AMD, Intel and Nvidia driver issues and last recommended version

[ryrynz]

and This is the definition of contradictory opinions.


No. These are priced attractively to sell now, that's the point. AMD will always be a competitor, it's to tempt people to upgrade now and upgrade they will, just look at all the cards going up for sale now on the second hand markets. AMD will lose some GPU market share likely this year and next as a result, it's a great release for Nvidia and sales will surge.

[NikosD]

No. These are priced attractively to sell now, that's the point. AMD will always be a competitor, it's to tempt people to upgrade now. AMD will lose some GPU market share likely this year and next as a result. nVidia is afraid of Big Navi and tried to make AMD believe that is not worth it to fight in the high-end segment since Big Navi of 80CUs will be a big die eating a lot of TSMCs wafers of AMD and with the aggressive pricing of 3080 and 3070 it needs to cut the margins a little more.
I said earlier that if AMD is going that way, it will be bad for customers leaving the monopoly of high-end segment to nVidia.
AMD has to compete with everything it has in its hands, releasing Biggest Navi of 80CUs and we have to support it.
If it does this, it will certainly INCREASE its GPU market share as there will be competitive products in the high-segment for the first time after many years.
If AMD doesn't release a high-end product after the extremely aggressive move of nVidia releasing a 320 Watt GA102 die product calling it just "3080" in the same price of the poor 2080, then it will certainly lose GPU market share and will support unintentionally the monopoly of nVidia in this segment.
I hope the margins to be OK for AMD and release the Biggest Navi of 80CUs and customers support it.

[NikosD]

And after writing the above post, I feel obliged to ask you to keep a moment of silence for those who bought Turing RTX 20 series cards, especially the 2080 Ti owners probably while reading the "Just buy it!" article of Tom's Hardware and after the extraordinary fear of nVidia because of Big Navi which made it release Ampere this way.

RIP brothers & sisters, owners of RTX 20 cards.
May the ebay be with you.

Also, I have to ask you to pray with me for all those nVIDIOTS who are going to buy 3090 cards for gaming.
They are the future RTX 20 owners with nVidia's next GPU release, but they don't know it yet -although they should.

Pray for them until the next RIP in around two years...

[chros]

And after writing the above post, I feel obliged to ask you to keep a moment of silence for those who bought Turing RTX 20 series cards
...
RIP brothers & sisters, owners of RTX 20 cards. May the ebay be with you.
:D That's a good one :) Let's be frank, those cards were crap without having competition.

Btw, M$ just confirmed the price of the new small Xbox: $300.
I hope that's a good news for the small cards in this generation. :)

Pray for them until the next RIP in around two years...
I'm sure that there won't be anything big in next years, see these posts why: link (https://forum.doom9.org/showthread.php?p=1922785#post1922785) and link (https://forum.doom9.org/showthread.php?p=1922791#post1922791)

3) NAVI was based on RDNA and according to Techspot the NAVI10 (5700 XT) chip of 10.3 billion transistors with a TDP of 225 Watts has the same performance with the TU104 (2070 Super) of 13.6 billion transistors with a TDP of 215 Watts
Huhn has a point, those were on 7nm already, that was not impressive at all. Let's see this year :)

with the aggressive pricing of 3080 and 3070
Guys, don't call this aggressive, it's still freaking high :)

xx70 is just a number.
That's where you are wrong: it's *not just* a number.
It's a price bracket and a performance bracket as well in a given generation!

if it is called 3070 for a 1000€ or 500€ doesn't matter
It does matter and matters a lot! See the 2 links above.

[huhn]

what link?

and no matter how you turn it it's just a name it does not mean that with a xx70 you will get a payable but not the best card if you interpret it as that then that's your own problem.
they can go and name then as they want and they can price them as they want as long as you get more performance per dollar you are not robbed.
if they sell you a product for half the price with the same performance as the last flag ship and massively cheaper in price and performance as the concurrence then this is fine no matter how it is named.
if you want a card for 350-450 wait for one but as long as this card also has the same performance increases per dollar is totally fine.

[NikosD]

Troll has a point, those were on 7nm already, that was not impressive at all. Let's see this year :)
In the name of Mighty Thor and the Elder Gods, how much power efficiency gain do you think AMD pulled out from TSMC 7nm vs TSMC 12nm of Turing ?
AMD targets clocks and performance mainly for its GPUs and power efficiency later.
You can't have both performance and power efficiency.
You target performance or power efficiency using a specific TSMC node.
They needed a small and fast die with IPC and power efficiency comparable to Turing and they did it.
We don't compare performance per area here, where the NAVI10 using TSMC 7nm simply destroys Turing with TSMC 12nm.
It's a TDP of 225 Watts for NAVI10 using fewer transistors vs TDP of 215 Watts for Turing that really matters.

[huhn]

you are comparing a 12 nm chip which actually is 16nm with 7 nm and you are not winning in clock rate and in power consumption.
the transistor number are 10800 to 10300.
does NAVI have dedicated INT32, RT and tensor cores?

just to hammer this in while a huge part of the turing cards has "wasted" or not often used transistors NAVI could still not win with a massively better process process.

and i have a hard time ever believing 7nm 251mm² chip is cheaper then a 12nm 445mm²...

[chros]

huhn is right in the above.

and no matter how you turn it it's just a name it does not mean that with a xx70 you will get a payable but not the best card if you interpret it as that then that's your own problem.
It's not my problem :) You don't understand what I'm telling. I don't mind that.

[NikosD]

As I wrote above TU104 of 2070 super has 13.6 billion transistors not 10.8 vs 10.3 billion of NAVI10 of 5700 XT.

If I have to repeat my posts two times in order to understand it correctly, it's a huge waste of time for me.

@ryrynz for example didn't read my post that RTX IO isn't supported by Turing GTX and he replied that it is supported by Turing RTX which is obviously a different card.

Read my posts two and three times before you reply.

[SamuriHL]

Especially if they crank up the memory, but, will it be GDDR6X or the equivalent? At the price points they're going to have to compete at, I have a hard time believing they're using a 7nm process AND fast memory. I bet they sacrifice memory speed to give you gobs of it.

A few things to correct, because ignorance reached the sky and is still going up.

So, before ignorance reaches the moon I have to say:

1) GDDR6X is a custom-made memory produced by Micron only and while we don't know what will happen in the future, the only customer for now is nVidia.
So Big Navi is going to have GDDR6 or if AMD surprises once more will have HBM2.

2) There is memory faster than GDDR6X and it's called HBM2 and HBM2e from Samsung, SK Hynix and AMD



Read my posts two and three times before you reply.

Yes, reading skills definitely have fallen by the wayside here. :) There's a reason I put "or the equivalent" in my sentence that you appear to be calling ignorant. We don't know what memory AMD will be utilizing. I VERY much doubt that it's going to be HBM2 at the price points that are rumored. It's likely going to be GDDR6, but this is pure speculation at this point. And I maintain what I said...to compete with the 3080 it will need faster memory, not just gobs of slower memory. I have no doubt AMD is going to best the 3070 with their top end Big Navi card. But I have SERIOUS doubts about it reaching the 3080. That's just my opinion until we see some comparisons in the real world for both cards instead of paper specs, for which we don't even have on Big Navi yet. Everything right now is mere speculation and conjecture.

[huhn]

a 5700 Xt is about as fast as a 2070 not a 2070 S.

[NikosD]

Zen3 on October 8th and Radeon RX 6000 on October 28th.
It's official from AMD:
https://videocardz.com/newz/amd-to-introduce-zen3-on-october-8-radeon-rx-6000-series-on-october-28

I can see that rumors will be huge till end of October for RX 6000.

[nevcairiel]

I hope Zen3 actually launches in October as well, and not very much later. Been eagerly awaiting those to rebuild my setup from ground up.

[chros]

And nvidia will lunch the 16GB and 20GB versions of their 3070/3080 in the middle of October as well, then 2 weeks later the 3060.
Actually this is a really clever move: they can sell the higher memory configs more ($100+ ?). Of course they could have give it away for "free" :D
And they can say: "we have a $500 card that is really good, if the memory isn't enough for you then buy the higher models" :)

[huhn]

yeah not going to get a new GPU before i get a new CPU. i really need zen 3.

[videoh]

RIP brothers & sisters, owners of RTX 20 cards. My 2080 Ti is not going to stop doing all the awesome things it does (and has done) for me just because a 30xx is released. Worst case, I can move it to replace the 1080 Ti in my backup machine (which will in turn replace the 1050 Ti in my second backup machine). But more likely I'll put my 3090 Ti in a new kick-ass machine with latest CPU goodness and PCIe 4. I've been benefiting from my 2080 Ti for almost two years so there's no reason for any regrets. Long live nVidia!

All this guy has is rhetoric, insults, forlorn hopes, and sour grapes. I will pray for him.

[NikosD]

All of you guys are a bunch of clueless nVidia fan boys, irresponsible customers and have real issues with facts and the real-world.

I see no real and solid arguments against mine.

All I see is the reproduction of the mainstream propaganda of nvidia's marketing BS.

You are living in a bubble and you are responsible for the monopoly of Intel and nVidia with all of the consequences of these monopolies - high prices and milking.

I feel pity.

[SamuriHL]

I hope Zen3 actually launches in October as well, and not very much later. Been eagerly awaiting those to rebuild my setup from ground up.

Traditionally I've stuck with Intel and do like my 9900k. However, they have definitely fallen behind and I no longer even remotely care about SGX and "official" UHD playback anymore. (I was only moderately interested before when I got the 9900k but now not even a little bit). So AMD is an option for me and I really want to see how balls out these Zen3 procs are. PCIe 4 nvme. Mmmmmmm :)

[huhn]

All of you guys are a bunch of clueless nVidia fan boys, irresponsible customers and have real issues with facts and the real-world.

I see no real and solid arguments against mine.

All I see is the reproduction of the mainstream propaganda of nvidia's marketing BS.

You are living in a bubble and you are responsible for the monopoly of Intel and nVidia with all of the consequences of these monopolies - high prices and milking.

I feel pity.
the unbeatable i'm right you are wrong argument i'm disarmed here.

from the guy with an intel CPU btw.

[SamuriHL]

Latest leak/rumor shows big navi with 16gb of GDDR6 at 256 bit bus. It's entirely possible this is not finalized, but, if it is, that hardly is going to beat a 3080.

[chros]

Do you have a link?

[NikosD]

Latest leak/rumor shows big navi with 16gb of GDDR6 at 256 bit bus. NO NO.
I have written before a few times that Big Navi21 of 80CUs should have 20GB of VRAM and larger bus.
The 16GB VRAM Navi should be the one competing with 3070.
If this that I describe is not the case, then AMD has lost this battle by far.

[ryrynz]

You are living in a bubble and you are responsible for the monopoly of Intel and nVidia with all of the consequences of these monopolies - high prices and milking.


Nvidia is going hard and delivering good prices too for what's being developed, Intel milked quad cores for far too long and didn't innovate quickly enough, they're paying the price now and so will anyone that languishes too long.

Competition is always there waiting in the wings. Like I said, nobody is complaining about how things are priced and what we're getting in any hardware market.

Can't have a fairy tale capitalist market, it's impossible. There are hardware limits, and you can't know the entire costs of running a large scale business.
The market is far from the doom and gloom you seem to view it as.


If this that I describe is not the case, then AMD has lost this battle by far.

The battle is far from lost, I'm sure they'll have a reasonably successful launch, now that I know that news of Big Navi will come very shortly, perhaps AMD will not lose market share as I expected but it will still be a tough battle.

This launch may be where things begin to turn around for them depending on the benchmark results and internet feedback.
We do want them to do well, the sentiment towards them is getting much better now they are delivering what the market wants, I hold hope.

[NikosD]

Competition is always there waiting in the wings. Like I said, nobody is complaining about how things are priced and what we're getting in any hardware market. NO.
Competition in the x86 CPU arena stopped in 2006 with the release of Core 2 Duo CPU and the dramatic mistakes of AMD.
x86 CPU arena like GPU arena are both near to monopoly situations or pseudo-duopoly if you prefer.
Intel has allowed to exist extremely few licenses and only AMD survived from this situation.
AMD brought up competition again, after 11 YEARS in 2017 with Zen architecture.
From 2006 to 2017 there was ZERO COMPETITION due to the pseudo-duopoly or covered monopoly of a Goliath that is Intel and David that is AMD.
No other real competitors in x86 CPU arena. ZERO.
This is a situation caused by anti-competitive practices of Intel for decades.
Take a look at the ARM architecture CPU arena.
Apple, Samsung,Qualcomm, Huawei (before Trump), MediaTek and Marvel, Ampere, Nuovo and so many other options for ARM.
THIS IS HEALTHY COMPETITION.
x86 "competition" is a disease, an illness, a disgusting, manipulating thing of Intel and its fanboys supporting this.
If AMD didn't bring Zen architecture after 11 YEARS, NOBODY of you would complain about the disease of Intel's monopoly and domination.
A lot of you are very unhappy with the situation and you cry like babies during the middle of the night, because Intel has become AMD's bitch since 2017.
You are Intel's slaves...You belong to Intel due to a Stockholm syndrome probably.
Take a look at this forum posts, not only this thread and you will find many of Intel's slaves.
The same situation is for the GPU arena.
In the beginning there were many GPU companies but the leather-jacket-man did many and very bad things and like a miracle nVidia managed to have another covered monopoly or pseudo-duopoly of AMD and nVidia this time.
You know, like AMD and Intel in x86 CPU arena.
AMD is too small to compete simultaneously with these two large companies.
Actually is too small to compete separately with each one, so to fight two wars at the same time is impossible.
The situation is ideal for both covered monopolies of Intel and nVidia to fight a little, tiny company like AMD.
They can legally say that they are not monopolies, because look there is AMD, but in the real-world we all know that Intel had 99% of the server CPU market before 2017 and nVidia still has 80% market of the discrete GPUs
If these are the numbers of YOUR COMPETITION, then you have clearly an issue of understanding the reality.
The battle is far from lost, I'm sure they'll have a reasonably successful launch, now that I know that news of Big Navi will come very shortly, perhaps AMD will not lose market share as I expected but it will still be a tough battle. NO NO.
If AMD releases the Biggest Navi, the highest and fastest of its RX 6000 products at the 3070 level of performance, then we will have a RDNA aka NAVI situation again.
I will say this one more time:
If the fastest product of Big Navi is at the 3070 level, then this means that RDNA2 (RX 6000) will compete with Ampere, like RDNA (RX 5000) competes with Turing aka 80% of discrete cards to nVidia.
WE DESPERATELY NEED the release of Big Navi of 80CUs that I personally believe can compete with 3080 and Big Navi using 20GB of VRAM with lower power consumption vs 10 GB of VRAM and higher power consumption of 3080 and this could bring some GPU market to AMD along with the lower tier models.
AMD and we as customers we can't afford AMD to bring its fastest RX 6000 model competing with the lowest of Ampere high-end cards.
THERE IS NO COMPETITION in a situation like this.

[huhn]

there is no need to have the biggest card they have barely any market share:
https://store.steampowered.com/hwsurvey/Steam-Hardware-Software-Survey-Welcome-to-Steam

it's pretty much look how great the 2080 ti is i get a 2060.

[NikosD]

I' ve just read a recent statement from the co-founder of Arm company, Hermann Hauser, to NS Tech regarding the possibility of nVidia buying Arm.

He said:
They are the semiconductor company that can buy Arm to destroy it—and it is very much in its interest to destroy Arm because they [would] gain a lot more than the $40 billion that they would pay for it
This statement pretty much summarizes what kind of bastard company ENVYdia really is.

[SamuriHL]

Do you have a link?https://wccftech.com/amd-radeon-rx-6000-big-navi-graphics-card-pictured-16-gb-gddr6-memory/

Remember these are leaks and speculation. Nothing to get all hot and bothered about just yet. But if this is true, and that's not guaranteed, then they will be competing with the 3070. Figure slightly faster with more ram for about 50 bucks higher. This may explain the supposed leaks by Lenovo of a 16gb 3070ti.

Quite frankly I want amd to succeed. I always have. I go back a very long ways in pc building and started out with amd for over 10 years until Intel annihilated all competition. Anyone remember cyrix? I used amd GPUs until the Nvidia 4xx cards, as well. I'm very seriously looking at the zen 3 but if these leaks turn out to be accurate about big Navi then I can't yet take them seriously for GPU.

Sent from my SM-G975U using Tapatalk

[butterw2]

Nvidia going with Samsung for Ampere is actually good for competition (vs TSMC) in the sub-12nm semiconductor manufacturing market.

If price/performance isn't there in mid/entry price range, don't buy, buy used, or buy a console instead. Not much point in complaining about lack of competition, Nvidia has a rockstar (or at least leather jacket) CEO and wants to become a datacenter company.

[chros]

I don't think that AMD is much bothered by desktop GPUs. They have the console market now and that was an important step for them, that's where the money comes from (along with CPUs, datacenter products). The desktop market compared to the rest is mimimal.

[nevcairiel]

I don't think that AMD is much bothered by desktop GPUs. They have the console market now and that was an important step for them, that's where the money comes from (along with CPUs, datacenter products).

"now"? :)
Consoles from the last ~7 years were based on AMD APUs, and both the XBox 360 and Wii before that used ATI GPUs - in combination with other CPUs, however.

NVIDIA, on the other hand, has been making more money from datacenter the last quarters then gaming products.

[SamuriHL]

But nVidia has the SHIELD and, uh, the Switch. Come on, those are killer! :D LMAO

[el Filou]

The comparison of TSMC 7nm vs Samsung 8nm for Ampere using GA100 vs GA102 gives a 47% higher density for TSMC and better efficiency.You can't compare the transistor and power budgets of a compute-only chip with a full graphics chip.As I have already mentioned, Ampere's TFLOPS are "fake" meaning they advertise a maximum theoretical value not reachable in real-world.As people have explained to you already but you seem to ignore, that's the case for all TFLOPS figures advertised by all GPU manufacturers.We don't compare performance per area here, where the NAVI10 using TSMC 7nm simply destroys Turing with TSMC 12nm.
It's a TDP of 225 Watts for NAVI10 using fewer transistors vs TDP of 215 Watts for Turing that really matters.Consumers (except maybe you?) don't care about "performance per chip area", only about performance per price or sometimes performance per watt.In the beginning there were many GPU companies but the leather-jacket-man did many and very bad things and like a miracle nVidia managed to have another covered monopoly or pseudo-duopoly.What bad things? Releasing technically superior (both in hardware and drivers) products that customers bought while Number Nine, Rendition, Intel, S3, PowerVR and Matrox (in decreasing order of awfulness) couldn't keep up and while 3dfx sabotaged itself by deciding to cut its board manufacturing partners? :rolleyes:

[NikosD]

As people have explained to you already but you seem to ignore, that's the case for all TFLOPS figures advertised by all GPU manufacturers. You and anyone else who have the guts to compare by any means the RDNA/TURING/RDNA2 TFLOPS vs Ampere TFLOPS and find them equal, I would declare you utterly ignorant and dangerous trolls of tech, that you should abandon writing in technical forums.
Nobody has explained to me anything of value.
You are still writing your opinions and mainly garbage.
Like this one you posted here.
I'll repeat it in case you eventually understand it.
Not all TFLOPS are equal.
It has already been reported by nVidia itself as I wrote before that the 30TFLOPS 3080 card has "up to 2x" the performance of the 10TFLOPS 2080 card.
Judging by the architecture of the Ampere cores (SM) it has already been processed that in real-world 1TFLOP of Ampere is 2/3 TFLOPS of Turing.
So the 3080 card of 30 TFLOPS produced by Ampere cores is equal around to 20 TFLOPS produced by Turing cores or "up to 2x" of the 2080 card.
This is the real-world performance of Ampere card.
Who cares what nVidia writes in its specs as "maximum theoretical value TFLOPS" of Ampere ?
Consumers (except maybe you?) don't care about "performance per chip area", only about performance per price or sometimes performance per watt.What bad things? You have serious issues understanding simple sentences not even technical.
I referred to performance per area as a metric that we don't care here for comparison reasons.
I mentioned it as a metric we don't compare, but instead we tried to compare performance per watt.
Next time be more careful with what you read, understand and write.
Unless you want to be the second troll of this forum.

[NikosD]

I have to be honest to this forum, like I always try to be.

The last few days dozens of tweets and more than 5 technical hardware news sites that I respect their opinion, have reported that Big Navi of 80CUs, the largest Navi21 according to rumors, is going to have 16 GB of VRAM and 256 bit bus.
Also, most of them if not all, seem to agree that Big Navi of 80CUs is going to compete with 3070 after all and not 3080.
On the other hand, I against all these respectful sites, believe that if AMD decides/wants/needs to release the Biggest/Baddest/BigFucking NAVI of all time using 80CUs (probably Navi21) is going to have 20 GB of VRAM and 320 bit bus.

So, it's ME against THE WORLD.

We'll see...

Now, according to a tech journalist of Extremetech site called Joel Hruska, Big Navi of 80CUs is going to be slower than 3070.
He claims that because of 3070 having 5888 cores and Big Navi of 80CUs having 2x40CUs = 2x2560 cores = 5120 cores, it will struggle to compete even with 3070.

Thank God there are actually so incompetent tech journalists comparing wrong figures in such ridiculous way that I'm now more confident that the 80CU RDNA2 chip is going to have 20GB of VRAM and will come close to 3080 with a suitable clock.

I'll leave it to the users of this forum to understand why this "tech journalist" is writing nonsense.

After so much training by me, you should be able to figure it out.

If not, I failed as a teacher.

[chros]

"now"? :)
Consoles from the last ~7 years were based on AMD APUs, and both the XBox 360 and Wii before that used ATI GPUs
:o Oh, sorry, I stopped playing a long time ago.

I just realized why nvidia was/is scared: M$ will provide netflix type of subscription for games: $25 for small and $35 for big Xbox per month for 2 years including the device. I think this is big for gamers (well not for me :)).

[NikosD]

A really interesting read for this afternoon:
https://www.mooreslawisdead.com/post/nvidia-s-ultimate-play

[ryrynz]

Therefore, I think this is Nvidia’s Ultimate Play: Intentionally causing an initial dearth of Ampere stock, allowing “supply and demand” to inflate the street price of Ampere when those beautiful $699 Founders cards instantly sell out, and then ultimately forcing AIBs to sell most of their models well above MSRP due to the required BOM Cost-Downs…that will be in ample supply once the street price is elevated.

So letting the market do it's thing, this is what a smart company does.. I applaud them for it, all hail Capitalism.

[NikosD]

These are the tricks of an abusive, ruthless, manipulative company acting as monopoly to the free market and NOT LETTING free, capitalistic market work as it should, in order to keep margins and prices high, while it looks like is selling in lower prices.

Marketing propaganda and nVidia fan boys help a lot here to increase the hype.

Hail to nVidia means hail to crippled "free" market, hail to high inflated prices for the customer.

Anyone who doesn't have nVidia shares in his hand or not being a mindless fan boy should see this and criticize this behavior.

[chros]

3060Ti leaks (https://www.youtube.com/watch?v=4-BWfmBP_WE), just like last year the 2060S, it's pretty close to 3070 (same die, lower cuda cores, etc.). If that will be the case then I don't see the point of 3070 since it's not good enough for 4K gaming with its 8GB, so a 3060Ti would be cheaper to get (let alone madvr usage). The only question remains with how much: if they follow last year prices (just like they did with higher end models) then it will cost $400. And that's way too high for me ...

The other thing that there are really confusing "leaks" about RDNA2, some state it's crap, some state it's an Ampere killer :D
How on earth can this happen with AMD at this stage that I don't understand ... :)

[SamuriHL]

Because there's 2 chipsets for AMD, and I suspect the leaks so far are on the lower end one. NO ONE should be buying a 3060 or 3070 until the AMD reveal. I'm still planning on getting my 3080 this week if I can find one. On the lower end cards, where AMD will absolutely compete, there is going to be a price war I think. Fun time to be a consumer.

[CZ Eddie]

Just noticed these up on the USA BestBuy.com website.
It says all three will be launching this coming week on the 17th.
https://www.bestbuy.com/site/searchpage.jsp?st=rtx+3080&_dyncharset=UTF-8&_dynSessConf=&id=pcat17071&type=page&sc=Global&cp=1&nrp=&sp=&qp=&list=n&af=true&iht=y&usc=All+Categories&ks=960&keys=keys
https://i.ibb.co/hH12Kwp/Untitled.jpg

[huhn]

i'm not far away from not caring about AMDs next GPU anymore after the nightmare of my 480 and the current nightmares with my 5700 XT.

freesync is a lie on these cards cap a game to 200 FPS make an overlay on and see the monitor OSD displaying between 120-240 while the game and the overlay clearly displays 200 FPS all the time the card isn't outputting it it just outputs what ever it wants to. take a much slower nvidia card and see how the monitor OSD and the overlay/ game FPS are spot on with an AMD verified freesync screen there are no word for this.

the pathetic QA. they just sometimes release driver where you can't switch the refresh rate they didn't even check if the OS get's reported the correct refresh rates i mean the driver can't switch refresh rate so no point in checking if the OS can right?

virtual super resolution = stick at highest refresh rate because reasons.

reporting the wrong native resolution they love to do that with TV by saying 4096x2160 is "native" so you get from time to time a user asking why is my screen cut of...

there wasn't a signal day in the life of the 5700 XT where deint ever worked correctly.

[chros]

Interesting to see the different cooler designs, for the 1st time nvidia did something completely different than its AIBs :) I'll be curious about the tests regarding to this, which will be more efficient.

[chros]

i'm not far away from not caring about AMDs next GPU anymore after the nightmare of my 480 and the current nightmares with my 5700 XT.
:D I can understand, no one can blame you ...

[SamuriHL]

The FE card has been up there since shortly after the reveal. That's likely where I'm going to try to order one from on Thursday. I can't wait and hope I'm able to grab one.

[NikosD]

The guy from MLID that I posted his article, says that Big Navi is not going to be that "Big" eventually.
The die is small and for the first time power consumption could be lower than the competing Ampere cards.
But he also says that memory bus and memory bandwidth will be lower, so no 320bit I guess - probably it's going to be 256 bit and 16GB even for Big Navi21.
At the same time he seems positive that is going to crash 3070 and compete directly with 3080.
With so low memory bandwidth ?
All these estimations are for rasterized performance.
Ray tracing is going to be faster than Turing and slower than Ampere, according to him.
I will definitely wait for RDNA2 cards which are ready months ago and will be released with polished drivers, probably for the first time in history.
But AMD cards are famous for keeping their value during their long life of supporting drivers and performing better in the long run than nVidia cards.
For example the RX 470/RX 570 card is clearly faster these days than 1060 3GB card and RX 480/RX 580 card is faster than 1060 6GB card.
These nVidia Pascal cards were faster than Polaris cards using their initial drivers.
Anyway since I saw also the video of Steve from Gamers Nexus bashing Radeon Marketing team for the hilarious history of misleading hype and underperforming cards, I hope this time history will not repeat itself with "Poor Volta" ridiculous hype.

I'm confident that RDNA2 cards will be more competitive than ever to nVidia's top Ampere cards.

[varekai]

Hello!

GPU NVIDIA Geforce GTX 960, drivers 398.11 win7 x64.
Would this old GPU benefit with latest drivers 452.06?

[chros]

I doubt it will.

Something else: nvidia just bought ARM (https://www.youtube.com/watch?v=1nYPNXwpNS4) for $40 billion. I have mixed feelings about this ... I like one of the comments: "The birth of Skynet" :D

[ryrynz]

Anyone who doesn't have nVidia shares in his hand or not being a mindless fan boy should see this and criticize this behavior.

The consumer cares only about what they can buy and it's perceived value. Everyone with a 30 series Geforce is going to be very happy indeed. *Begin applause* the world moves on as the market continues as it always has.


Something else: nvidia just bought ARM (https://www.youtube.com/watch?v=1nYPNXwpNS4) for $40 billion. I have mixed feelings about this ... I like one of the comments: "The birth of Skynet" :D

An excellent move, finally the beginning of the end for x86, lots of cash to put into finally putting this old bird to roost. I called this happening back in 2006.

[SamuriHL]

Where the hell is the new driver?! I'm starting to get impatient with them. I know it exists because all the people out there who've gotten prerelease cards are using them. We need to see if their claims of a madvr fix holds true.

Sent from my SM-G975U using Tapatalk