2027/2028 Nvidia/AMD/Intel GPU speculation

[Marees]

RDNA 5 Is Coming Earlier Than Anyone Expected. What That Means for Your Next Build​

Plan for Q3 (2027), be pleasantly surprised if it’s Q2, and don’t lose sleep if it slips to Q4.

The DRAM and NAND Problem Nobody Is Talking About Loudly Enough​

RDNA 5 arriving in mid-2027 is not automatically good news for build costs. Tom’s Hardware reported in July 2026 that AI-driven demand is keeping DRAM and NAND prices elevated through at least Q3 2026, with no clear signal of sustained relief. GDDR7, the memory standard RDNA 5 is expected to use, is being consumed heavily by AI inference hardware before it ever reaches consumer graphics cards. That supply pressure doesn’t vanish just because AMD hits a launch date.


What this means practically: RDNA 5 cards may launch with strong performance and painful MSRPs.


GDDR7 memory, expected on RDNA 5 flagship and upper-mid cards, is under supply pressure from AI hardware demand through at least late 2026. If that pressure doesn’t ease, expect RDNA 5 MSRPs to open higher than historical equivalents. Mid-tier cards will likely hit reasonable street pricing two to three months after launch as channel inventory catches up.

 

[El_Capitan]

Any new GPU coming out will have had to pay the absurd AI-caused DRAM tax, and will be more expensive. My hope is that they're holding their GPU's back from release until the AI RAMpocalypse is over, but they're the ones directly benefitting from it, so I doubt it.

 

[Marees]

Any new GPU coming out will have had to pay the absurd AI-caused DRAM tax, and will be more expensive. My hope is that they're holding their GPU's back from release until the AI RAMpocalypse is over, but they're the ones directly benefitting from it, so I doubt it.

they are holding back the "budget" cards

AT2 will launch in the $500 to $1000 range next july / Q3 2027

AT0 above that will probably not release for gamers

unfortunately the "budget" AT3 & AT4 cannot be released because AMD made the "foolish" decision of making them lpddr rather than gddr

that means minimum 32gb vram — for a BUDGET card !!!

the end result is for $500 & below, RDNA 4 will continue service until RDNA 6 (2029 / 2030)

the AT3 & AT4 will release as part of framework AI PCs with 128gb/256gb ram !!!!

 

[LukeTbk]

unfortunately the "budget" AT3 & AT4 cannot be released because AMD made the "foolish" decision of making them lpddr rather than gddr

that means minimum 32gb vram — for a BUDGET card !!!

Or if quite limited bandwith but large vram is quite useful, just not make them budget cards...

B70 are not that powerful for 32GB (and juste 600 gb/s...), sometime even the mini RTX 2000 beat them, they can sell them for a lot of money. The gddr vs lpddr price split will determine how bad-good of a pick it was I feel like, maybe they will need to keep them for the small form factor AI Pc going crazy high... if Moe and compression keep them interesting even at very low bandwith.

 

[Marees]

RDNA 5 to beat Blackwell in CPU bottlenecked situations — MLID

https://www.technetbooks.com/2026/08/amd-rdna-5-graphics-cards-target-gaming.html

 

[WilyKit]

RDNA 5 to beat Blackwell in CPU bottlenecked situations — MLID

https://www.technetbooks.com/2026/08/amd-rdna-5-graphics-cards-target-gaming.html

So almost never gonna beat it

 

[El_Capitan]

I have a rare AMD Ryzen AI Max 395+ with 16GB LPDDR5 8000MHz mini-pc incoming. Not good for LLM's due to the lack of memory, but it'll do well for gaming using the iGPU. I'll have to figure out how much memory to allocate to the 8060S. It'll likely perform somewhere between the RTX 4060 and RTX 4070. It has a dedicated OCuLink port (as well as USB4), so I could always just use an eGPU. It's definitely a better deal than the Steam Machine (and performs better). I'm still wondering whether to install Windows 10 Pro on it (it comes with Win 11 Pro), or Steam OS.

Also, I've been looking into rumors that a 9000G series could possibly come out this year. I doubt it'll have a 8060S, but it's possible. The 9700G with a 8060S, and the 9600G with a 8050S would be nice.

 

[ZeroBarrier]

So almost never gonna beat it

Got to have those positive marketing claims whether they are only theoretical or not.

 

[LukeTbk]

RDNA 5 to beat Blackwell in CPU bottlenecked situations — MLID

https://www.technetbooks.com/2026/08/amd-rdna-5-graphics-cards-target-gaming.html

If blackwell which is almost Lovelace is the bar for RDNA 5.... the 4080 super, a 2022 gpu people called a 4070 will still be near the top of all non-nvidia GPU until spring 2029 ?

or course rdna 5 on TSCM 3 should beat very old by then tsmc 5 family node Blackwell and not just during CPU limited scenario, always and of course. That must be not what MLID said (or misleading), rdna5 will beat blackwell even more in cpu bottleneck situtations or something specific like that.

 

[Marees]

If blackwell which is almost Lovelace is the bar for RDNA 5.... the 4080 super, a 2022 gpu people called a 4070 will still be near the top of all non-nvidia GPU until spring 2029 ?

or course rdna 5 on TSCM 3 should beat very old by then tsmc 5 family node Blackwell and not just during CPU limited scenario, always and of course. That must be not what MLID said (or misleading), rdna5 will beat blackwell even more in cpu bottleneck situtations or something specific like that.

looks like I got my architectures mixed up then

he said AT0 will beat 5090 in CPU bottlenecked situations

 

[LukeTbk]

he said AT0 will beat 5090 in CPU bottlenecked situations

Listening to it, seem to mean beat it all the time and that the 5090 TI would not be enough, because AT0 will be big and on TSMC 3N, the gap with blackwell and future RDNA 5 if it exist cannot make up for such a node upgrade.

There is no way a TSMC 3N AMD gpu need some * for when it would be faster than blackwell, or something went really wrong. It's competition must be TSMC 3 rubin...

He seem way too excited in the video about CDNA/AI workload deciding what RDNA 5 for game will look like cache wise, yes AI predictable stream prefer ultra wide HBM high bandwith to local cache, so it make sense to go that way, but gaming love cache a lot, the GLobal L2 being small per core is not some great news for gaming performance here, at least not necessarily so.

That not how it would report it if it was coming from Nvidia.... and he does not seem to present clearly how normal (what they had before and others have, that gL2, talk about it as if it was new)

 

[kram182]

Listening to it, seem to mean beat it all the time and that the 5090 TI would not be enough, because AT0 will be big and on TSMC 3N, the gap with blackwell and future RDNA 5 if it exist cannot make up for such a node upgrade.

There is no way a TSMC 3N AMD gpu need some * for when it would be faster than blackwell, or something went really wrong. It's competition must be TSMC 3 rubin...

He seem way too excited in the video about CDNA/AI workload deciding what RDNA 5 for game will look like cache wise, yes AI predictable stream prefer ultra wide HBM high bandwith to local cache, so it make sense to go that way, but gaming love cache a lot, the GLobal L2 being small per core is not some great news for gaming performance here, at least not necessarily so.

That not how it would report it if it was coming from Nvidia.... and he does not seem to present clearly how normal (what they had before and others have, that gL2, talk about it as if it was new)

He should stick to leaks - he's an AMD fanboy and an Nvidia hater so him and all his predictions go together and have panned out like icebergs and the Titanic

 

[LukeTbk]

Rumours from MLID that Rubin gaming desktop discrete GPU will be a bit late (as expected by everyone by now), H1 2027:


View: https://www.youtube.com/watch?v=CP98gJdcTCo

Not sure how much I believe that the source is actually much in the know (that the delay would be for AI to gobble up as much TSMC capacity as possible this year... really doubt tsmc compute die is what limit AI capacity and that they need to save from gaming gpu at all, complex packaging gaming GPU will almost certainly not use and memory is what limit AI deployment I would think, there is not that much mm of TSMC 3N into those rack...)

The crack down on leak do seem believable, they were able to make the announcement of DLSS 5 themselve... despite many studio in the know, there so much money in play now and worldwide geopolitical value that they could have deployed a lot of no-leak effort.

 

[tunatime]

Rumours from MLID that Rubin gaming desktop discrete GPU will be a bit late (as expected by everyone by now), H1 2027:


View: https://www.youtube.com/watch?v=CP98gJdcTCo

Not sure how much I believe that the source is actually much in the know (that the delay would be for AI to gobble up as much TSMC capacity as possible this year... really doubt tsmc compute die is what limit AI capacity and that they need to save from gaming gpu at all, complex packaging gaming GPU will almost certainly not use and memory is what limit AI deployment I would think, there is not that much mm of TSMC 3N into those rack...)

The crack down on leak do seem believable, they were able to make the announcement of DLSS 5 themselve... despite many studio in the know, there so much money in play now and worldwide geopolitical value that they could have deployed a lot of no-leak effort.

Unless they can do something to make it not cost $6090 who cares?

 

[zandor]

I think we'll be lucky to see a next gen gaming card any time in 2027. By this point in the last cycle we had lots of leaks on the upcoming specs. Now? Crickets. Nothing for gaming cards.

 

[LukeTbk]

AMD next-gen gaming card because they are partner heavy for the next gen (Xbox/PS6, etc...) has leaked a lot i think ?

Nvidia grock "aquistion" was kept on the downlow, DLSS5, Intel deal was also achieved to be announcement by them first.

So they are trying hard to keep lids on the leaks... but this would be much harder to keep secret, lot of third party, lot of lower level people, drivers stack, AIBs, hard to believe anything before Q3/Q4 2027, maybe an announcement in H1 2027

 

[zandor]

So they are trying hard to keep lids on the leaks... but this would be much harder to keep secret, lot of third party, lot of lower level people, drivers stack, AIBs, hard to believe anything before Q3/Q4 2027, maybe an announcement in H1 2027

Seems like that's where the leaks come from. It's not NV/AMD/Intel corporate. Once all the assorted other companies involved in actually bringing cards to market get involved is when we get the leaks. So the lack of leaks makes me think cards are a ways off. The rest of the supply chain has to be engaged months before launch. No leaks suggests that they haven't been, so launch is a ways off.

I've seen a little on next gen consoles. Not a lot, just something about a CPU with 3 full sized Zen cores and 8 compact cores. Not sure if that was XBox or PS. That actually makes sense for console gaming. I know people love to hate on e-cores, but the reality is games only have a small number of hot threads and the rest are just getting work done. So main graphics thread, main game thread, and an extra full size core if the dev needs it - seems reasonable.

 

[pippenainteasy]

Kepler L2 says MLID is pulling stuff out of his rear end, so I guess we'll see who is right.

 

[MistaSparkul]

View attachment 826183

Kepler L2 says MLID is pulling stuff out of his rear end, so I guess we'll see who is right.

So...This AT2 GPU seems like a 5080 competitor, which means the 4090 from 2022 will literally be faster than any and all AMD cards until at least 2028, imagine not being able to beat a 4090 for 6+ years.

 

[Marees]

So...This AT2 GPU seems like a 5080 competitor, which means the 4090 from 2022 will literally be faster than any and all AMD cards until at least 2028, imagine not being able to beat a 4090 for 6+ years.

gaming GPUs are low margin business. bragging rights are not worth burning the money for AMD

 

[MistaSparkul]

gaming GPUs are low margin business. bragging rights are not worth burning the money for AMD

It's not about bragging rights, this is how tech is supposed to go, at least that WAS how it went. Think back 10 years ago, a 2017 mid range card would easily destroy the flagship from 2012.

That was how tech progressed, the newest card from 3 generations onward would be much faster than the old flagships. Now here in 2026 going into 2027 the mid range lineups from both camps fail to beat the 4090 until well into 2028, and even then there is no guarantee that they actually will and the 4090 could still be faster than both Nvidia's and AMD's mid range lineups that come out in 2028. That is just sad, but I just think it's even more sad for AMD not being able to clear the 4090 for over a half a decade which is a long time in tech terms. Finally beating a 6 year old flagship isn't exactly bragging rights worthy at all, so rather than have something to beat the 4090 for "bragging rights", they should have something to beat the 4090 to prove they aren't absolutely pathetic.

 

[Marees]

It's not about bragging rights, this is how tech is supposed to go, at least that WAS how it went. Think back 10 years ago, a 2017 mid range card would easily destroy the flagship from 2012.

View attachment 826196

That was how tech progressed, the newest card from 3 generations onward would be much faster than the old flagships. Now here in 2026 going into 2027 the mid range lineups from both camps fail to beat the 4090 until well into 2028, and even then there is no guarantee that they actually will and the 4090 could still be faster than both Nvidia's and AMD's mid range lineups that come out in 2028. That is just sad, but I just think it's even more sad for AMD not being able to clear the 4090 for over a half a decade which is a long time in tech terms. Finally beating a 6 year old flagship isn't exactly bragging rights worthy at all, so rather than have something to beat the 4090 for "bragging rights", they should have something to beat the 4090 to prove they aren't absolutely pathetic.

TSMC jacked up the prices. even ancient 6nm & 7nm processes !!!!

 

[zandor]

It's not about bragging rights, this is how tech is supposed to go, at least that WAS how it went. Think back 10 years ago, a 2017 mid range card would easily destroy the flagship from 2012.

View attachment 826196

That was how tech progressed, the newest card from 3 generations onward would be much faster than the old flagships. Now here in 2026 going into 2027 the mid range lineups from both camps fail to beat the 4090 until well into 2028, and even then there is no guarantee that they actually will and the 4090 could still be faster than both Nvidia's and AMD's mid range lineups that come out in 2028. That is just sad, but I just think it's even more sad for AMD not being able to clear the 4090 for over a half a decade which is a long time in tech terms. Finally beating a 6 year old flagship isn't exactly bragging rights worthy at all, so rather than have something to beat the 4090 for "bragging rights", they should have something to beat the 4090 to prove they aren't absolutely pathetic.

Yep. The next card down roughly matched the one a notch up from the previous generation. Maybe a little faster or slower, but about there. So 600, 700, 900, 1000... by the time the 1000 series came out the 680 was a "50" card. That's the old way, and it doesn't hold up anymore. 4060 couldn't even beat a 3060Ti, 5080 can't beat a 4090, 5060 can't beat a 4060Ti, and the 5060 still can't reliably beat that pesky 3070.

 

[kram182]

It's not about bragging rights, this is how tech is supposed to go, at least that WAS how it went. Think back 10 years ago, a 2017 mid range card would easily destroy the flagship from 2012.

View attachment 826196

That was how tech progressed, the newest card from 3 generations onward would be much faster than the old flagships. Now here in 2026 going into 2027 the mid range lineups from both camps fail to beat the 4090 until well into 2028, and even then there is no guarantee that they actually will and the 4090 could still be faster than both Nvidia's and AMD's mid range lineups that come out in 2028. That is just sad, but I just think it's even more sad for AMD not being able to clear the 4090 for over a half a decade which is a long time in tech terms. Finally beating a 6 year old flagship isn't exactly bragging rights worthy at all, so rather than have something to beat the 4090 for "bragging rights", they should have something to beat the 4090 to prove they aren't absolutely pathetic.

My RTX 4070 is faster than a Titan RTX

My RTX 5070 goes toe to toe with 3090 (not Ti though)

 

[WilyKit]

My RTX 4070 is faster than a Titan RTX

My RTX 5070 goes toe to toe with 3090 (not Ti though)

I'd go a step further than the person you quoted and say next gen #2 card is often faster than previous gen flagship with the exception of 50 series. My 4070Ti Super is right there with a 3090Ti while pulling roughly half the power. A 3080 is faster than a 2080Ti. A 2080 was faster (albeit barely) than a 1080Ti

 

[MistaSparkul]

I'd go a step further than the person you quoted and say next gen #2 card is often faster than previous gen flagship with the exception of 50 series. My 4070Ti Super is right there with a 3090Ti while pulling roughly half the power. A 3080 is faster than a 2080Ti. A 2080 was faster (albeit barely) than a 1080Ti

View attachment 826269

Yeah and once you go two generations ahead then that's when the 60 series should start passing the flagship from 2 generations prior and we don't see that happening anymore. A 5060 Ti is still falling short of a 3090 let alone a 3090 Ti.

But this wasn't even my main point. My main point is that AMD literally cannot beat a 4090 until 2028 by which then it will be 6+ years old and that is just sad. Imagine if AMD had nothing that could beat the original GTX Titan, a 2013 GPU, until 2019. At least back then they managed to come up with a competitor in less than a year with the R9 290X.

 

[tunatime]

Yeah and once you go two generations ahead then that's when the 60 series should start passing the flagship from 2 generations prior and we don't see that happening anymore. A 5060 Ti is still falling short of a 3090 let alone a 3090 Ti.

View attachment 826291

But this wasn't even my main point. My main point is that AMD literally cannot beat a 4090 until 2028 by which then it will be 6+ years old and that is just sad. Imagine if AMD had nothing that could beat the original GTX Titan, a 2013 GPU, until 2019. At least back then they managed to come up with a competitor in less than a year with the R9 290X.

Amd will never beat Nvidia again unless they somehow get the AI tech Nvidia has in games now. I kinda think though we never get a new Gen and everything goes to subscription based gaming

 

[BravO)))]

Yeah and once you go two generations ahead then that's when the 60 series should start passing the flagship from 2 generations prior and we don't see that happening anymore. A 5060 Ti is still falling short of a 3090 let alone a 3090 Ti.

View attachment 826291

But this wasn't even my main point. My main point is that AMD literally cannot beat a 4090 until 2028 by which then it will be 6+ years old and that is just sad. Imagine if AMD had nothing that could beat the original GTX Titan, a 2013 GPU, until 2019. At least back then they managed to come up with a competitor in less than a year with the R9 290X.

It isn't a matter of AMD matching or doing better than Nvidia in performance, but what is financially viable from what the market decides. The market has shown many times, even if AMD outperforms Nvidia by a small to medium margins, Nvidia still holds the market for reasons beyond being faster. If AMD spent the money on R&D and produced something that could compete at the top end, would it be worth it? Think about the situation we are in now. The 5060 ti gets pounded by the 9070 xt in gaming, but both are around the same price now. Why would anyone pick the 5060 ti over a 9070 xt? How would AMD fair if they tried for top tier at the moment? IMO, AMD would be better off spending the money on ROCM and competing for top tier in another cycle.

 

[MistaSparkul]

It isn't a matter of AMD matching or doing better than Nvidia in performance, but what is financially viable from what the market decides. The market has shown many times, even if AMD outperforms Nvidia by a small to medium margins, Nvidia still holds the market for reasons beyond being faster. If AMD spent the money on R&D and produced something that could compete at the top end, would it be worth it? Think about the situation we are in now. The 5060 ti gets pounded by the 9070 xt in gaming, but both are around the same price now. Why would anyone pick the 5060 ti over a 9070 xt? How would AMD fair if they tried for top tier at the moment? IMO, AMD would be better off spending the money on ROCM and competing for top tier in another cycle.

I actually did buy a 6900XT when it was able to go toe to toe with the 3090, but yeah you're right most people would have probably settled for a 3080 at the same price just due to features than buy AMD. This is probably now more relevant than ever with DLSS5.

 

[kram182]

The thing is people think if AMD beats Nvidia 1 time 1 generation at one tier/card level every 5 years it means people should be buying AMD

AMD needs to consistently beat Nvidia and prove to people they won't lose again the next 3 generations thereafter and so forth before they ever get people to switch

With them not even making HALO cards seemingly anymore that's (even more of a) a tall order

 

[BravO)))]

The thing is people think if AMD beats Nvidia 1 time 1 generation at one tier/card level every 5 years it means people should be buying AMD

AMD needs to consistently beat Nvidia and prove to people they won't lose again the next 3 generations thereafter and so forth before they ever get people to switch

With them not even making HALO cards seemingly anymore that's (even more of a) a tall order

AMD doesn't need to be the best of the best for people to buy them. They just need to have a great value in relation to Nvidia. Regardless, pc gaming is in a weird spot. APUs are gaining market share in gaming, to the point they will be the dominate hardware for most gamers in the future imo. I would expect game makers to develop around this trend, similar to steam deck compatibility on games.

 

[kram182]

AMD doesn't need to be the best of the best for people to buy them. They just need to have a great value in relation to Nvidia.

People have been saying this for years, decades. Yet as you point out when they do beat Nvidia on just 1 card for just 1 generation, or when people like to hold up and point to frame to dollar graphs etc, it still doesn't equal the sales.

So they obviously do need to be the best of the best (consistently, repeatedly - and not just as a fluke or once in a blue moon occurrence) for people to buy them.

 

[BravO)))]

People have been saying this for years, decades. Yet as you point out when they do beat Nvidia on just 1 card for just 1 generation, or when people like to hold up and point to frame to dollar graphs etc, it still doesn't equal the sales.

So they obviously do need to be the best of the best (consistently, repeatedly - and not just as a fluke or once in a blue moon occurrence) for people to buy them.

Most of the time I've bought AMD, it wasn't because they had the best, but were the best I could afford. Regardless, 5060 ti isn't the price of a 9070 xt because it is faster at gaming. It isn't. It has nothing to do with gaming at all. How would AMD benefit from trying to compete in a higher segment, given the current climate?

 

[Marees]

Most of the time I've bought AMD, it wasn't because they had the best, but were the best I could afford. Regardless, 5060 ti isn't the price of a 9070 xt because it is faster at gaming. It isn't. It has nothing to do with gaming at all. How would AMD benefit from trying to compete in a higher segment, given the current climate?

AMD has the right ideas for RDNA 5

• AT0 = xcloud
• AT2 = series x2
• AT4 = series s2
• AT3 = DGX Spark/N2X/mac mini competitor

 

[kram182]

Most of the time I've bought AMD, it wasn't because they had the best, but were the best I could afford. Regardless, 5060 ti isn't the price of a 9070 xt because it is faster at gaming. It isn't. It has nothing to do with gaming at all. How would AMD benefit from trying to compete in a higher segment, given the current climate?

And you alone and all the others weren't enough to give them the sales numbers still

If AMD competed - and succeeded at competing - then people would be buying their cards, for AI as well as gaming, as is the case with Nvidia - but they've continuously failed and flopped on both fronts - explaining the situation in its entirety

And your previous 'APU cope' is the same cope we heard would occur because of the consoles going AMD, over a decade ago, which never manifested into reality - aside from the fact very soon Nvidia is gonna be the iGPU for Intel's upcoming APUs - so AMD will still have to compete -and beat- Nvidia there as well if they want the sales

 

[BravO)))]

And you alone and all the others weren't enough to give them the sales numbers still

If AMD competed - and succeeded at competing - then people would be buying their cards, for AI as well as gaming, as is the case with Nvidia - but they've continuously failed and flopped on both fronts - explaining the situation in its entirety

And your previous 'APU cope' is the same cope we heard would occur because of the consoles going AMD, over a decade ago, which never manifested into reality - aside from the fact very soon Nvidia is gonna be the iGPU for Intel's upcoming APUs - so AMD will still have to compete -and beat- Nvidia there as well if they want the sales

What are you talking about? AMD bought ATI specifically for what they are doing now... and that was 20 years ago... and they are still around. And through those 20 years, AMD has only beaten Nvidia at the top a few times, and even then, Nvidia still dominated the market. What AMD needs to do is focus on ROCM, because if they were going for top tier, they would be up against CUDA. CUDA is currently what is making AMD worth less, but not worthless. I say that with all AMD systems. One has a 7945hx3d with a 9070 xt and the other is a hx 370.

 

[kram182]

What are you talking about?

What part would you like clarification on?

And through those 20 years, AMD has only beaten Nvidia at the top a few times, and even then, Nvidia still dominated the market.

Exactly - like I said - they needed to beat Nvidia consistently, repeatedly - not every once in a blue moon - if they wanted to overtake them/sell more than them - they never did - Radeon has only lost more and more marketshare to Nvidia under AMD's ownership/leadership - ATI alone was more competition to Nvidia than AMD Radeon is/has been

Edit: And even for iGPUs AMD has lost consistently to/been dominated by Intel all this time if you mean 'that' product/market/segment is what AMD bought ATI for

What AMD needs to do is focus on ROCM, because if they were going for top tier, they would be up against CUDA

What do you think ROCM is competing against?

 

[BravO)))]

What part would you like clarification on?



Exactly - like I said - they needed to beat Nvidia consistently, repeatedly - not every once in a blue moon - if they wanted to overtake them/sell more than them - they never did - Radeon has only lost more and more marketshare to Nvidia under AMD's ownership/leadership - ATI alone was more competition to Nvidia than AMD Radeon is/has been

View attachment 827716



What do you think ROCM is competing against?

What you showed argues my point even more lol. CUDA came out in 2006. Notice anything after that point? On that chart, what really gained AMD ground, after losing market share for years, were the things that had nothing to do with gaming. After mining tanked, what happened? AMD lost market share.

 

[kram182]

What you showed argues my point even more lol. CUDA came out in 2006. Notice anything after that point? On that chart, what really gained AMD ground, after loosing market share for years, were the things that had nothing to do with gaming. After mining tanked, what happened? AMD lost market share.

That's not why Nvidia was beating AMD back then - that was all gaming - CUDA wasn't a hit when it first came out - it wasn't even until AlexNet in 2012 that the idea of using GPUs for AI was shown 'successfully'/as a good/performant idea - that's when CUDA took off, if you want to say

Also you posted at the same time I edited this in but:

Edit: And even for iGPUs AMD has lost consistently to/been dominated by Intel all this time if you mean 'that' product/market/segment is what AMD bought ATI for

 

[BravO)))]

That's not why Nvidia was beating AMD back then - that was all gaming - CUDA wasn't a hit when it first came out - it wasn't even until AlexNet in 2012 that the idea of using GPUs for AI was shown 'successfully'/as a good/performant idea - that's when CUDA took off, if you want to say

Also you posted at the same time I edited this in but:

It started with supercomputers, universities, and STEM related work. And hit isn't the word I would use to describe the evolution. The growth for the most part has been steady, as indicated by your chart. Until AMD can replace Nvidia in industry, R&D towards being the fastest in gaming is like winning the special Olympics. If AMD can't apply their fastest cards towards industry, it limits how much R&D can be recouped.