NVIDIA Is Absorbing Memory Costs to Protect Gamers “For Now”; Here’s the Other Side of GPU Shortages

There have been a lot of rumors floating around NVIDIA’s RTX 50 GPUs and how memory shortages are impacting the supply/prices, and we thought we should do a detailed report on our findings, which paints a whole different picture of the ongoing situation.

Is NVIDIA Canning GeForce RTX 50 GPUs In Light of The Recent Memory Shortages? Here’s What Our Sources Tell Us

Since November, when the memory crisis started to creep into the consumer hardware segment, there have been various reports highlighting that GPU vendors, both NVIDIA & AMD, will increase prices, and more recently, reports that NVIDIA will be entirely ending production of certain GPUs in its inventory.

We went over and asked around some reputable industry sources on the matter, sources close to NVIDIA, sources close to their AICs, and sources at various AICs. Our findings have provided us with more information on the matter, and while the memory shortage’s impact is real and affects the entire consumer market, most of the information is either not conveyed appropriately or presented in a way that’s utterly devoid of truth. So here’s what we were able to get, and bear with us, the report will tackle various aspects of the consumer GPU market, particularly NVIDIA GPUs.

NVIDIA’s Official Statement – Strong Demand in a Memory-Constrained Ecosystem

First and foremost, we have to mention NVIDIA’s most recent and official statement, which was provided to us just in time for this report. The report aligns with what most AICs & partners are saying: that demand for GeForce RTX 50 GPUs remains strong despite prices rising. In Q4, NVIDIA’s GeForce RTX GPUs saw strong sales, and the Deals Season (Black Friday, etc.) prompted a buying spree among gamers and consumers.

There was an extra push from consumers to get these cards, driven by factors beyond the forthcoming memory crisis, such as tariffs. Yeah, we all thought at some point that tariffs would be the end of the consumer gaming segment, and that is also reflected in the recent IDC report, which showed big growth in the PC market in Q4 2025. But enter memory shortages, that’s just a whole new level of misery, and there’s little that consumers can do to avert it.

NVIDIA confirms in its official statement that memory supply is constrained, but it is working closely with its suppliers to maximize memory availability. NVIDIA also confirms that they continue to ship all GeForce SKUs. These two statements will be covered in detail later on in this post, but for now, let’s list down NVIDIA’s official statement.

Demand for GeForce RTX GPUs is strong, and memory supply is constrained. We continue to ship all GeForce SKUs and are working closely with our suppliers to maximize memory availability.

NVIDIA to Wccftech

Both GDDR6 and GDDR7 Memory Supply Is Constrained

Continuing our report, we thought about clarifying what we mean by memory shortages. We have seen DDR5 getting expensive, DDR4 is following suit, and the most recent reports suggest that DDR3 prices are on the rise. Surely our gaming graphics cards don’t use these memory standards, so why do GPU makers say that memory supply is constrained?

To answer it, it’s not that a single memory type is experiencing shortages, but it is the entire memory industry. DRAM, or memory as we call it, is used across several products and platforms. Computers use DDR standards for PCs and Data Centers, Graphics Cards use GDDR, and HPC/AI chips use HBM memory standards.

There are a handful of major companies that produce DRAM itself; the most widely known are Samsung, SK hynix, and Micron. These three are responsible for the bulk of the DRAM output on the global stage, & considering how huge they are, each has dedicated fabs and production plants where these memory chips are made. But as demand for a particular type of DRAM rises, production lines are prioritized for that type.

Currently, HBM and DDR are the most widely used DRAM products due to rising AI demand. Data Centers and AI powerhouses are gobbling up these memory types like nothing, and major companies have already back-ordered years’ worth of supply, which these DRAM manufacturers are currently working to fulfill. GDDR for graphics cards and GPUs is also being produced, but not at the same level as the other two memory types, which is causing the memory shortages. At the same time, prices to procure such memory also rise, and the usual supply-and-demand logic applies.

So it’s not just GDDR7 graphics cards, like the RTX 50 series, that will be expensive; even older GDDR6 cards will be expensive, as memory prices have shot up. But there’s some good news on this matter.

Gaming GPU Prices Are Affected By Memory Cost Increases

One thing that our sources wanted to point out from the start is that NVIDIA has held off on passing through any memory cost increases to their AICs and consumers longer than other brands.

NVIDIA held off passing through any memory cost increases longer than most companies. Not all of the memory costs increases are being passed through to partners and end users.

Industry Source to Wccftech

As we stated above, memory prices have gone up due to supply constraints, but despite that, it looks like NVIDIA has favored its customers and limited the amount of memory costs passed through to consumer-centric GPUs. The source also pointed out that they have received word on the memory price increases for GPUs, & they are in fact lower than what most have been quoting online. You can recall our month-old post where some channels highlighted $20-$40 US hikes with bumps throughout 2026.

And even if NVIDIA were to purchase memory from DRAM manufacturers for its GPU at exorbitant prices, there’s still the demand for RTX 50 GPUs, and that sort of supply is currently not available for the latest gaming graphics cards. That leads to an overall shortage of consumer graphics card output, leaving consumers with no option but to pay higher prices.

They (NVIDIA) are not passing through all memory cost increases to partners, but memory supply is causing the constraints.

Industry Source to Wccftech

Is NVIDIA Cutting Off GPU+Memory Bundles To AICs?

Another story on the same matter, related to AICs, that appeared recently said that NVIDIA has cut off its GPU+DRAM bundles to AICs, forcing AICs to procure memory for each GPU themselves. Like AMD & Intel, NVIDIA provides its partners with a GPU+DRAM combo.

So let’s say an AIB asks 100 units of the GeForce RTX 5060 GPU from one of NVIDIA’s distributors. The AIC will receive two things: the GPU or chip used in the RTX 5060 graphics card, and the DRAM modules. It’s been the case since forever. The AIC will then produce a PCB that aligns with NVIDIA’s guidance spec and perform the final retrofitting, inspection, validation, and packaging before sending it to their warehouse, where it’ll be shipped to retail distributors in bulk.

To answer this story, we can confirm with our industry sources that it was, fortunately, false, and that there have been no major changes in how NVIDIA manages DRAM supply with its partners (AICs). They still receive the aforementioned GPU+DRAM combo.

No. There have been no major changes in how NVIDIA manages memory with its customers.

Industry Source to Wccftech

No GPUs Are Being EOL’d, & No 16 GB Or Higher VRAM Models Are Being Axed

Now, coming to the main question, and something that was reported recently. But first of all, let me make this clear: all journalists and tech outlets do their best in sourcing their information, and present it in the best way possible to their readers. Every one of us is responsible for presenting facts as they are, and not a matter of distrust or controversy. This is why everything you hear on the web, from reputable outlets, has its weight. But oftentimes, it’s not the outlet itself but the source or means by which that information was gathered that causes the discrepancy.

We are talking about ASUS’s recent comment on the RTX 5070 Ti being EOL’d. This was covered by Hardware Unboxed, a highly reputable outlet I have great respect for, their work, and how they present their data. Surely, if someone at ASUS tells me that a particular GPU is being EOL’d, I’d take their word, because they are one of, if not the, biggest NVIDIA AIC partners out there. The report’s intention was right, but it looks like ASUS was wrong about one thing: the RTX 5070 Ti is not EOL.

There have been reports in the past, too, about the 5090 being EOL, the 5080 being EOL. It’s almost like the past few months, all RTX 50 cards have been declared EOL by someone, and this time, it was ASUS. The thing is, no card is being EOL’d; they are still up there on retail, sometimes out of stock, sometimes back in stock at exorbitant prices, and sometimes below MSRP (though these days you’ll be lucky to find one within $100 of the MSRP).

All RTX 50 VRAM Models Continue To Be In Production

When speaking with our industry source and the respective AICs, we confirmed that NVIDIA is not ending shipments of any of the said GPUs to them, and that they continue to supply all RTX 50 GPUs. All of these GPUs are still in production, and none is being prioritized over another, nor is any VRAM config favored over another.

All 8 GB, 12 GB, 16 GB, and 32 GB models in the existing RTX 50 series remain in stock. The source even wanted to quote NVIDIA directly, saying, “They support every market segment.” We cannot ignore the fact that there will be a certain level of supply drop to AICs based on memory supply, as we reported earlier, but all things our source points to the fact that NVIDIA won’t outright EOL any of their cards, nor have they done so for any RTX 50 GPU at the moment.

Certain cards will be more popular than others, and certain cards have always been produced in higher quantities than others. For example, the 5090, being the flagship, won’t have the same supply output as the RTX 5050, 5060, or 5070, nor will the 5070 Ti, a very popular option, have the same production level as the RTX 5060 series. So it going out of stock when supply constraints are affecting literally everyone is to be expected, and if they aren’t showing up on a specific retailer, it’s not because they are EOL; it’s simply ran out of stock.

We see this happen even during the launch period, where a certain supply is allocated for the first few weeks, and retailers have to put up “Out of Stock” once the supply dries up. That doesn’t mean that GPU makers are EOL’ing their cards just a few days after launch. The same has happened with the Founders Edition models, which have always been limited-edition. As soon as these models are out of stock, the first thing you’ll see pop up is that NVIDIA has EOL’d the entire series.

ASUS Clarifies That No RTX 50 GPU Has Been EOL’d

And just as we were about to publish our report, ASUS issued another statement confirming what our source just said: that their representative passed on incorrect information and that none of the RTX 50 cards they mentioned during the CES tradeshow to media have been EOL’d. The full statement is as follows:

We would like to clarify recent reports regarding the ASUS GeForce RTX™ 5070 Ti and RTX™ 5060 Ti 16 GB. Certain media may have received incomplete information from an ASUS PR representative regarding these products. The GeForce RTX 5070 Ti and GeForce RTX 5060 Ti 16 GB have not been discontinued or designated as end-of-life (EOL). ASUS has no plans to stop selling these models.

Current fluctuations in supply for both products are primarily due to memory supply constraints, which have temporarily affected production output and restocking cycles. As a result, availability may appear limited in certain markets, but this should not be interpreted as a production halt or product retirement.

ASUS will continue to support the GeForce RTX 5070 Ti and RTX 5060 Ti 16 GB and is working closely with partners to stabilize supply as conditions improve.

via ASUS

Similar rumors also surfaced a few months back, claiming the RTX 5090 had been EOL’d; however, at CES, MSI, ASUS, and Gigabyte unveiled new in-production and retail-ready designs. If the card was really EOL, then why would manufacturers still produce more custom designs for it?

So that gives you an idea that most of the information being shared on the rumor mill, or by companies themselves, as big as ASUS, isn’t based on hard facts of what’s really going on in the industry right now, and that’s why we had multiple pointers and sources within this post that reflect that actual situation & what consumers should expect.

Hardware Unboxed has provided its own timeline on the matter:

With that said, that’s about what we were able to get out for now, and hopefully this report answers a few questions about the current state of NVIDIA’s RTX 50 GPUs and its gaming lineup in general, while also addressing some rumors on the web.

NVIDIA GeForce RTX 50 GPU Prices (Newegg):

NVIDIA GeForce RTX 50 GPU Prices (Amazon):

Follow Wccftech on Google to get more of our news coverage in your feeds.