Nvidia gets anti-competitive with unsavory GeForce Partner Program

Scorpus

Posts: 2,156   +238
Staff member

Late last week an interesting story emerged surrounding Nvidia and the GeForce Partner Program (GPP). The GPP was quietly established by Nvidia as a means of working closer with OEMs and add-in board (AIB) partners on new products, product launches, marketing and so forth. At least, that is the intention of the program at the surface level, but an investigation by HardOCP has exposed a much less savory purpose behind the GPP, with elements of an anticompetitive and anti-consumer nature.

The story begins around two weeks ago, after HardOCP began their investigation into GPP. Getting ahead of the story, Nvidia’s director of partner marketing, John Teeple, put up a blog post outlining Nvidia’s new initiative.

Most of the blog post is pure Nvidia marketing speak, but a couple of lines are interesting. Nvidia claims the GPP is “designed to ensure that gamers have full transparency into the GPU platform and software they're being sold, and can confidently select products that carry the NVIDIA GeForce promise.” The blog post continues:

"This transparency is only possible when NVIDIA brands and partner brands are consistent. So the new program means that we'll be promoting our GPP partner brands across the web, on social media, at events and more. And GPP partners will get early access to our latest innovations, and work closely with our engineering team to bring the newest technologies to gamers

Partners are signing up, fast. They see the benefit of keeping brands and communication consistent and transparent.

The program isn't exclusive. Partners continue to have the ability to sell and promote products from anyone. Partners choose to sign up for the program, and they can stop participating any time. There's no commitment to make any monetary payments or product discounts for being part of the program.”

Nvidia mentions transparency several times, but it’s not clear what Nvidia is referring to exactly. Is it transparency into who Nvidia partners with for the GeForce Partner Program? Does the GPP now provide transparency into the partnerships Nvidia have held with OEMs and AIBs for years? As HardOCP points out, despite this transparency claim, it’s actually not clear which companies are part of the program.

A second interesting point is Nvidia’s statement that the GPP isn’t exclusive. Partners can continue to sell and promote products from anyone, according to the blog post. This is an important statement and we’ll get into why shortly.

Now on the surface, it doesn’t sound like there is anything inherently wrong with the GPP. It just sounds like a typical partnership arrangement between Nvidia and OEMs, delivering both sides mutual benefits like additional marketing and brand placement. But when HardOCP started talking to the companies involved, questions of anticompetitive and anti-consumer behaviour became apparent.

Note: MSI may or may not be part of the GeForce Partner Program. We don't know.

While investigating the GPP, Kyle Bennett from HardOCP spoke with seven companies, none of which wanted to go on the record. However they did speak anonymously about the GPP, and, according to Bennett, all the people he spoke to had similar opinions about the program.

Bennett summarizes the opinions as follows:

  • The terms of the GPP agreement are potentially illegal
  • The GPP will hurt consumer choices
  • The GPP will hurt a partner’s ability to do business with other companies like AMD and Intel

These opinions stem from a key component of the GPP agreement document, which Bennett read but decided not to publish. This component states that GPP partners must have their “gaming brand aligned exclusively with GeForce”. In other words, if a company like Asus wanted to join the GPP, they would not be allowed to sell AMD graphics cards as Republic of Gamers products. No more ROG-branded Radeon video cards, no more ROG laptops with AMD graphics inside.

Of course it’s up to the partners whether they want to be part of the GPP and accept these terms – Nvidia isn’t explicitly forcing anyone to sign up – but there are some pretty significant consequences to not signing up. Specifically, the GPP provides benefits to partners such as launch partner status, high-effort engineering engagements, marketing development funds, social media and PR support, game bundling and more.

Let's not forget Nvidia holds a dominant market position in the GPU market and has not only held but solidified that position in recent years -- due in large part to a superior product offering.

The GPP requires participants to align their gaming brands exclusively with GeForce, and if they don’t sign up to the program, their direct competitors that are part of the GPP will get special treatment from Nvidia. So there is a pretty strong incentive for OEMs and AIBs to sign up otherwise they’ll be left in the dust by the dominant player in the graphics market.

And it allegedly goes beyond the terms outlined in the GPP document. Some AIBs expressed concerns that it they do not sign up to the GPP, Nvidia would restrict GPU allocations and preference GPP members instead. This isn’t in the GPP agreement itself, but is allegedly happening through under-the-table agreements.

The biggest issue that stems from this kind of arrangement is that OEMs and AIBs are essentially forced into signing up to remain competitive among Nvidia partners. If they don’t join the GPP, they won’t get benefits like marketing development funds or launch partner status. And if a competitor does join, they will receive a genuine advantage, which puts anyone that decides not to join the GPP in a disadvantageous position.

Nvidia is requiring participants to align their gaming brands with GeForce exclusively, so in joining the GPP to remain competitive, OEMs and AIB partners wouldn't be able to use their gaming brand on AMD, Intel or any other competitor’s products. Technically a larger company would still be able to sell AMD GPUs under a non-gaming brand or run parallel promotions and push two brands, though evidently that would translate into higher costs and the gaming brands are almost always the most attractive to gamers.

The potential end result is less competition and less consumer choice. An everyday buyer looking for a graphics card for gaming would see almost exclusively Nvidia products from partners that join the GPP. AMD products would be pushed over to the side by an agreement that most manufacturers would find very difficult not to sign.

There’s a stark difference between what Nvidia is saying publicly about the GPP, and what the actual document allegedly states. Nvidia claims the program isn’t exclusive, yet the agreement asks for exclusivity with a partner’s main gaming brand. Nvidia claims partners can stop participating at any time, but in reality if they do, they will put themselves in a disadvantageous position. Nvidia claims the program is about transparency, but no one is saying which companies are part of the GPP.

This sort of behaviour is a prime candidate for investigation by regulators. Intel has been stung and stung hard by regulators in the US and Europe for anticompetitive conduct in the past.

The situation is a bit puzzling on the motives. Nvidia holds a dominant market position with their GeForce graphics products: they are already the most popular brand among buyers, they are essentially unopposed at the high end and aspects of their technology are ahead of their nearest competitor. There is no need to force OEMs into selling Nvidia products; buyers were already buying mostly Nvidia stuff anyway.

But like with Intel’s rise to power, Nvidia is finding itself in a dominant position and they're coming up with ways of crippling the competition. Perhaps it’s not the first time we’ve seen these sort of practices either...

GameWorks has famously been a source of angst among the gaming community, with Nvidia “partnering” with game developers to integrate their effects that mostly cripple performance on AMD GPUs – and sometimes hurt Nvidia performance, too – while providing little visual benefit.

HairWorks, for example, notoriously overused tessellation to harm AMD performance while AMD’s competing technology, TressFX, actually provided better performance on both Nvidia and AMD hardware. TressFX was an open technology that didn’t gain much traction while Nvidia signed partnerships to ensure HairWorks was used in games instead.

AMD are clearly concerned about the GPP, as they were the ones that shopped around this story in the first place, as HardOCP acknowledges. It’s always interesting when a company pitches publications like ours with a negative story on a competitor. It doesn’t happen too often, but when it does, someone at the company is usually pretty annoyed with the practices of their competitor (to say the least).

It's a fine line that divides fierce competition and the sort of practices that end up hurting consumer choice in the long run. Competition is healthy and we’re all a bit sick of companies abusing their dominant position to further their own interests rather than using it to create a vibrant hardware landscape.

Permalink to story.

 
Intel may have gotten fined but they actually haven't paid a cent in the EU yet. Not to mention, the amount of marketshare and revenue they stole from AMD was much more than any fine they had to pay.

I'm guessing Nvidia is thinking that the potential fines won't outweigh the revenue either. They probably want to create an iron grip on the market, just in case AMD / Intel do come out with a superior architecture. Navi is coming in 2019 and may very well be a MCM. GPUs are bigger than CPUs and thus have allot more to gain yield wise by making smaller dies. Cost would go way down and yields way up. It's why AMD can charge half of what Intel does for it's flagship CPU and pack in more cores at the same time.
 
Last edited:
Intel may have gotten fined but they actually haven't paid a cent in the EU yet. Not to mention, the amount of marketshare and revenue they stole from AMD was much more than any fine they had to pay.

I'm guessing Nvidia is thinking that the potential fines won't outweigh the revenue either. They probably want to create an iron grip on the market, just in case AMD / Intel do come out with a superior architecture. Navi is coming in 2019 and may very well be a MCM. GPUs are bigger than CPUs and thus have allot more to gain yield wise by making smaller dies. Cost would go way day and yields way up. It's why AMD can charge half of what Intel does for it's flagship CPU and pack in more cores at the same time.

Well, you saved me a fair bit of typing :) There's little doubt that between AMD's domination on the console front and Intel's impending use of AMD tech in their APUs the suits at Nvidia are getting jumpy. That can lead to stupid mistakes that attract the scorn of consumers and the hairy eyeball of regulators. Nvidia seems to falling right into that trap.
 
NVidia will continue to dominated the graphic card business. AMD can hardly compete vega cards could hardly compete with
nvidia's cards that were over a year old.
 
The next generation is going to be insanely expensive and not just because of miners. Expect baseline prices to remain higher indefinitely. Bottom line is being a PC gamer is getting more expensive across the board.
 
This looks like a total non-story. The whole article relies on this: "which Bennett read but decided not to publish. This component states that GPP partners must have their “gaming brand aligned exclusively with GeForce”." which is clearly a direct contradiction of Nvidia's published statement.
So we have an unsupported claim by a news generator (aka agent provocateur) versus a published statement from Nvidia. I think the first thing any investigation (or even a reputable tech blog), would do is find out if the unsupported claim is in fact correct.
 
AMD has the technology to KILL the GPU AIB market. A small taste of this is happening with Intel Kaby-G: i7-8809G Processor with Radeon RX Vega and Intel Core i7-8709G Processor with Radeon RX Vega.

If AMD were to release say an 8 core 16 thread Ryzen 2000 processor in the same package with maybe 2 7nm Vega gpus and 8gb HBM2 VRAM that would be the end of GPU Add in Boards.

The reduction in electrical path latency ALONE would see that Ryzen/Vega APU outperforms anything nVidia has on top tier.

AMD is probably concerned that the perceived disruption in the GPU AIB market would hurt them more than it helped, but Intel doesn't have to see it that way when they launch their own version of KABY-G with Intel HD graphics and lock out BOTH nVidia and AMD.

Intel has been eager to eliminate the PCIe bus for at least 10 years and in fact this was prohibited by the FTC in the ORDER by the FTC in the Intel vs AMD Settlement Agreement Docket no. 9431 until at least 2018.

If you think I am wrong, the why didn't Intel use nVidia GTX silicon for it's Kaby-G?

Still do not believe? Read this "Design and Analysis of an APU for Exascale Computing" published Feb 2017.

from here: http://www.computermachines.org/joe/publications/pdfs/hpca2017_exascale_apu.pdf

One final point, AMD does not dominate the console market because of it's graphics, but rather the economical combination of it's Jaguar CPU and Radeon graphic cores into a heterogeneous Accelerated Processing Unit. All for less than $125.00.

That level of performance just can not be provided by either Intel nor nVidia. Or it would have been.
 
Last edited:
This looks like a total non-story. The whole article relies on this: "which Bennett read but decided not to publish. This component states that GPP partners must have their “gaming brand aligned exclusively with GeForce”." which is clearly a direct contradiction of Nvidia's published statement.
So we have an unsupported claim by a news generator (aka agent provocateur) versus a published statement from Nvidia. I think the first thing any investigation (or even a reputable tech blog), would do is find out if the unsupported claim is in fact correct.

Then why hasn't Nvidia responded back to Kyle's email when he voiced his concerns & release a formal statement to the public to set the record straight once and for all?
 
This looks like a total non-story. The whole article relies on this: "which Bennett read but decided not to publish. This component states that GPP partners must have their “gaming brand aligned exclusively with GeForce”." which is clearly a direct contradiction of Nvidia's published statement.
So we have an unsupported claim by a news generator (aka agent provocateur) versus a published statement from Nvidia. I think the first thing any investigation (or even a reputable tech blog), would do is find out if the unsupported claim is in fact correct.
Leaving aside the obvious PR talk from Nvidia, what you are doing is trying to do (and say) is weird. From the very beginning it was made pretty clear that this isn't a copy-paste article on a no-name blog.
Not posting the contract is normal. That should be done only after Nvidia gets a chance to answer the accusations and if they continue to lie (and only then) he has the option of posting the whole thing. We are talking about confidential contracts here, not a random document, it's something that affects not just Nvidia.
 
AMD has no market share and sells all of it's gpus to miners lol, what are we supposed to do? wait until 2020 until they release gpus to compete with the 10 series?
 
AMD has the technology to KILL the GPU AIB market. A small taste of this is happening with Intel Kaby-G: i7-8809G Processor with Radeon™ RX Vega and Intel® Core™ i7-8709G Processor with Radeon™ RX Vega.

If AMD were to release say an 8 core 16 thread Ryzen 2000 processor in the same package with maybe 2 7nm Vega gpus and 8gb HBM2 VRAM that would be the end of GPU Add in Boards.

The reduction in electrical path latency ALONE would see that Ryzen/Vega APU outperforms anything nVidia has on top tier.

AMD is probably concerned that the perceived disruption in the GPU AIB market would hurt them more than it helped, but Intel doesn't have to see it that way when they launch their own version of KABY-G with Intel HD graphics and lock out BOTH nVidia and AMD.

Intel has been eager to eliminate the PCIe bus for at least 10 years and in fact this was prohibited by the FTC in the ORDER by the FTC in the Intel vs AMD Settlement Agreement Docket no. 9431 until at least 2018.

If you think I am wrong, the why didn't Intel use nVidia GTX silicon for it's Kaby-G?

Still do not believe? Read this "Design and Analysis of an APU for Exascale Computing" published Feb 2017.

from here: http://www.computermachines.org/joe/publications/pdfs/hpca2017_exascale_apu.pdf

One final point, AMD does not dominate the console market because of it's graphics, but rather the economical combination of it's Jaguar CPU and Radeon graphic cores into a heterogeneous Accelerated Processing Unit. All for less than $125.00.

That level of performance just can not be provided by either Intel nor nVidia. Or it would have been.
Except that, no matter what you put into any processor, a dedicated add in board will have a much larger power budget and cooling system to work with, and thus will be more powerful, and customers will buy that instead.

I mean, hell, if AMD had the technology to put "2 7nm Vega gpus and 8gb HBM2 VRAM" on an APU, why not put, say, 4 or 6 on a dedicated board for 2-3x the performance?
 
Last edited:
Leaving aside the obvious PR talk from Nvidia, what you are doing is trying to do (and say) is weird. From the very beginning it was made pretty clear that this isn't a copy-paste article on a no-name blog.
Not posting the contract is normal. That should be done only after Nvidia gets a chance to answer the accusations and if they continue to lie (and only then) he has the option of posting the whole thing. We are talking about confidential contracts here, not a random document, it's something that affects not just Nvidia.


"...versus a published statement from Nvidia."

...you have drunk the kool-aid! A true nVidia zealot. Have faith in the word of the gospel according to nVidia.
 
"Technically a larger company would still be able to sell AMD GPUs under a non-gaming brand or run parallel promotions and push two brands, though evidently that would translate into higher costs and the gaming brands are almost always the most attractive to gamers."

I know that branding is always pushed in our faces when it comes to new gaming hardware. But in my experience Gamers are more performance orientated than they are brand orientated. As a gamer I will always select the hardware of best performance. (which does happen to be Nvidia right now)

I can only pity AMD (and other lesser known brands) because of Nvidia's tactics on this. But ultimately complaining and forcing and investigation into Nvidia is only going to go so far. If AMD really want to hurt Nvida they will need to offer a product that completely outstrips their competition.
 
Except that, no matter what you put into any processor, a dedicated add in board will have a much larger power budget and cooling system to work with, and thus will be more powerful, and customers will buy that instead.

I mean, hell, if AMD had the technology to put "2 7nm Vega gpus and 8gb HBM2 VRAM" on an APU, why not put, say, 4 or 6 on a dedicated board for 2-3x the performance?

THAT IS EXACTLY WHAT AMD INTENDS TO DO!!!


What you are failing to understand is Intel Kaby-G is an Intel CPU chiplet, a Radeon Vega chiplet and HBM2 on an Interposer (Intel calls it an EMIB but same thing) in a single package. It is EXACTLY what I pointed out AMD could make with Ryzen 2000 and 2 7nm Vgea cores. Except Kaby-G is a smaller version.

As far as TDP is concerned, It would likely draw not much more power than a 16 core 32 thread Threadripper does now. I don't know what your babbling about power has to do with anything.

You have also not read the White Paper that I provided a link to and again would have understood that keeping the electrical path extremely short, the data carrying energy demand is also very low. Again another failure of you trying to understand before you shoot from the hip.

In actual fact, the concept AMD is exploring is an 4 - 4 core CPU's with 4 GPU Multi Chip Module along with HBM2 memory. AMD calls this ENA or Exascale Node Architecture. The package is called an EHP or Exascale Heterogeneous Processor. I suspect that 4 core 8 thread cpu's have incredibly high yields. But there should no reason that 8c/16t cpu chiplets could not be used with advanced lithography.

From the White Paper that YOU failed to read:

"The EHP is an accelerated processing unit (APU) consisting of a balanced mix of CPU and GPU compute resources integrated with in-package 3D DRAM. The overall structure makes use of a modular “chiplet” design, with the chiplets 3D-stacked on other “active interposer” chips. The aggressive use of advanced packaging technologies enables a large amount of computational and memory resources to be located in a single package. Fig. 2 shows a conceptual rendering of the EHP. The EHP consists of two primary types of resources. In the center of the EHP are two CPU clusters, each consisting of four multi-core CPU chiplets stacked on an active interposer base die. On either side of the CPU clusters are a total of four GPU clusters, each consisting of two GPU chiplets on a respective active interposer. Upon each GPU chiplet is a 3D stack of DRAM (e.g., some future generation of JEDEC high-bandwidth memory (HBM) [12]). The DRAM is directly stacked on the GPU chiplets to maximize bandwidth (the GPUs are expected to provide the peak computational throughput) while minimizing memoryrelated data movement energy and total package footprint. CPU computations tend to be more latency sensitive, and so the central placement of the CPU cores reduces NUMA-like effects by keeping the CPU-to-DRAM distance relatively uniform. The interposers underneath the chiplets provide the interconnection network between the chiplets"

AMD also discusses this idea in Patent # 20140089609 granted March 2014. The patent discusses the Use of Memory controllers imbedded in an Interposer with CPU and GPU chiplets and stacked HBM.

As I pointed out before and AGAIN you failed to understand was likely AMD has concerns regarding the disruption to not only the GPU AIB market but the desktop CPU market as well. There is always unintended consequences along with the deployment of disruptive technology.

Intel did hire Raja Koduri to do what exactly?
 
Last edited:
Well, you saved me a fair bit of typing :) There's little doubt that between AMD's domination on the console front and Intel's impending use of AMD tech in their APUs the suits at Nvidia are getting jumpy. That can lead to stupid mistakes that attract the scorn of consumers and the hairy eyeball of regulators. Nvidia seems to falling right into that trap.

Given that nVidia had to sue Intel for access to their transport microarchitecture back in 2010 or so you are probably right. Intel does not want nVidia Inside anymore. They do not want AMD either but since AMD produces x86 cpu's that impact will not be felt... much.

Intel wants to eliminate the PCIe transport bus. And in actual fact it really serves no purpose in CONSUMER gaming or workstations other than to move data to GPU's.

If Intel HD was better then PCIe would be gone!
 
Except that, no matter what you put into any processor, a dedicated add in board will have a much larger power budget and cooling system to work with, and thus will be more powerful, and customers will buy that instead.

I mean, hell, if AMD had the technology to put "2 7nm Vega gpus and 8gb HBM2 VRAM" on an APU, why not put, say, 4 or 6 on a dedicated board for 2-3x the performance?

Hmmm... you have not heard of AMD Fenghuang APU then? It is not yet quite what I had pointed out but it is getting there. It is probably the best that can be done economically with 14nm lithography. Expect better wth 7nm.

Fenghuang is supposedly 8 core 16 thread Zen cpu chiplet with 28 Compute Unit Vega GPU chiplet and HBM2. I say supposedly as this has been a pretty consistent rumour since early December of 2017 and in actual fact was discussed obliquely several times in 2016. Given the existence of Kaby-G, the strength of the White Paper and Patents, I would say AMD has something to surprise the market with. Considering the strength and favorable acceptance of Ryzen APU's this past month I would say Fenghaung will surprise.

http://www.userbenchmark.com/UserRun/7679982
 
It's a story. Sounds a bit like sour grapes.
No one is stopping anyone from buy anything. It's called marketing. .


SOUR GRAPES?

TechSpot has no dawg in this hunt?

However Intel too believed it was all just marketing until they got sued by the FTC for illegal trade practices.
 
......
So we have an unsupported claim by a news generator (aka agent provocateur) versus a published statement from Nvidia. I think the first thing any investigation (or even a reputable tech blog), would do is find out if the unsupported claim is in fact correct.

Really FUD?? The point of journalism as someone as famously said is to "comfort the afflicted and afflict the comfortable".

Calling people name like "agent provocateur" and denoucing it as "unsupported claims" does nothing but to show that you like to provide marketing cover FUD for whatever corporation that you like worship, in this case nVidia. People might even suspect you might be a paid agent to do social media damage control.

But bottome line, nVidia is NOT denying the existence of the NPP and fact that the NPP exists is more than enough to support the author's claims such that legal law enforcement entities should to the investigation. Even if the legal investigation does not find any wrongdoing, it is fully within the rights, protected by 1st amendment free speech, for writers journalists to make such speculations. You can keep your fascist states in Russia and China where the government and corporations are operate via unholy corruption on a daily basis and working hard to control the media.
 
It's a story. Sounds a bit like sour grapes.
No one is stopping anyone from buy anything. It's called marketing. .
If this is truly what the program is about then Nvidia will indeed get fined just like Intel. There are laws in place to prevent/punish such things. It's not "marketing" if laws are being broken, you can't just do anything.

"No one is stopping anyone from buying anything" - no, they are not stopping others from buying AMD stuff (not completely at least), but they do put significant pressure on you to not do it and Nvidia has the monetary power and market share to force you to do it if they want to. And the gaming brands exclusivity is without a doubt illegal (IF it turns out to be true).
 
The next generation is going to be insanely expensive and not just because of miners. Expect baseline prices to remain higher indefinitely. Bottom line is being a PC gamer is getting more expensive across the board.

I sure hope PC gamers will not be insane and eagerly pay the "insanely expensive" prices. The market will self correct when they are not moving the goods and prices will come down.

As of this moment bitcoin price is under 9k
https://charts.bitcoin.com/chart/price

It is said "in a gold rush, the people that make the money are the ones selling the shovels." Well eventually there will be too many expensive shovels, and not enough people mining.
 
The next generation is going to be insanely expensive and not just because of miners. Expect baseline prices to remain higher indefinitely. Bottom line is being a PC gamer is getting more expensive across the board.
Once the second hand market gets saturated the prices for new products should go down. I predict that this will happen by then end of summer, but don't quote me on that :D.
 
Once the second hand market gets saturated the prices for new products should go down. I predict that this will happen by then end of summer, but don't quote me on that :D.


The second hand market has zero impact. Everyone likes to believe that somehow eBay has some major impact. But when you consider that MAYBE a dozen or so second hand gpu's are sold on eBay each week, this is actually a drop in the OCEAN in a world wide GPU market.

I don't know anyone who waste money on a second hand burned out mining GPU with no warrantee.

Dealer stock is different. There are dealers on eBay and they are approved dealers selling NEW stock.
 
Last edited:
Back