Friday, January 23rd 2026

NVIDIA Ends "Open Price Program," Pushing GPUs Way Past MSRPs

According to Roman "der8auer" Hartung and HardwareLuxx, NVIDIA has effectively ended its Open Price Program (OPP). This program was intended to help NVIDIA's add-in card partners sell GPUs at NVIDIA-suggested MSRPs. The program reportedly ended a few days ago, which means the pricing of GPUs from NVIDIA's AIC partners will now fluctuate beyond MSRP. NVIDIA supplies its GPU dies and GDDR7 memory as a DIY kit to AIC partners, who purchase these kits at predetermined prices. They then install the GPU and memory on their customized PCBs to create third-party GPU designs. However, due to the scarcity of memory, GDDR7 prices have been rising significantly, and NVIDIA is passing these cost increases onto its AIC partners.

As per to HardwareLuxx, the OPP program helped NVIDIA's AICs maintain MSRPs for their custom GPUs. Participation in the OPP program allowed manufacturers to list their GPUs at MSRPs, making them more appealing to consumers due to their perceived value. However, with the end of the OPP program, prices are expected to rise significantly since NVIDIA will no longer "subsidize" these listings. This change is likely influenced by the strained supply chain in the memory segment, where manufacturers are running low on available memory, leading to increased lead times and higher costs. Fortunately, NVIDIA continues to provide GDDR7 memory with its GPUs, so AIC partners don't need to source it independently.
There were initial rumors about NVIDIA discontinuing the GeForce RTX 5070 Ti GPUs. Both the RTX 5070 Ti and the RTX 5080 use the same GB203 base die and 16 GB of GDDR7 memory. NVIDIA is reportedly prioritizing the RTX 5080, which has a higher MSRP. The RTX 5070 Ti has not been discontinued, but its production will make up a smaller portion of the manufacturing capacity. Since the RTX 5080 has a $250 higher MSRP, it makes sense to favor the higher-end model. In typical GPU manufacturing, chips that fully meet the RTX 5080's performance and power specifications are allocated to that SKU, while those that fall slightly short can be repurposed through binning for the RTX 5070 Ti. This approach optimizes yield without increasing silicon consumption, as both models come from the same die family.
Source: HardwareLuxx
Add your own comment

34 Comments on NVIDIA Ends "Open Price Program," Pushing GPUs Way Past MSRPs

#26
notoperable
We need to clarify NVIDIA's interpretation of open. There is no doubt that NVIDIA's interpretation of "Open source" for example - it's differs a lot from the publicly acknowledged one, so Open Price might follow the same pattern.
Posted on Reply
#27
mechtech
Meh

Charge $2k for a 5070. Nvidia owners will pay it no questions asked.
Posted on Reply
#28
Lycanwolfen
Glad I bought my MSI Expert 5080 RTX when I did last year. Great card btw.
Posted on Reply
#29
Dirt Chip
EndymioI'm willing to lay odds that 2027 RAM prices will be lower than 2026. Are you willing to take that bet? People still haven't internalized that HBM capacity coming online this year will -- by reducing the price differential between it and DDR5 -- also act to depress DDR5 prices.
The big corps make up there mind to migrate most pc to the cloud, leveraging AI as catalytic engine weather it had anything to do with it or not.
The wheels are in motion to push any kind of pc component out of $$ reach for larger and larger segment's.
In the end it will be left as boutique hobby, much like human drived cars will be (in 10 years or so).

We will see alternating excuses for the cost to go up but in the end the only real reason is that they decide it's more profitable to sell compute by cloud as monthly subscription then to sell the different components to make a PC.

The end is (hopefully not so) nigh, brace yourself.
Posted on Reply
#30
Unregistered
OnasiDon't remember saying anything about DDR prices though? They might improve, but, last I checked, this is a GPU price thread.
If you're solely interested in GPUs, then you should be even happier with AI's growth, as videogames and gamers have long since ceased to provide enough revenue to fund GPU development. If it wasn't for AI and datacenter revenues, you'd still be using your 1080ti, longing for some future upgrade.
Posted on Edit | Reply
#31
Onasi
EndymioIf you're solely interested in GPUs, then you should be even happier with AI's growth, as videogames and gamers have long since ceased to provide enough revenue to fund GPU development. If it wasn't for AI and datacenter revenues, you'd still be using your 1080ti, longing for some future upgrade.
I have no idea why you are inferring things I have never said. I made a comment on the pricing, that’s all. I couldn’t care less for what happens with the gaming industry in terms of GPU availability, I have long advocated for a freeze on graphics development for games. I sat on a 1070 for a decade without any need for upgrade and bought a 5070 basically just cause. I barely play anything intensive and was more interested in a boost to compute to work from home now and then. That’s my position. If you are here just to tell me that the AI revolution is actually awesome and stuff - I am not interested in having that mind-numbing discussion, sorry.
Posted on Reply
#32
xorbe
I think it's possible we see total stagnation of consumer gaming GPU progress, and we just "limp along" with games targeting 8GB mid-range for a few years, with the death of high end for a while.
Posted on Reply
#33
Youseethis
To summarize the video alleges:

Nvidia cuts the OPP program to subsidize the cards to msrp
Nvidia reduces the 5070 ti production in favor of the rtx 5080
Both will get massive price hikes.

Also posed the question: why nvidia bothered in the first place to pus a low msrp.

I wonder something else: were the premium models also subsidized in some capacity?
If not then the subsidy was somewhat predictable because the pricing for the premium rtx 5080 models for example was 1200-1300$ (except asus).

So, up to 200$?

The memory (16gb) was less than 80$ because that was the price for 8x3gb gddr7 kits according to this:
www.techpowerup.com/337853/samsung-3-gb-gddr7-chips-sold-in-chinese-retail-attracts-memory-modders

So if we assume a 4x increase that would be less than 320$ for a price differential of 320-80=240$
Add them together and we get 440$

Checks out on vram and scarcity alone because for now the increases are about 300$ for the US and around the same in euros for the EU market.
Can we expect an sudden 100-200$ price hike then+a further 80-160$ from the ram?

Meaning at worst a 1200+300+200+160=1860$ for a premium rtx 5080?

Posted on Reply
#34
Dirt Chip
YouseethisMeaning at worst a 1200+300+200+160=1860$ for a premium rtx 5080?
1800$ is the average 5080 cost at my home country today..
Those 2500$ 5080 will come here sooner than later…
Posted on Reply
Add your own comment
Nov 16th, 2026 07:41 CST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts