Eshelmen
Supreme [H]ardness
- Joined
- Feb 3, 2004
- Messages
- 7,302
I've been following Nvidia's pricing and the recent reports about the 5090, and I'm starting to wonder how much longer this can go on before somebody seriously investigates their supply practices.
The 5090 launched at $1,999. We're now seeing listings around $5,000, sometimes higher. And frankly, MSRP has been a joke for most of this generation.
Now there are reports that Nvidia may have stopped allocating new GB202 chips to consumer 5090 manufacturers, favoring professional and AI products instead.
Nvidia says the 5090 isn't discontinued. Okay, but that doesn't necessarily mean they're still allocating new chips to board partners at the same rate. Are they actually maintaining production for consumers, or just continuing shipments from existing inventory?
I'd like to know. Because Nvidia controls the chips. ASUS, MSI, Gigabyte, and the other manufacturers can't build these cards without them. So while Nvidia might not directly set retail prices, it absolutely influences how much product makes it to market.
And there's another part to this that I think deserves more attention. If the 5090 becomes practically unavailable, what's Nvidia's next flagship for consumers? The 5080.
Now you've got people who would've bought a 5090 looking at a 5080 instead. That puts additional demand on the 5080, which is already expensive and impossible to find at MSRP. Buyers who can't justify that price start looking at the 5070 Ti, and the same thing can happen further down the lineup.
Look at some of the retail prices we've been seeing:
Now consider Nvidia's financial position. For fiscal 2026, they reported $193.7 billion in Data Center revenue compared to $16 billion in Gaming.
Twelve times the revenue. So obviously theres a massive incentive to prioritize AI hardware. I don't have a problem with Nvidia making money, and I'm not suggesting they should be forced to manufacture unlimited gaming GPUs.
What I have a problem with is the idea that Nvidia has little responsibility for what's happening to consumer pricing when it controls the supply of the chips these cards depend on.
There's a difference between being unable to meet demand and deciding to allocate less product to a market. Both can be legal, but they're not the same thing.
And if any of those allocation decisions involve restrictive agreements or practices that unlawfully protect Nvidia's market position, then we're getting into antitrust territory.
Section 2 of the Sherman Act deals with maintaining monopoly power through exclusionary conduct. High prices alone aren't enough to establish a violation. Neither is prioritizing AI over gaming.
But that doesn't mean Nvidia's supply and distribution practices are automatically beyond scrutiny.
Nvidia has already disclosed that competition regulators have requested information about its GPU sales and supply allocation practices.
I'd like to see some actual numbers. How many consumer GPUs are they capable of producing? How many are being allocated to board partners? Are those allocations being reduced? And what restrictions, if any, are placed on those partners?
Especially with the 5090 reports.
I'm not saying Nvidia has definitely stopped producing consumer 5090s. Those reports haven't been confirmed, and Nvidia disputes them.
But if the flagship is effectively being pushed out of the consumer market, that potentially changes demand throughout the rest of the lineup. The 5080 becomes the top realistically available card, and everything below it can experience additional pricing pressure.
I don't think it's unreasonable to question whether that's simply an unavoidable consequence of manufacturing constraints or whether Nvidia's business decisions are contributing to it.
And to be clear, I'm not claiming that expensive GPUs automatically equal an antitrust violation. I'm questioning whether there's enough market influence, financial incentive, and potential competitive harm here to justify an investigation into Nvidia's actual allocation and distribution practices. I think there is.
If Nvidia's shortages are legitimate, fine. But I'd like to see that established through evidence rather than just taking their word for it. Because when a $1,999 graphics card is being listed for $5,000, and the rest of the lineup is following with ridiculous pricing of its own, I think consumers have every reason to start questioning who's making the supply decisions.
And whether Nvidia is deliberately creating or maintaining shortages that drive up consumer prices while protecting its own market position.
And there's also the GDDR7 memory shortage to consider. Nvidia isn't just competing for GPU manufacturing capacity; its massive AI business also gives it considerable purchasing power with memory suppliers like Samsung, SK Hynix, and Micron.
I understand AI accelerators primarily use HBM rather than GDDR7, but these companies still have to decide where to invest their manufacturing capacity and which customers get priority. Given Nvidia's financial influence, I think it's fair to question how much leverage it has over those decisions and whether that affects consumer GPU availability.
A legitimate memory shortage doesn't automatically mean Nvidia has no influence over how limited supplies are allocated.
Ultimately, I'd like to see more transparency around Nvidia's supply and allocation decisions. There may be perfectly legitimate reasons for the shortages and pricing we're seeing, but given Nvidia's influence across the GPU industry, I think there's enough here to warrant a closer look by regulators.
The 5090 launched at $1,999. We're now seeing listings around $5,000, sometimes higher. And frankly, MSRP has been a joke for most of this generation.
Now there are reports that Nvidia may have stopped allocating new GB202 chips to consumer 5090 manufacturers, favoring professional and AI products instead.
Nvidia says the 5090 isn't discontinued. Okay, but that doesn't necessarily mean they're still allocating new chips to board partners at the same rate. Are they actually maintaining production for consumers, or just continuing shipments from existing inventory?
I'd like to know. Because Nvidia controls the chips. ASUS, MSI, Gigabyte, and the other manufacturers can't build these cards without them. So while Nvidia might not directly set retail prices, it absolutely influences how much product makes it to market.
And there's another part to this that I think deserves more attention. If the 5090 becomes practically unavailable, what's Nvidia's next flagship for consumers? The 5080.
Now you've got people who would've bought a 5090 looking at a 5080 instead. That puts additional demand on the 5080, which is already expensive and impossible to find at MSRP. Buyers who can't justify that price start looking at the 5070 Ti, and the same thing can happen further down the lineup.
Look at some of the retail prices we've been seeing:
- 5090: $1,999 MSRP, around $5,000 retail
- 5080: $999 MSRP, around $1,700 retail
- 5070 Ti: $749 MSRP, around $1,200 retail
- 5070: $549 MSRP, around $900 retail
- 5060 Ti 16GB: $429 MSRP, around $800 retail
Now consider Nvidia's financial position. For fiscal 2026, they reported $193.7 billion in Data Center revenue compared to $16 billion in Gaming.
Twelve times the revenue. So obviously theres a massive incentive to prioritize AI hardware. I don't have a problem with Nvidia making money, and I'm not suggesting they should be forced to manufacture unlimited gaming GPUs.
What I have a problem with is the idea that Nvidia has little responsibility for what's happening to consumer pricing when it controls the supply of the chips these cards depend on.
There's a difference between being unable to meet demand and deciding to allocate less product to a market. Both can be legal, but they're not the same thing.
And if any of those allocation decisions involve restrictive agreements or practices that unlawfully protect Nvidia's market position, then we're getting into antitrust territory.
Section 2 of the Sherman Act deals with maintaining monopoly power through exclusionary conduct. High prices alone aren't enough to establish a violation. Neither is prioritizing AI over gaming.
But that doesn't mean Nvidia's supply and distribution practices are automatically beyond scrutiny.
Nvidia has already disclosed that competition regulators have requested information about its GPU sales and supply allocation practices.
I'd like to see some actual numbers. How many consumer GPUs are they capable of producing? How many are being allocated to board partners? Are those allocations being reduced? And what restrictions, if any, are placed on those partners?
Especially with the 5090 reports.
I'm not saying Nvidia has definitely stopped producing consumer 5090s. Those reports haven't been confirmed, and Nvidia disputes them.
But if the flagship is effectively being pushed out of the consumer market, that potentially changes demand throughout the rest of the lineup. The 5080 becomes the top realistically available card, and everything below it can experience additional pricing pressure.
I don't think it's unreasonable to question whether that's simply an unavoidable consequence of manufacturing constraints or whether Nvidia's business decisions are contributing to it.
And to be clear, I'm not claiming that expensive GPUs automatically equal an antitrust violation. I'm questioning whether there's enough market influence, financial incentive, and potential competitive harm here to justify an investigation into Nvidia's actual allocation and distribution practices. I think there is.
If Nvidia's shortages are legitimate, fine. But I'd like to see that established through evidence rather than just taking their word for it. Because when a $1,999 graphics card is being listed for $5,000, and the rest of the lineup is following with ridiculous pricing of its own, I think consumers have every reason to start questioning who's making the supply decisions.
And whether Nvidia is deliberately creating or maintaining shortages that drive up consumer prices while protecting its own market position.
And there's also the GDDR7 memory shortage to consider. Nvidia isn't just competing for GPU manufacturing capacity; its massive AI business also gives it considerable purchasing power with memory suppliers like Samsung, SK Hynix, and Micron.
I understand AI accelerators primarily use HBM rather than GDDR7, but these companies still have to decide where to invest their manufacturing capacity and which customers get priority. Given Nvidia's financial influence, I think it's fair to question how much leverage it has over those decisions and whether that affects consumer GPU availability.
A legitimate memory shortage doesn't automatically mean Nvidia has no influence over how limited supplies are allocated.
Ultimately, I'd like to see more transparency around Nvidia's supply and allocation decisions. There may be perfectly legitimate reasons for the shortages and pricing we're seeing, but given Nvidia's influence across the GPU industry, I think there's enough here to warrant a closer look by regulators.
Last edited: