As we’re all coming to understand, AI takes precedence over human consumers. The latest casualty, caused by a focus on feeding AI datacenters with as much memory as the world’s manufacturers can supply, is NVIDIA’s RTX 5070 Ti graphics card.
Yes, the 16GB Blackwell GPU that was one of the more popular options from the new lineup — thanks mainly to its ample memory allocation and its stellar performance — is seemingly on its last legs as NVIDIA’s AIB partners (aka the companies that build GPUs around NVIDIA’s chips) struggle to receive a supply of the GPU.
Hardware Unboxed notes that retailers have not been explicitly warned that the 5070 Ti has hit end-of-life status. However, speaking with those same retailers, it’s clear that there is no stock from AIB partners to buy and sell.
This news aligns with December rumors I covered, suggesting that NVIDIA would cut RTX 50-series production in 2026 to cope with memory shortages. At the time, it looked like production might shrink by 30-40%, but it could be a lot worse.
So, the headline story is that NVIDIA has essentially killed off the GeForce RTX 5070 Ti from the market. ASUS, the largest NVIDIA AIB partner, explicitly told us this model is currently facing a supply shortage, and as such, they have placed the model into end-of-life status. This means ASUS has no plans to produce any more 5070 Ti models from this point forward. What is currently on store shelves is it from them; no more production of that card. Their GPU division is shifting focus towards other models they expect will continue to be supplied.
Hardware Unboxed
New, more recent rumors stemming from Board Channels and picked up by Videocardz seemed to indicate that NVIDIA had adjusted its supply chain to focus on 8GB versions of its RTX 50-series GPUs, including the RTX 5060 and RTX 5060 Ti. In an update, it appears that this rumor is only partly true.
According to HKEPC, a Hong Kong-based outlet, NVIDIA is now prioritizing its GPU production based on memory allocations. For example, out of the 8GB cards, NVIDIA will focus on the RTX 5060 Ti. For 16GB cards, it will focus on the RTX 5080. There’s no information yet about the 12GB RTX 5070, but the RTX 5090 will continue production.
This new prioritization effectively leaves the 16GB RTX 5060 Ti and RTX 5070 Ti behind. If you currently have one of these GPUs in your PC, I probably don’t need to tell you to protect it at all costs. There’s a lot of discussion about VRAM and how much you need to game effectively in 2026, but it essentially boils down to “more is better.”
The RTX 5070 Ti, which launched with a $749 MSRP, is indeed seeing price spikes at retailers. Some versions are still available at Newegg for as low as $829.99, but the majority of options have climbed to $1,250 and well beyond.
The 16GB RTX 5060 Ti, which launched at a $429 MSRP, is facing a similar predicament. Some models are still available for as low as $489.99 at Newegg, but the majority have hit or surpassed $530.
NVIDIA’s Blackwell SUPER cards have been postponed indefinitely
Traditionally, SUPER variants of NVIDIA’s cards are released about a year after the initial cards’ launch. Rumors began in November 2025 of memory shortages killing the RTX 50 SUPER cards, and Hardware Unboxed has now confirmed that the SUPER cards have been postponed indefinitely.
Should memory prices calm down later this year, which isn’t likely, NVIDIA could revisit its next-gen products that were supposed to debut at CES 2026.
It’s worth pointing out that NVIDIA’s laptop GPUs have seemingly not yet suffered the same prioritization fate, but I expect that 2026 will see a lot of low- and high-end models rather than a mid-range spread.
NVIDIA focusing on the costly RTX 5080 as its only 16GB GPU is surely going to leave a lot of PC gamers in the lurch. Are you ready to make the jump to AMD or Intel? Let me know in the comments section below!
Follow Windows Central on Google News to keep our latest news, insights, and features at the top of your feeds!
