The current Citrix workstations are Threadripper 3970’s and an RTX Titan. That gives enough juice that I can run 5 students off a single workstation don’t want to be buying more Titans with the 3090’s right there but I don’t have the budget to be dropping the A6000’s in these things unless I replace all those workstations with some EPYC’s and blah blah blah. Fingers crossed on the 6900XT suppprt.Sure its a gaming card. They have a line of professional cards. Quatro cards are the workstation cards its the way it is.
Nvidia in their greed ages ago realized there is a Halo type of customer... that I would call super armature 3D artists. This could include some kids in collage type classes... that can justify the spend but 9 times out of 10 will never really "work" in the industry. As well as just general admitted armatures with bank accounts that allow them to buy Quadro lite cards. They have no need for Quatro level cards cause they in general have little clue what they are doing... but they can render monkey heads in blender really good... and ya they are probably talented enough to render themselves some basic youtube title cards ect.
If someone is buying for a Gov agency... perhaps they should actually be looking at the Quatro cards. Of course if they are in Gov.... chances are they are fine just buying the gamer cards if that is what makes sense. I have sold to gov before... and bean counters understand the concept of acceptable substitutions in regards to tenders.
Anyway where probably getting off topic. To pull it back a bit... I get the 3090 market placing price wise may still make sense. Nvidia has done a good job of locking some important software into Cuda. Cuda is a legit advantage for them... and your right of course the 3090 with 24gbs of ram and plenty of cuda horsepower, does have a market place for that armature level pro card consumer. And in regards to tech in gov agencies... I can't see the problem justifying "gaming" hardware for the job. Heck the US gov has done plenty of crazy things over the years including building servers out of playstation 3s.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Where Gaming Begins: Ep. 2 | AMD Radeon™ RX 6000 Series Graphics Cards - 11am CDT 10/28/20
- Thread starter FrgMstr
- Start date
Thanks some how the last few weeks I completely missed that bit of news. I stand corrected on the naming.
Fear of government is not because of buying for a Gov agency, most company in most jurisdiction can get tax credit, R&D credit, no sales taxes when they buy good and services for their company, specially when you buy to use from home buying something with a lot of gaming marketing surrounding it make it more uncomfortable, what if we get audited on this point (an audit can go quite in details sometime and you have to justify what hours you did put as R&D and so on).If someone is buying for a Gov agency... perhaps they should actually be looking at the Quatro cards. Of course if they are in Gov.... chances are they are fine just buying the gamer cards if that is what makes sense. I have sold to gov before... and bean counters understand the concept of acceptable substitutions in regards to tenders.
That one is a gaming card (a strange one, in how many year would it make any sense ?) but not all 3090 are marketed as such, say this:
https://www.gigabyte.com/Graphics-Card/GV-N3090TURBO-24GD#kf
vs this
https://www.gigabyte.com/Graphics-Card/GV-N3090GAMING-OC-24GD#kf
At the moment is there any card that come close to a 3090 in some professional application ? Before the new titan/quadro line up come up, not only the 3090 is much more than a gaming card it is the most powerful pro card in some scenario by a really good amount, no ? Even the 3070 beat the latest Titan in some case.
Hopefully AMD is an option... but don't be shocked if you hear a we require CUDA reply. Hopefully you can sneak "gaming" hardware in. I'm not sure who does your oversight... would it be possible to run the idea by them. Explain the current titans are in fact also gaming cards.
I hear ya purchasing is a pita job. lolI know Citrix has OpenCL options because they list all the AMD workstation class cards as compatible but none of their residential ones. Where in the NVidia stack they support the GTX 980 and up.Hopefully AMD is an option... but don't be shocked if you hear a we require CUDA reply. Hopefully you can sneak "gaming" hardware in. I'm not sure who does your oversight... would it be possible to run the idea by them. Explain the current titans are in fact also gaming cards.
Hopefully. Kyle's recent comment that AMD should have way more launch stock than Ampere is encouraging, but I suspect 6900XT will nevertheless end up $2000+ on ebay.
Last edited:
Well hopefully.... I assume the main factor will being able to run AMDs pro drivers. I'm not an expert with Citrix I hope the people you rely on figure it out. Just so you know though I believe you would probably be golden... although these are new cards so not 100% sure.
https://drivers.amd.com/relnotes/amd-radeon-pro-software-for-enterprise-20.Q2.1.pdf
"Starting in 19.Q2, AMD is including Radeon products in our Enterprise driver. This is being done in consideration of the fact that enterprises use both professional, commercial, and consumer hardware. This driver is recommended for Radeon hardware being used in commercial use cases with productivity applications. This includes users who need to review CAD data as part of their workflow. It is not recommended for gaming. Workstation performance, features, and application certifications do not apply to Radeon hardware and no end user benefit will be observed without Radeon Pro hardware."
They specifically list the 5700 cards as supported by their pro driver. Hopefully Navi 2 will gain support in a reasonably quick fashion. I am not sure I would expect launch day support for sure though. (would be nice though.... who knows perhaps one of the early reviewers that read these forums will read this and do a quick test if they get a review unit to see if the pro driver installs launch day.)
Randall Stephens
[H]ard|Gawd
- Joined
- Mar 3, 2017
- Messages
- 1,820
Too much tech news, so easy to miss stuff that doesn't directly interest you. I know how you feel.Thanks some how the last few weeks I completely missed that bit of news. I stand corrected on the naming.
It's more on the Citrix side if you have a card they don't deem compatible or that they don't officially support it just doesn't even detect it so when you try to assign GPU to the pool it just won't show in the dropdown.
X1900XTX vs 7900GTX I believe... it's been too long that's for sure!
Flogger23m
[H]F Junkie
- Joined
- Jun 19, 2009
- Messages
- 14,373
As with all high ticket electronics, availability will be bad at launch.
when the R9 290x was released vs a more expensive GTX 780 ? October 2013 ?
- Joined
- May 18, 1997
- Messages
- 55,634
Well here is what is effed up about the whole thing, and I am sure there is going to be a "shortage" of cards for AMD. Thing about planning this out months and months ago and getting wafer production lined out in what you project you can sell. You are confident, and you shoot a little high as you feel as though you can take back some of the market. THEN....NVIDIA totally, 100%, shits the bed on supply, when you were expecting them to sell over 300K cards into the market in Q4. Instead they sell less than 50K worldwide. Now there is a huge demand that you expected NVIDIA to fill, and historically, NVIDIA does not fuck up this bad....ever. I think AMD will deliver well over 10X what NVIDIA has shipped, and it still will not be enough to fill the void. Yeah, AMD is going to sell out, not in 5 seconds like Green did, but AMD is going to sell out well before Christmas. I for one would not have wanted to be the one four months ago telling AMD execs to shift 4X the projected allocation to Big Navi delivery, because quite frankly no one could have predicted just how bad NVIDIA was going to short the retail channel. I have been around this market for a long time and I have never seen NVIDIA fuck things up so badly. And after they fucked it up, they tripped over their pile of burning PCBs and set themselves on fire again.
I really do not know what the contributing factor in this is (I have to believe it is GA102 ASIC production), but I firmly believe that Samsung is going to get thrown under the bus come the next NVIDIA earnings call.
I agree who could have seen that coming. In hind sight it sure does make that move to Samsung seem very desperate. I think tis probably hard to fault Samsung a ton... Nvidia is asking them to fab a chip 3-4x larger then anything they have ever done. I am sure Nvidia got caught sleeping when Navi 1 hit. Not that it was a barn burner... but they are smart enough to see that it was a architecture AMD was going to be able to squeeze a ton of performance per watt out of for the consoles. I honestly believe 2 years ago Nvidia may have thought AMD was going to sell some form of GCN+++ to MS and Sony for a Pro+ console rather then a true next gen. By the time they figured out they may have to do something less incremental... TMSCs fabs where spoken for. Samsung was a huge gamble. So far doesn't look like its paid off. Nvidia may have been better letting AMD take a lead for 2-3 months, and hitting back with a proper 7nm TMSC ampere in Q1 or even Q2 if need be.
Hopefully your right and AMD can mange to ship 500k cards. Must be nice to be TMSC right now though.... we might not get a price war right away, but TMSC might be looking at one. lol We have space to run 200k chips by February... and where taking bids. I suspect Nvidia isn't going to get great pricing if the TMSC rumors are true.
ZeroBarrier
Gawd
- Joined
- Mar 19, 2011
- Messages
- 1,011
Ah, memories. The X1900XTX was literally my last AMD card. That was back when we still had to download Omega drivers, because AMD have been shitting the bed driver-wise that long. Blows my mind really.
Last edited:
chameleoneel
Supreme [H]ardness
- Joined
- Aug 15, 2005
- Messages
- 7,602
Had to get omega drivers for Nvidia so that the texture filtering didn't look like vasoline. Something they didn't fix until the 6800 series.
cybereality
[H]F Junkie
- Joined
- Mar 22, 2008
- Messages
- 8,789
Yeah, well no one can predict the future, at least not consistently.
I don't know how much inventory AMD will have, but if their slides were real, it will be sold out no doubt.
And we still don't even know the mining angle (AMD has traditionally been strong there) , add in bots, well, it's going to be a fun time.
I don't know how much inventory AMD will have, but if their slides were real, it will be sold out no doubt.
And we still don't even know the mining angle (AMD has traditionally been strong there) , add in bots, well, it's going to be a fun time.
When AMD gets more details out on the cards through whitepapers and the like it might get interesting. Probably going to be great miners... and them saying they have ray generation hardware on each compute unit. Makes me wonder if someone is going to find a novel way to mine on them. I know its just math... but the Navi 1 had its DCU dual compute cores. I wonder if Navi 2 is still a "DCU" or if there is in fact some third compute path on each DCU. With the right software they could be compute beasts perhaps... or perhaps it simply doesn't work that way. Looking forward to reading... and hearing what smarter people think.
dub77nj
Limp Gawd
- Joined
- Oct 10, 2007
- Messages
- 304
This very stiff twig and berries brought to you with enhanced performance by AMD.
- Joined
- May 18, 1997
- Messages
- 55,634
Well, I was mixing some NA and EU and Worldwide numbers there.....If AMD could put that many in channel we would all have a merry Christmas. That ain't gonna happen. Sorry you get stream of consciousness replies here, not fully edited and checked content like on HardOCP.
Gideon
2[H]4U
- Joined
- Apr 13, 2006
- Messages
- 3,558
I wonder if someone decided to listen to the life long accountant at Nvidia that said, "Hey I know a great way to save money on each gpu and the process is almost as good as TSMC's." Seen that happen before in big companies, saving a buck because it's almost as good.
Holy fookin’ shit.
I follow them closely because I have money in it but even with my bullish views and the large amount we could reasonably infer, I’m actually pretty stunned. If this is real, it’s huge.
I could not agree with Kyle more about how badly Nvidia have fucked up with the supply stuff. Now AMD hit them on price again after nv already preemptively chopped margins massively.
I always saw this gen as a mindshare one for amd. Get relevance and be considered, but they’ve actually swung way harder than I thought, and tbh it feels harder than they needed to be on price. Absolutely crushed the ball.
If the nascent trends to kneecap nvidias mortal lock on GPU compute through the various attempts at open or cloud based standards (I.e not cuda) start getting traction it could actually change the game in a way that I thought couldn’t happen in this generation. Be clear nvidia are not well liked. Amazon and Google are as a big a driving threat as anyone else but there’s a whole industry that wants to take nvidia down a peg or two so they don’t get insurmountable (they’re a bit of a dick to deal with, even aside from no one wanting another trillion dollar gorilla with no competition)
it’s gonna get interesting. I’m almost tempted to increase my holdings and I never thought I’d be thinking that.
Plus, in pure [H] angst terms it’s now super obvious why the 3090 exists and it was still barely enough.
Daaaamn.
Also, I owe Kyle a charity donation now.
I follow them closely because I have money in it but even with my bullish views and the large amount we could reasonably infer, I’m actually pretty stunned. If this is real, it’s huge.
I could not agree with Kyle more about how badly Nvidia have fucked up with the supply stuff. Now AMD hit them on price again after nv already preemptively chopped margins massively.
I always saw this gen as a mindshare one for amd. Get relevance and be considered, but they’ve actually swung way harder than I thought, and tbh it feels harder than they needed to be on price. Absolutely crushed the ball.
If the nascent trends to kneecap nvidias mortal lock on GPU compute through the various attempts at open or cloud based standards (I.e not cuda) start getting traction it could actually change the game in a way that I thought couldn’t happen in this generation. Be clear nvidia are not well liked. Amazon and Google are as a big a driving threat as anyone else but there’s a whole industry that wants to take nvidia down a peg or two so they don’t get insurmountable (they’re a bit of a dick to deal with, even aside from no one wanting another trillion dollar gorilla with no competition)
it’s gonna get interesting. I’m almost tempted to increase my holdings and I never thought I’d be thinking that.
Plus, in pure [H] angst terms it’s now super obvious why the 3090 exists and it was still barely enough.
Daaaamn.
Also, I owe Kyle a charity donation now.
Last edited:
Probably not a completely price based choice.... seems like they waited to long to contract TMSC. And their dance card got filled.
However you could be on to something. They may have been ready to go TMSC long ago (early rumors where a 600mm or so 7nm ampere)... they probably could have contracted TMSC nice and early. However Nvidia being Nvidia they likely held out for a better price... and Huawei had their situation pop up and said Yes TMSC sir we will pay whatever you want but start fabing yesterday. lol I get the feeling Samsung was the best option as the alternative may have been to wait for TMSC space to free up and let AMD beat them to market by a quarter. Hey in their defense I can respect that they took the swing... if Samsung had nailed it and they where overflowing with parts they could have looked like Geniuses too.
Deep in the tinfoil hat here but hear me out. NVidia knew that their cards weren’t going to beat out AMD and they knew they would need a Super/Ti card sooner than planned. So they shorted their initial stock knowing that beyond the initial wave they weren’t going to do well against AMD. But if AMD sells out fast before Christmas it’s not like they can just make more, they have their Threadripper Pro and Ryzen Pro’s they need to start delivering soon, plus the next wave of consoles for after Christmas, then they have the newest EPYC’s and Threadripper’s after that. Then they have the mid/low range Radeons to deliver on which brings them firmly into April. So what if come November when AMD starts putting them on the shelves and they sell out PDQ NVidia gets to ride in on their white horse in their leather bound glory and announce a large supply of 3070 TI & 3080 TI cards with supply to match. Not too many early adopters are burnt because they haven’t actually gotten that many cards out the door, the Scalpers are left holding the bag for product they have a harder time moving, and people waiting for either card now have a new third option.
- Joined
- May 18, 1997
- Messages
- 55,634
Nope.
Succinct.
It’s Occam’s razor with this shit. If you analyse whether some company had a really complex plan to try and bamboozle their competition, or they just fucked up. One of those is vanishingly less likely to be true.
Large corporations are not that smart. Remember that shit is corporate strategy and marketing. An mba ain’t that hard. You get well paid but nowhere close to the smartest folks. You might find some very expensive advice from a McKinsey or a similar got them trying to play too cute, but it seems unlikely.
I’m nowhere close to the sector and people that know, but the rumour that nvidia tried to play negotiation hard ball and fucked it, sounds very plausible in my experience. If they then responded by going a bit left field and that’s not gone as well as they expected, well it just compounds it.
Yay, competition in gpu’s and cpus. About damned time.
Large corporations are not that smart. Remember that shit is corporate strategy and marketing. An mba ain’t that hard. You get well paid but nowhere close to the smartest folks. You might find some very expensive advice from a McKinsey or a similar got them trying to play too cute, but it seems unlikely.
I’m nowhere close to the sector and people that know, but the rumour that nvidia tried to play negotiation hard ball and fucked it, sounds very plausible in my experience. If they then responded by going a bit left field and that’s not gone as well as they expected, well it just compounds it.
Yay, competition in gpu’s and cpus. About damned time.
Last edited:
I think this is going to be the irony of situation this round. It's obvious nVidia completely shit the bed with the release and no real stock. It would literally be impossible for AMD or any other company to be able to make up for the massive vacuum left by nVidia. No company tries to allocate more than twice the stock they need to fill initial orders because 9,999 times out of 10,000 the result is massive overstock. There is no way AMD could have known that there would be a literal need to fill a void this big much less known far enough in advance to take measures to fully capitalize on it.
Future "complaints" about AMD having no stock after selling out should be fun. Even better will be when the estimated numbers of card sold gets compared.
I suspect the nVidia release was rushed way ahead knowing what AMD had. What I'm curious about is how much the bungled release is going to hurt nVidia and how much could have been avoided had nVidia let AMD release basically unopposed.
There is no way nVidia pulled back production or stock. The company needs the revenue and profit from as many card sales as possible no matter how few they can provide. Cutting back production to a small fraction of what you need and what your opponent has is giving the market leader position to the competitor. It would hurt nVidia financially (unless every single chip was sold at a loss) but the loss of marketshare and mindshare would be devastating. It would literally be sabotaging your own company in order to make it fail.
nVidia is trying to get as much stock out as possible because it's the only decision which makes sense. Any Super or Ti or whatever model comes out is going to take time. Respin of the silicon is going to be required. The power needed to get the performance currently is bad enough so throwing more juice at the cards to overclock them isn't going to work. If a different fab process is necessary that's going to take even longer. During this time nVidia must still supply cards. Low supply is the bare minimum simply to keep operating.
No matter how bad or where the problems are with the 3xxx series nVidia can still sell every one of them made right now. If more 3xxx cards could get to market, nVidia would be supplying them in a heartbeat.
WorldExclusive
[H]F Junkie
- Joined
- Apr 26, 2009
- Messages
- 11,548
Let's take a look at history. In the beginning of 2010, GTX 480 was fast but power hungry. ATI won that round with cards just as fast that ran cooler with less power.
Nvidia decided to dump the GTX 480 series altogether only after a few months and released the 580 by the end of the year.
Don't put it past Nvidia to jump to a new generation 2021. They've respun their cards in relaunched them within the same year.
Like you said, Super and Ti cards may never come because of the power problem, but a respun RTX 4080 isn't too far fetched.
Nvidia said shortages into 2021. In other words, they are working on something different.
RDNA3 is scheduled for 2022 so they can't ride 3080s until then because Nv desires a clear win over AMD. Trading blows isn't good for mindshare.
I’m not sure it would be literally impossible. Given the console generation refresh and alignment, plus their capacity, AMD could
have the awful problem of which high margin product do I dedicate my enormous capacity towards. (Capacity for low margin console chips being contractually watertight)
I think it’s unlikely of course that they change the game in 2 years, just because mindshare is really hard to get. However nvidias ability to pivot is a bit limited without throwing the baby out.
If it’s all true, I can see AMD just going “ooo a window” and just throwing huge amounts of money at it, I don’t know what tech PR costs but if they can follow through and get stock through, spending $200m on a PR avalanche to get mindshare wouldn’t be the stupidest thing they could do. Think Radeon in every fucking tangential tenuous place you can think of. This I have experience of in mobile. It’s how it works. It works.
For the first time in maybe ever, they’re in a position to do that. Working in AMD treasury and wanting to borrow a few hundred million in working capital would be the easiest fucking job going at the moment.
have the awful problem of which high margin product do I dedicate my enormous capacity towards. (Capacity for low margin console chips being contractually watertight)
I think it’s unlikely of course that they change the game in 2 years, just because mindshare is really hard to get. However nvidias ability to pivot is a bit limited without throwing the baby out.
If it’s all true, I can see AMD just going “ooo a window” and just throwing huge amounts of money at it, I don’t know what tech PR costs but if they can follow through and get stock through, spending $200m on a PR avalanche to get mindshare wouldn’t be the stupidest thing they could do. Think Radeon in every fucking tangential tenuous place you can think of. This I have experience of in mobile. It’s how it works. It works.
For the first time in maybe ever, they’re in a position to do that. Working in AMD treasury and wanting to borrow a few hundred million in working capital would be the easiest fucking job going at the moment.
Do we have Navi 21 die size yet? The die that was showed off in the presentation looked quite chunky. It’s far from obvious that AMD is enjoying great margins on the 6x00 cards. CPU margins are likely much healthier.
That would cost a lot more than $200m when you include marketing folks salaries and operational costs. It’s also unnecessary. The discrete GPU market is primarily driven by OEMs selling prebuilts not by desperate people lining up outside Microcenter.
If AMD can get in good with OEMs and provide sufficient supply then they can do real damage to market share. Otherwise it will be the same story we’ve seen over and over. Everyone gets excited that the underdog is back in the fight, reviews look good and AMDs core fanbase is smiling again but market share goes nowhere.
nVidia is not going to come up with a completely new architecture within a few months or even a year. There's not enough time for that. That's the reason for the respin. A respin is tweaking the current architecture in order to gain performance, lower power, decrease manufacturing costs, etc. That is exactly what nVidia will need to do in order to push past the issues with the current 3xxx line.
This is the same problem AMD had with the release of the RDNA. We never saw anything bigger than the 5700XT because of inherent issues which needed to be worked out. RDNA2 is more a respin on steroids rather than a new architecture. It took time for AMD to do all this and it was planned years in advance. These chips are way too big and complex to change on the fly. I have no doubt nVidia had plans for a respin for the 3xxx series but it's no matter how much of that work is done, it can't be released tomorrow. And this is just for a respin.
There are zero indications nVidia could release a new architecture within the next year even if the company wanted to. The typical nVidia cycles are around 24 months and it's been just over 24 months since the 2xxx series released. nVidia may have had a respin planned to release in another 6-9 months such as the Ti or Super models but the likelihood of anything showing up before that is slim with no chance of a new architecture. Current and future contracts and planning for components, fab space and expected yields means fast turnaround is not possible.
nVidia has a shortage problem and will for the foreseeable future outside of a miracle. That doesn't indicate at all that nVidia is working on something different. That nVidia is working on something different is correct because it's always working on something different. It doesn't mean that something different is going to be in stores anytime soon.
It doesn't really matter when RDNA3 is coming out and we don't know how RDNA3 will turn out anyway with regards to nVidia. However, nVidia has no choice but the ride the current 3xxx series out. It doesn't matter if nVidia likes being outperformed or outsold. nVidia can't make Jensen's desires reality by saying a word. This is called competition.
A paraphrase from a book series I enjoy. "The Law of Battle: Everything gets fucked up as soon as the enemy arrives. That's why he's called the enemy." In this case AMD arrived with RDNA2 and fucked up all of nVidia's plans. Still, it wouldn't have mattered if nVidia had stock. The 3xxx series and architecture isn't bad, it's simply not available.
- Joined
- May 18, 1997
- Messages
- 55,634
Ecosystem. So you think AMD is going to push A+A when it can't fulfill those needs? Or maybe it will just pull an NVIDIA? AMD is not against the wall, NVIDIA and Intel are.
Say it with me, "execution." Something that AMD has been on point on. Can't say that for NVIDIA or Intel this last gen.
I would think it goes without saying that in order to gain market share you need to produce and sell more product. I was replying to the post above regarding a marketing push - clearly that only helps with demand not supply.
All good points.
What i would say you don’t pay marketing people’s salaries though for a campaign like that. PR is almost always contracted (except if you’re Apple or a movie studio)
You just pick up the phone to companies like WPP and say I’d like to spend $200m, give me a proposal. Then you give them the money and they go do your ad buys, line up the campaigns, start spamming the shit out of media outlets with releases, getting coverage in magazine shows etc etc etc
So I don’t know how much it costs for this type of stuff and pulled a number out of air but it’s not AdWords and I do know how much it costs to get your product on a number one rated tv show or an upcoming movie like (well actually) a Mission Impossible. Not nearly as much as youd think.
I’m pretty sure you could get a whole series of content from some of the tech tubers not called Linus or Jay for the budget equivalent of a couple of t-shirts and a hand on the leg from a pretty girl.
Get people aware, get people to know you’re back, get stuff in people’s hands. That’s your momentum. They did it with Ryzen remember.
What i would say you don’t pay marketing people’s salaries though for a campaign like that. PR is almost always contracted (except if you’re Apple or a movie studio)
You just pick up the phone to companies like WPP and say I’d like to spend $200m, give me a proposal. Then you give them the money and they go do your ad buys, line up the campaigns, start spamming the shit out of media outlets with releases, getting coverage in magazine shows etc etc etc
So I don’t know how much it costs for this type of stuff and pulled a number out of air but it’s not AdWords and I do know how much it costs to get your product on a number one rated tv show or an upcoming movie like (well actually) a Mission Impossible. Not nearly as much as youd think.
I’m pretty sure you could get a whole series of content from some of the tech tubers not called Linus or Jay for the budget equivalent of a couple of t-shirts and a hand on the leg from a pretty girl.
Get people aware, get people to know you’re back, get stuff in people’s hands. That’s your momentum. They did it with Ryzen remember.
- Joined
- May 18, 1997
- Messages
- 55,634
The only problem with demand that AMD is going to have to deal with is the gaping hole that NVIDIA has left in the market by not executing.
I agree that's the case in terms of status quo where AMD only supplies 20% of the market.
- Joined
- May 18, 1997
- Messages
- 55,634
Ah! Holy shit! So now you finally understand TAM and how market projections work? Kudos.
I think you're arguing with the boogeyman. And yes I know quite a bit about it thanks.
