems to be pretty on top, at least CPU wise.

I am no longer hoping too much for RTG to improve. They are really far behind. I have more faith in Raja’s effort at Intel

 

[bug]

Solid performing card. Definitely not for any demanding RTRT applications but good 1440p card nonetheless.

RTRT is more or less in demo mode for this generation (the fact that the very first title to support it actually managed to get passable frame rates at 1440p is rather surprising for me, past technologies haven't been so lucky). If I were a developer trying to get a hang of how RTRT works and what it can do, this is the card I'd buy.
I never spend more than $300 on a video card and I rarely update from one generation to the next. Yet I feel compelled to get one one of these. If only to spite those talking trash about Turing whenever they can,

 

[wolf]

This card is the pick of the litter of the 20 series so far! what a beast for the price! now even if AMD just cuts vega prices we might see a price/perf shift.

 

[B-Real]

Very nice, Vega 64 performance for $350, whilst being way more efficient to boot.

For a $100 jump in price compared to the GTX 1060, which i think will be more likely around $400-450 in real life circumstances as the 1070Ti is available around $420+ and the 2070 is too close to it (nearly halfway compared to 1060-1070). And still there are the AIB Vega 56 models cheaper for less than $350 with a $150 3 game bundle. A better choice compared to the RTX 2060.

So, in practice this is a 1070Ti @ $349?

I would be quite surprised if it would cost $350, as the 1070 Ti costs around $420. And there are 2 Gigabyte GTX 1070 that cost around $320. And a reference 1070 for $300.

 

[Vya Domus]

I have more faith in Raja’s effort at Intel

Intel is getting into the dedicated GPU game at the worst time possible, they have to fight a towering monopoly, they are at least 1-2 years away from delivering anything remotely competitive which also means they'll miss the initial 10/7nm boat.

Only place where they can make a dent is datacenter/AI which is where I suspect they'll focus their efforts on anyway. Best you'll see from Intel for the gaming crowd in the foreseeable future will be better iGPUs and that's about it. But we'll see.

 

[birdie]

A very nice review, @W1zzard, I especially liked this part in the conclusion:

"By all intents and purposes, the RTX 2060 belongs to a higher market segment than the GTX 1060, and this is reflected in the card's performance."

Yes, the RTX 2060 is expensive to be a truly midrange card but it's not in the same class/league as the GeForce 1060 was/is.

Still, I'm gonna skip this entire generation because it looks like a stopgap before NVIDIA releases updated Turing on 7nm and the more important reason is that I only update my GPU if the next one is at the very least 2,5 faster.

My last four GPUs have been: 7600 GT, 8800 GT, GTX 660, GTX 1060 (the current one).

 

[hat]

Neat. It beats my 1070 by a healthy margin for not too much money.

 

[Fluffmeister]

For a $100 jump in price compared to the GTX 1060, which i think will be more likely around $400-450 in real life circumstances as the 1070Ti is available around $420+ and the 2070 is too close to it (nearly halfway compared to 1060-1070)

Well that makes sense at the end of the day, Nvidia can't go on competing with themselves forever... so focus on what competiton there is.

Vega needs a price cut.

 

[GoldenX]

HBM is too expensive, Vega needs normal desktop cards with GDDR6.

 

[efikkan]

For a $100 jump in price compared to the GTX 1060, which i think will be more likely around $400-450 in real life circumstances as the 1070Ti is available around $420+ and the 2070 is too close to it (nearly halfway compared to 1060-1070). And still there are the AIB Vega 56 models cheaper for less than $350 with a $150 3 game bundle. A better choice compared to the RTX 2060.

Come on, be real
RTX 2060 is both cheaper and higher performing than Vega 56 in general. Discounts are exceptions, and should be evaluated when buying, but in general, you can't honestly claim Vega 56 is a better choice.

 

[ab3e]

GTX1060 competed with GTX980 and yet was sold as a x60 tier card and priced accordingly. Now the RTX2060 competes with GTX1080 and it's sold with a $100 price increase. No way you can defend Nvidia, this should be seen as a anti consumer tactic, a bunch of greedy green space monkeys.

 

[moproblems99]

I'm not sure what to make of the late high profile figures abandoning that boat though.

Well, considering the state of the GPU division, is it really a loss? A shake up is well needed. Fresh blood. Or scapegoat...

Navi better be good. xx60 matching previous gen flagship this is just sad for RTG. Not to mention the perf per watt.

Considering that Pascals/Vegas are ~3 years old, I should hope so. It would be a disappointment for them not to be.

someone will always steal the show with halo products ,sweet talk and unorthodox practices

They need a way to convince the average consumer that their product is just as good or better than Nvidia. When they had better products (HD5000 vs GTX 400), people still flocked to Nvidia even though, ironically, the 400 series has all the problems AMD has had recently. They can't have that this time. Beyond cleaning house in the GPU department, I would also clean us in the PR/marketing because they aren't that good. Poach some from Nvidia as they can convince people that corn kernels in a turd are really gold nuggets.

I am no longer hoping too much for RTG to improve. They are really far behind. I have more faith in Raja’s effort at Intel

That is what they all said about AMD with CPUs. Look where we are now. I'm not sure why you have hope for Raja at Intel. Isn't he the one who developed Fury, Polaris, and Vega which you love so much?

EDIT:

GTX1060 competed with GTX980 and yet was sold as a x60 tier card and priced accordingly. Now the RTX2060 competes with GTX1080 and it's sold with a $100 price increase. No way you can defend Nvidia, this should be seen as a anti consumer tactic, a bunch of greedy green space monkeys.

So, why should they sell the 2060 that performs better than the Vega 64 in some cases for less than the Vega 56?

EDIT 2:

It is actually amazing that the 2050 is going to perform fairly close to the Vega 56 or maybe even the Vega 64. That might be the largest gap I have ever seen.

 

[Fluffmeister]

Pascal and Vega didn't launch at the same time, Vega was the dream team red clung too. Hell with PhysX now open source, and the usual suspects picking holes Nvidia's day-0 VRR across multiple dodgy FreeSync displays support, before the driver is even available... suggests to me at least, you guys really hope AMD turn up soon.

 

[Deleted member 50521]

With high HBM pricing i think the better choice is just stop production of Vega as a whole. They need Navi asap.

 

[TheinsanegamerN]

They need a way to convince the average consumer that their product is just as good or better than Nvidia. When they had better products (HD5000 vs GTX 400), people still flocked to Nvidia even though, ironically, the 400 series has all the problems AMD has had recently. They can't have that this time. Beyond cleaning house in the GPU department, I would also clean us in the PR/marketing because they aren't that good. Poach some from Nvidia as they can convince people that corn kernels in a turd are really gold nuggets.

That's quite misleading. The 5000 series from AMD sold INCREDIBLY well, and it was the closest AMD ever came to achieving majority market share VS nvidia. AMD's driver support, however, was BEYOND abysmal, which is the only reason the 400s sold at all outside of the surprisingly competitive 460/465. This was absolutely the time you kept multiple different catalyst drivers on tap depending on what game you wanted to play, because every fix brought 3 new bugs in different titles, and known issues took months if not years to fix.

They then lost that market because AMD, deciding the 5000's were great, immediately pulled a rebrandeon and rehashed the ENTIRE lineup. Nvidia struck back with the 500s at the time and caught AMD completely off guard, they had assumed Fermi was beyond repair. They rushed production and final release of the 6900s, which were not enough to close the gap and nvidia managed to, somehow, reclaim the performance crown with the hot, power hungry fermi chips, while AMD let their smaller more efficient chips rot on the vine. After ONE GENERATION.

AMD absolutely deserves the position they are in. They dug this hole over the course of a decade. The current incompetence of RTG, getting beaten by mid range nvidia cards, is the logical conclusion of a GPU maker that simply is incapable of consistently churning out decent products and supporting them properly. Ditching the head of marketing, losing Raja, and other high profile losses are great for AMD, the old baggage was just as bad as Hector Ruiz, Dirk Meyer, and Rory Read, the unholy trinity of C suites that just about killed AMD by letting go of the mobile line instead of licensing to qualcomm, green lighting bulldozer, and wasting hundreds of million on things like SeaMicro.