nVidia GeForce Partner Program: Well Intention Marketing or Anti-Competitive

[Razor01]

24 minutes ago, hobobobo said:

Aint gonna argue with RTG marketing being shit, it is. In fairness, the problem with vegas geometry troughput comes from AMD not delivering on some of the promised features, who knows why they failed. Marketing probably got overzealous hearing all the promised feautures and their implications and shot themselves in the foot

 

 

Those promised features though, are based on special programming needs.  I think that is why they decided to drop it.  We had conflicting reports from two different AMD employees on that too.  One said they were automatic another one said sometimes they will work sometimes they won't lol.  From my point of view, to expose things in an API the features must be there, but if exposed with different extensions, programming must be done.  I really don't think something was broken in Vega. 

 

This is also because the way chips are designed, and how drivers are done prior to the chips going to risk production.  Drivers are already almost complete by the time Tape out is done.  This is why tape out occurs, by that time the designer knows how the chip will function.

 

Everything is emulated on other hardware prior to going to mass production, to emulate the hardware, the drivers must be done or close to it.

[hobobobo]

1 minute ago, Razor01 said:

Those promised features though, are based on special programming needs.  I think that is why they decided to drop it.  We had conflicting reports from two different AMD employees on that too.  One said they were automatic another one said sometimes they will work sometimes they won't lol.  From my point of view, to expose things in an API the features must be there, but if exposed with different instructions, programming must be done.

From what ive heard, they postponed it till Navi, which doesnt make much sense, since navi is supposed to be the last GCN and vega was supposed to serve as poc for those features.

 

The corporate and marketing side of this business seems so much fun, almost makes me sad of not having such opportunities in Russia.

[Notional]

33 minutes ago, Razor01 said:

But there is a graphics fidelity difference between to the two and that is what causes the performance differences. 

Yes, TressFX is vastly better in graphical fidelity, performance and features:

And that still runs better than any HairWorks implementation.

 

No one is arguing NVidias rights. We are arguing their business tactics on an ethical level. Specifically from a consumer point.

 

All markets are regulated. Anarki would be the alternative. And good riddance.

 

Indeed. But NVidia is desperate. With the new Kaby Lake G, Intel and AMD + AMD's Ryzen APU's, have basically destroyed most of NVidia's laptop market space. And this will only get worse. The same thing could happen in the desktop space. Especially if AMD makes a "ryzen comeback GPU". It's clever that NVidia is focusing on AI, autonomous vehicles, etc. They know their consumer GPU space is shrinking by the day. They are holding on with scummy shit like GameWorks and this GPP bs.

[Razor01]

 

Her hair looks like dried grass, I haven't been impressed implementation of Tress Fx yet, but the wolves in Witcher 3, loved their fur, it flowed like it would in real life. but even that if the camara got too close, you can say it looked like grass cause you can see the polygons easily delinted.

 

Are we back to assumptions again?  I'm going to assume Kaby Lake G is going against Pascal, then yes you are correct, If I'm going to assume Kaby lake G is going up against next gen GPU from nV then no.

 

Keep this in mind we have a good idea where Kaby lake G will end up in performance and power usage, it ends up around a 1050 -1060 but depending on the variant its power usage is higher than that of an equivalent separate CPU (either AMD or Intel) and GPU from nV currently.

[Notional]

4 minutes ago, Razor01 said:

 

Her hair looks like dried grass, I haven't been impressed implementation of Tress Fx yet, but the wolves in Witcher 3, loved their fur, it flowed like it would in real life. but even that if the camara got too close, you can say it looked like grass cause you can see the polygons easily delinted.

 

Are we back to assumptions again?  I'm going to assume Kaby Lake G is going against Pascal, then yes you are correct, If I'm going to assume Kaby lake G is going up against next gen GPU from nV then no.

 

Keep this in mind we have a good idea where Kaby lake G will end up in performance and power usage, it ends up around a 1050 -1060 but depending on the variant its power usage is higher than that of an equivalent separate CPU (either AMD or Intel) and GPU from nV currently.

You can't really compare fur with hair on the head. It will look very different. But dried grass? Really? Now you're just being silly. Also show me HairWorks that gets wet or gets snow and grime on it. Doesn't exist.

 

Right now, laptop vendors have access to Intel mobile CPU's with iGPU, Kaby Lake G for great gaming performance and Ryzen mobile APU's being between the two former in performance. The only space NVidia has left is highend mGPU's. And they just aren't as widely common as a lot in here thinks.

[Razor01]

18 minutes ago, hobobobo said:

From what ive heard, they postponed it till Navi, which doesnt make much sense, since navi is supposed to be the last GCN and vega was supposed to serve as poc for those features.

 

The corporate and marketing side of this business seems so much fun, almost makes me sad of not having such opportunities in Russia.

 

 

Interestingly Russia has some of the best graphics artists in the world lol.

 

All I can say is looking at where this problem started from, their first DX11 chip, and since they haven't changed much in their fixed function pipeline.  I think it was a trade off of where to use transistors. 

 

I hope to god Navi is not going to be GCN based, if it is, its going to have the same damn problems AMD is having now lol.

 

The reason why we see nV gaining market share is because their products have been better, but why have their products been better.  They have been slowing adopting and making those features better than AMD's counterparts.  If we go back to pre DX11 hardware and DX10 even, AMD had better geometry through put than nV hardware, but a lot.  They also had better shader through put pre (DX10).   The x19xx, x18xx series from ATi, was much better at pixel shader and dynamic branching performance than their gf counter parts.  G80 changed all that.  R600 was still better than g80 when it came to triangles.  But it had weakness in other areas which over shadowed this.  With nV sitting fairly still through out that time till Fermi.  AMD was able to catch up, but only because they stuck with VLIW architecture which is much better for graphics than the SIMD.  So they were able to get a similar performance chip as they advanced their VLIW architecture with a smaller chip vs a general all purpose compute and graphics solution.  Fermi had the same problems too against AMD.  But as compute started to pick up in the gaming world, AMD needed to shift to something like nV's architecture, they needed better compute performance.  Hence GCN. Now GCN has been AMD's go to architecture for how many generations?  5 gens going to be 6?  2 generations was the limit at ATI for new architectures.  That's where the problem lies, they must have a new architecture that can accommodate the changes needed to be competitive again.

[Razor01]

Just now, Notional said:

You can't really compare fur with hair on the head. It will look very different. But dried grass? Really? Now you're just being silly. Also show me HairWorks that gets wet or gets snow and grime on it. Doesn't exist.

 

Right now, laptop vendors have access to Intel mobile CPU's with iGPU, Kaby Lake G for great gaming performance and Ryzen mobile APU's being between the two former in performance. The only space NVidia has left is highend mGPU's. And they just aren't as widely common as a lot in here thinks.

 

 

I haven't seen a single Kaby Lake G product on the selves yet man.

[Notional]

1 minute ago, Razor01 said:

I haven't seen a single Kaby Lake G product on the selves yet man.

 

https://www.pcper.com/news/Mobile/CES-2018-HP-Spectre-x360-15-Updated-Kaby-Lake-G-8th-Gen-Intel-Core-i7-CPUs-Paired-Radeon

[Razor01]

13 minutes ago, Notional said:

https://www.pcper.com/news/Mobile/CES-2018-HP-Spectre-x360-15-Updated-Kaby-Lake-G-8th-Gen-Intel-Core-i7-CPUs-Paired-Radeon

 

 

That price doesn't compare to a notebook with a 1060 lol. Most 1060 laptops are going for 1200 bucks.  The lowest one in that group, which will be the slowest Kaby Lake G is considerably higher. 

 

here is a max q 1070 laptop,

 

https://www.newegg.com/Product/Product.aspx?Item=9SIAATJ6140564&cm_re=max_q-_-2WC-0019-000J1-_-Product

 

1800 bucks,

 

https://www.amazon.com/dp/B07665GKSX/ref=sspa_dk_detail_0?psc=1&pd_rd_i=B07665GKSX&pd_rd_wg=kdqrZ&pd_rd_r=G74GA1FHNZZ50A4DGKB1&pd_rd_w=0x24O

 

1850

 

These are 1070's max q's man.  They are only 150 bucks more than  top end  Kaby Lake G. 

 

The 1060 max q's are going to be costing around the lowest performance Kaby Lake G's.

 

Pricing is a major problem for kaby lake G parts, that HBM 2 memory price rearing its ugly head again.

[hobobobo]

9 minutes ago, Razor01 said:

 

 

Interestingly Russia has some of the best graphics artists in the world lol.

 

All I can say is looking at where this problem started from, their first DX11 chip, and since they haven't changed much in their fixed function pipeline.  I think it was a trade off of where to use transistors. 

 

I hope to god Navi is not going to be GCN based, if it is, its going to have the same damn problems AMD is having now lol.

 

The reason why we see nV gaining market share is because their products have been better, but why have their products been better.  They have been slowing adopting and making those features better than AMD's counterparts.  If we go back to pre DX11 hardware and DX10 even, AMD had better geometry through put than nV hardware, but a lot.  They also had better shader through put pre (DX10).   The x19xx, x18xx series from ATi, was much better at pixel shader and dynamic branching performance than their gf counter parts.  G80 changed all that.  R600 was still better than g80 when it came to triangles.  But it had weakness in other areas which over shadowed this.  With nV sitting fairly still through out that time till Fermi.  AMD was able to catch up, but only because they stuck with VLIW architecture which is much better for graphics than the SIMD.  So they were able to get a similar performance chip as they advanced their VLIW architecture with a smaller chip vs a general all purpose compute and graphics solution.  Fermi had the same problems too against AMD.  But as compute started to pick up in the gaming world, AMD needed to shift to something like nV's architecture, they needed better compute performance.  Hence GCN. Now GCN has been AMD's go to architecture for how many generations?  5 gens going to be 6?  2 generations was the limit at ATI for new architectures.  That's where the problem lies, they must have a new architecture that can accommodate the changes needed to be competitive again.

Not only graphic designers, IT specialists as well) The problem is, im none of those, and quite frankly, dont want to be. Ive studied economics and became dissillusioned with it. Maybe, one day, ill pick some engeneering and be able to work in some tech design, but for now i dont see a feasable way to join something like AMD/QCOM/Samsung.

 

Navi is supposed to be the last of gcn, work as a proof of concept for mcm gpu and escape some of GCN limitations via mcm. Dunno how they gonna break 8 rops limitation or others, but in theory it should be quite good, since GCN is exceptional at a lower number of CUs, just sucks at scaling up.

 

New architecture is reportedly in pipelines for 2020, b4 that theres was going to be vega node shrink, now probably cancelled, next is navi on 7nm (if gloflo manages that) and after that their next gen architecture on 7nm too. Considering the financial state of AMD in the past 5 years its a damn engineering miracle they managed to make GCN last this long without it completly falling to the sidelines. I mean, beyond 1080ti and Titans tiers, its quite competative. nVidia just managed to win mindshare, with far superior marketing, long before amd fell into deep troubles.

[Razor01]

3 minutes ago, hobobobo said:

Not only graphic designers, IT specialists as well) The problem is, im none of those, and quite frankly, dont want to be. Ive studied economics and became dissillusioned with it. Maybe, one day, ill pick some engeneering and be able to work in some tech design, but for now i dont see a feasable way to join something like AMD/QCOM/Samsung.

 

Navi is supposed to be the last of gcn, work as a proof of concept for mcm gpu and escape some of GCN limitations via mcm. Dunno how they gonna break 8 rops limitation or others, but in theory it should be quite good, since GCN is exceptional at a lower number of CUs, just sucks at scaling up.

 

New architecture is reportedly in pipelines for 2020, b4 that theres was going to be vega node shrink, now probably cancelled, next is navi on 7nm (if gloflo manages that) and after that their next gen architecture on 7nm too. Considering the financial state of AMD in the past 5 years its a damn engineering miracle they managed to make GCN last this long without it completly falling to the sidelines. I mean, beyond 1080ti and Titans tiers, its quite competative. nVidia just managed to win mindshare, with far superior marketing, long before amd fell into deep troubles.

 

 

well AMD's architecture ROPS are decoupled from the memory banks, so they shouldn't have a problem with increasing ROP's if they had the space in silicon to do so.  nV on the other hand they haven't decoupled them yet.  So this is why we see ROP sizes increasing for them as more blocks are added on.

 

Vega is competitive in the 1070ti bracket, but that is because the architecture was made for a certain voltage and certain frequencies, once those are pushed up higher Vega can't compete.  This comes from base transistor layouts.  nV took their time with Maxwell and improved it further with Pascal.  There was an interview done by Hexus, where nV engineers talk about this. 

 

Yeah as you stated this disparity didn't come about in recent years.  Its been a long time coming.  Everything started from the G80 and CUDA, we are going to see this even more in DL technologies too, where Intel and AMD have no chance in these markets in the short or mid term.  But that was the first GPU AMD went up against and the r600 failed miserably against its competition.

[leadeater]

40 minutes ago, Razor01 said:

Yeah as you stated this disparity didn't come about in recent years.  Its been a long time coming.  Everything started from the G80 and CUDA, we are going to see this even more in DL technologies too, where Intel and AMD have no chance in these markets in the short or mid term.  But that was the first GPU AMD went up against and the r600 failed miserably against its competition.

Even back in the ATI days with the 9x00, x00 and 1x00 GPUs Nvidia still far out sold them or had done for so long the short period where ATI sold better Nvidia still had more GPU share installed in computers. Hasn't really been that close long enough to really gain real market share.

 

 

If 1x00 had been slightly better and GeForce 8 been slightly worse the landscape could be quite different, who knows.

 

Nvidia's always had the bigger market presence.

[MegaTechPC]

7 hours ago, Razor01 said:

 

 

That is only if this is really what the contract is about.  If it wasn't the case and that's all nV wants is separate brands?

 

What is the recourse on that.  [H] is slime?  Or should it be taken a bit further, and actually punish the people for giving [H] the piece to begin with?  [H] is culpable yeah I would agree, they are of age people lol, they can think for themselves.  So are the people that got this story to them.

 

People here are saying nV's ethics are bad and should be stopped correct?  What about ethics of the others involved?

 

We can't have laws govern some of these things at all.  Feelings, marketing, all this stuff is ephemeral. 

 

If nV is telling other sites that they just want two seperate brands one for their products and one for their competitions, but we have Kyle stating contrary to that, who are we to believe?  The truth will come out soon enough.  We know Asus signed up, MSI too.  just have to keep our eyes open right?

That's fine if Nvidia wants to control how its cards are marketed by Asus (for example) when talking about the creation of a new sub-brand.  What is not OK is for Nvidia to use the leverage of their current position in the market to force Asus to make ROG an Nvidia-only sub brand when it was never Nvidia alone that helped to develop the popularity and reputation of ROG over the last decade.  This is simple stuff.

 

Oh and by the way, arguing that AMD ostensibly "deserves" to suffer from inherently anti-competitive Nvidia policy because their cards aren't as fast as Nvidia's right now is a really terrible argument that supports monopolistic behavior.  Those of us who aren't Nvidia fanboys actually want competition in the GPU sector and AMD is really the only company out there that can challenge an Nvidia monopoly in discreet graphics.

[Razor01]

1 hour ago, MegaTechPC said:

That's fine if Nvidia wants to control how its cards are marketed by Asus (for example) when talking about the creation of a new sub-brand.  What is not OK is for Nvidia to use the leverage of their current position in the market to force Asus to make ROG an Nvidia-only sub brand when it was never Nvidia alone that helped to develop the popularity and reputation of ROG over the last decade.  This is simple stuff.

 

Oh and by the way, arguing that AMD ostensibly "deserves" to suffer from inherently anti-competitive Nvidia policy because their cards aren't as fast as Nvidia's right now is a really terrible argument that supports monopolistic behavior.  Those of us who aren't Nvidia fanboys actually want competition in the GPU sector and AMD is really the only company out there that can challenge an Nvidia monopoly in discreet graphics.

 

 

Look even if nV wants the ROG brand to be selling only their brand and AMD's brand to be something else, that is ok too.  It all depends on if their partners want to do that.  Its up to the partners.  ROG is nothing without the products that it has, which the first one was a nforce 4 SLI motherboard, for the first 2 years of ROG's existence it was nforce that gave them the ROG brand life lol.  Now if ROG becomes only nV graphics cards and other brands for other things I see that fine too.  So ROG can still have AMD motherboards, Intel motherboards, just nothing to do with GPU's outside of nV's.

 

AMD doesn't deserve our charity either or our sympathy for being second fiddle to a company that has shown time and time again why they keep coming out on top.  AMD did it to themselves.  When ATi had the money, had the markets share, they came out with a bomb, which AMD paid for ATi at a crazy ass price well above what ATi was worth.  keep this in mind nV was worth 5 billion or so as a comparison at that time, ATi was worth around 3 to 3.5 billion maybe a bit less if I'm not mistaken.  Yet AMD purchased ATi and 5.4 billion stock and cash on the premise there were up coming technologies that would catapult them to number 1 in both CPU and GPU.  Looked like a sound investment, but the truth of the matter was there was goose egg coming down the pipeline that ATi was fully aware of.  ATi screwed AMD over, and AMD screwed itself over, because of that buy out and the r600 AMD was forced to cut down R&D for CPU's which took over a decade to finally come back into the race with Intel, and that is because Intel sat on its ass for the past 5 generations with chip design.  Then in the crisis of this situation to keep afloat they spun off their fabs, which handcuffed them to GF's, they might get a good price for their wafers but if the node GF's is working on doesn't match up well with competitors nodes, then AMD is tied to a sinking ship.  To get out of that one they had to make a cash payment to give them access to purchase wafers outside of the GF, which by the by they have to pay GF a per wafer charge lol.   All the while, because of nV's lukewarm approach to consoles, they wanted more money per chip and MS and Sony pretty much said no, AMD took that low margin business and called it a day.  They sold their products for much less margin then they should be sold at.  Great its cash in hand but, next time they go up to MS and Sony, they don't have anything to leverage from outside of compatibility.  No one likes to make less then a product is truly worth.  AMD is not an AIB where they get kick backs and pass through.  AMD's overhead for R&D and manufacturing for future tech changes up word fast those low margin products can't sustain them, not only that they will only drag them down as their other products improve in margins.

 

To get back to even keel if we want to call it that, on the CPU side, they had to decimate their GPU side.

 

Back to the time ATI had the marketshare and the money to compete with nV, they should have also pushed their game programs, just like nV did when they released a bomb of the FX, the FX came out right before TWIMTBP program came out.  nV pushed that program hard because they knew it didn't matter if they marketed the FX series much or not, they just weren't going to be able to stop ATi from taking marketshare.  Did we hear about nV talking about the FX, yeah ever now and then but when it came down to it.  Not much, there were whimpers from a far distant corner.

 

You can't make this shit up, its all there, they f'ed themselves over time and time again.

 

 

[leadeater]

2 hours ago, Razor01 said:

Look even if nV wants the ROG brand to be selling only their brand and AMD's brand to be something else, that is ok too.  It all depends on if their partners want to do that.  Its up to the partners.

But it's not a choice partners can make though, it's the terms of the GPP so it's a contractual obligation. Before you bring up that GPP is optional, no it really isn't, not if you actually want to be competitive and actually stay in the GPU AIB market for Nvidia cards. No matter if it's all of ROG or just Strix or what ever it's certainly not a choice, it's about as much choice as you have to breath air or drink water. You can go without but not for long.

 

Forget all the back story and history to these companies, the products they make, who's is better it's all irrelevant because this is 100% a contractual issue.

[LAwLz]

3 hours ago, leadeater said:

But it's not a choice partners can make though, it's the terms of the GPP so it's a contractual obligation. Before you bring up that GPP is optional, no it really isn't, not if you actually want to be competitive and actually stay in the GPU AIB market for Nvidia cards. No matter if it's all of ROG or just Strix or what ever it's certainly not a choice, it's about as much choice as you have to breath air or drink water. You can go without but not for long.

So do I understand things correctly...

In order to be part of the GPP, manufacturers has to sell Nvidia products under their own brand. Correct?

For example if Nvidia makes a ROG card they now have to choose if they want the entire ROG brand to be Nvidia, or AMD. They can't make both ROG Nvidia and ROG AMD products.

Correct?

 

In that case I think this is a pretty bad move, but not anywhere near as bad as the original article presented it (which I interpreted as, AMD cards can no longer be marketed as gaming products).

 

 

But with that being said, I don't think having separate brands is a big deal. There is already a market precedent that manufacturers are allowed to force others to market and brand their products in a certain way, for example Apple's "store-within-a-store" concept deployed in stores like BestBuy. I am fairly sure Apple prohibits non-approved brands within the Apple section in those stores.

Anyway, I think it really is up to the OEMs on this. They are in charge of which brand gets assigned to which GPU manufacturer. I think the main worry people have is that recognizable brands like ROG will become exclusive to one GPU brand, but who gets the ROG label is up to Asus to decide. They get to choose if ROG becomes AMD or Nvidia, and if Federation of Gamers (or whatever it gets to be called) gets the other one.

It might be:

ROG = AMD

FOG = Nvidia

That is completely in the hands of Asus, unlike joining the GPP (which is not really a choice if they want to stay competitive, they have to join).

[leadeater]

3 hours ago, LAwLz said:

So do I understand things correctly...

In order to be part of the GPP, manufacturers has to sell Nvidia products under their own brand. Correct?

For example if Nvidia makes a ROG card they now have to choose if they want the entire ROG brand to be Nvidia, or AMD. They can't make both ROG Nvidia and ROG AMD products.

Correct?

 

In that case I think this is a pretty bad move, but not anywhere near as bad as the original article presented it (which I interpreted as, AMD cards can no longer be marketed as gaming products).

 

 

But with that being said, I don't think having separate brands is a big deal. There is already a market precedent that manufacturers are allowed to force others to market and brand their products in a certain way, for example Apple's "store-within-a-store" concept deployed in stores like BestBuy. I am fairly sure Apple prohibits non-approved brands within the Apple section in those stores.

Anyway, I think it really is up to the OEMs on this. They are in charge of which brand gets assigned to which GPU manufacturer. I think the main worry people have is that recognizable brands like ROG will become exclusive to one GPU brand, but who gets the ROG label is up to Asus to decide. They get to choose if ROG becomes AMD or Nvidia, and if Federation of Gamers (or whatever it gets to be called) gets the other one.

It might be:

ROG = AMD

FOG = Nvidia

That is completely in the hands of Asus, unlike joining the GPP (which is not really a choice if they want to stay competitive, they have to join).

Yea it's not really clear exactly what the requirements are other than there is branding requirements in GPP, along with other stuff. It is a time will actually tell thing though.

[LAwLz]

22 minutes ago, leadeater said:

Yea it's not really clear exactly what the requirements are other than there is branding requirements in GPP, along with other stuff. It is a time will actually tell thing though.

Yeah, it's not like a sudden change in all Nvidia/AMD branded products will go unnoticed.

[Humbug]

29 minutes ago, leadeater said:

Yea it's not really clear exactly what the requirements are other than there is branding requirements in GPP, along with other stuff. It is a time will actually tell thing though.

There will be individual variations in the customized GPP contracts which Nvidia signs with various vendors. And those agreements will remain confidential.

 

What we can do is be on the alert for lineup and branding changes. E.g. Will products such as the Asus ROG strix rx 580 based on Radeon become less common...

[asus killer]

9 minutes ago, Humbug said:

What we can do is be on the alert for lineup and branding changes. E.g. Will products such as the Asus ROG strix rx 580 based on Radeon become less common...

what if there really is a brand change? on the alert why, who really cares.

Anyone informed will not care at all. Those that shop blindly would most probably go with nvidia anyway, and that's mainly AMD's fault anyway as we can see by the historical market share graph that was posted

[leadeater]

14 minutes ago, asus killer said:

what if there really is a brand change? on the alert why, who really cares.

Anyone informed will not care at all. Those that shop blindly would most probably go with nvidia anyway, and that's mainly AMD's fault anyway as we can see by the historical market share graph that was posted

I kind of don't see how it's AMD's fault for people just defaulting to buying Nvidia when it's always had the much larger market share, that's just how it goes with brand recognition. Even in the later parts where AMD had the better products over the complete product stack Nvidia still outsold them, people just want Nvidia because Nvidia is viewed as the best because it's the most known.

 

Even if AMD's marketing was as good as Nvidia's this would still hold true just to a lesser extent. You don't blame AMD for this, it's to Nvidia credit it's the way it is because they worked hard to make sure it was this way.

 

There are a lot of markets where people want a product that isn't as good as many other options in the price range because they want the brand, cars are a good example and the audiophile world (some of these people are just stupid though lol).

[Razor01]

7 hours ago, leadeater said:

But it's not a choice partners can make though, it's the terms of the GPP so it's a contractual obligation. Before you bring up that GPP is optional, no it really isn't, not if you actually want to be competitive and actually stay in the GPU AIB market for Nvidia cards. No matter if it's all of ROG or just Strix or what ever it's certainly not a choice, it's about as much choice as you have to breath air or drink water. You can go without but not for long.

 

Forget all the back story and history to these companies, the products they make, who's is better it's all irrelevant because this is 100% a contractual issue.

 

 

It is a choice, instead of all the per vendor "at the time negotiations", everything is up front, you get them if you are part of the GPP or you don't need to ask about it at a later date, we won't give you those.   Actually for the smaller guys, things like launch partner status, and early access things, are extras that they will never get before, those were reserved for exclusive partners or large volume order partners.  Free game bundles were never there even for those guys too, the cost of the game bundles were tagged on to GPU costs before (at cost, nV got no profits from this)

 

Everything right now is the same way outside of branding.  nV now is being upfront with its partners.  If they aren't part of the GPP, there are no extra benefits.  You buy the product, you support the product on your own for your customers.

 

This is where AMD's weakness in the market gives nV leverage to do what they want to do with their products.

 

The only part I see a problem is when there are pass through rebates, but even those are done on a per vendor basis (and other parts of the MDF). Since this is where market changes in the channel which happen quite often, seasonality right, can do some major damage to partners.  Again though it was done on a per vendor basis in the past.  Linus mentioned this in his video, there was no overall contract for all vendors on this.

[Razor01]

Just now, leadeater said:

I kind of don't see how it's AMD's fault for people just defaulting to buying Nvidia when it's always had the much larger market share, that's just how it goes with brand recognition. Even in the later parts where AMD had the better products over the complete product stack Nvidia still outsold them, people just want Nvidia because Nvidia is viewed as the best because it's the most known.

 

Even if AMD's marketing was as good as Nvidia's this would still hold true just to a lesser extent. You don't blame AMD for this, it's to Nvidia credit it's the way it is because they worked hard to make sure it was this way.

 

There are a lot of markets where people want a product that isn't as good as many other options in the price range because they want the brand, cars are a good example and the audiophile world (some of these people are just stupid though lol).

 

 

Market share when nV and AMD cards were close in all metrics was always around a 5% swing in either direction.  That 5% - 10% is your mind share and marketing abilities of each company.

[asus killer]

4 minutes ago, leadeater said:

I kind of don't see how it's AMD's fault for people just defaulting to buying Nvidia when it's always had the much larger market share, that's just how it goes with brand recognition. Even in the later parts where AMD had the better products over the complete product stack Nvidia still outsold them, people just want Nvidia because Nvidia is viewed as the best because it's the most known.

 

Even if AMD's marketing was as good as Nvidia's this would still hold true just to a lesser extent. You don't blame AMD for this, it's to Nvidia credit it's the way it is because they worked hard to make sure it was this way.

 

There are a lot of markets where people want a product that isn't as good as many other options in the price range because they want the brand, cars are a good example and the audiophile world (some of these people are just stupid though lol).

i'm not going to disagree with you, still Maketing, nice boxes, ROG brand, bla bla bla vs 1080ti, who wins?

 

when you have 2 brands, and when we talk about gpus (not graphic cards) and 1080ti is almost in everyones mouths when someone mentions gpus what can you actually do with marketing against that? Even if i can't get the 1080ti i want me some of that, a 1070, a 1060,... you see what i mean.

Even when AMD has good gpu's they're always overshadowed. If i were AMD that's were i would focus, beat the Nvidia flagship, and of course without needing a nuclear powered PSU or some other drawback. And not beat them once and then go back to be overshadowed by Nvidia products, changing customers perceptions of a brand is hard, and no marketing can help you if you constantly trail behind. I guess they tried with Vega, sadly once again it was more of the same.

And all those years of comically bad cpus didn't help the AMD brand one bit.

 

I would be as rich as Linus if i got a euro every time a youtuber mentions "1080ti". 

 

Most uninformed customers and even shops prefer to buy/sell Intel or Nvidia over AMD products. We all know the stories of employees trashing AMD's products even if they are good products.

[leadeater]

49 minutes ago, Razor01 said:

It is a choice, instead of all the per vendor "at the time negotiations", everything is up front, you get them if you are part of the GPP or you don't need to ask about it at a later date, we won't give you those.   Actually for the smaller guys, things like launch partner status, and early access things, are extras that they will never get before, those were reserved for exclusive partners or large volume order partners.  Free game bundles were never there even for those guys too, the cost of the game bundles were tagged on to GPU costs before. 

 

Everything right now is the same way outside of branding.  nV now is being upfront with its partners.  If they aren't part of the GPP, there are no extra benefits.  You buy the product, you support the product on your own for your customers.

 

This is where AMD's weakness in the market gives nV leverage to do what they want to do with their products.

 

The only part I see a problem is when there are pass through rebates, but even those are done on a per vendor basis (and other parts of the MDF). 

I think you need to watch the WAN show and listen to Linus explain it, it's nothing like what you are saying it is. As for smaller partners those are the ones that need to be in GPP the most, again see WAN show.

 

Also Nvidia already enforces branding requirements, again WAN show.

 

GPP is nothing but forcing even more control over the AIB's, that's all any partnership program is at the end of the day for anyone. It's just a set of terms that both party's come to an agreement on but normally that's done with a reasonable level of equal partnership and influence in those negotiations, in this case it's not balanced and Nvidia hold all the leverage so you aren't going to get mutual agreement you're going to get comply or lose out.

 

And you still fundamentally don't get it, GPP is not a choice you must be part of it or forget selling Nvidia products at all otherwise lose money trying to do so.

 

Call it up front all you like but it's anything but. Partnership deals already existed before and those are clear already, they have to be. All the things in those existing contracts are now excursively in GPP, this is just a contract renegotiation but now with potentially unfair terms with no option other than to agree to them.

 

I have no idea why you want to keep supporting this and painting it as everything is fine don't worry about it, you can't possibly know that, you or anyone else here have not seen the GPP terms. Literally no one knows how good or bad it is other than what has been reported and even if you remove some potential truth stretching there is reasonably clear signs Nvidia is trying to force more control over partners that they should not have.

 

The only thing on a graphics card that Nvidia own is the GPU die, nothing else. It's not their product it's the partners and that is even more the case when it's a custom design not reference, but good luck trying to do that now if you're not in GPP. There are very few industries where parts suppliers like Nvidia, or even AMD for that matter, get as much control over the product it's being put in.