Nvidia's RTX 50 Super Goes Missing as Memory Shortage Freezes the GPU Market

The refresh did not appear at CES, board partners were told it is delayed, and gamers face higher prices on the cards that already exist.

Key takeaways

  • Nvidia reportedly told partners the RTX 50 Super refresh is delayed; it was absent from CES 2026.
  • Reported cause: scarce 3GB GDDR7 modules as DRAM goes to AI customers.
  • Radeon RX 9070 reportedly rose to $599, and AMD is not expected to launch a new consumer GPU in 2026.

The graphics cards PC gamers expected at CES 2026 never showed up. Nvidia's RTX 50 Super refresh, which was meant to put more VRAM on mid-range and high-end GeForce cards, has reportedly been pushed back, and the reason is the same memory shortage that is lifting prices across consumer hardware.

What was reported

Several outlets, including Guru3D and TweakTown, reported that Nvidia told its add-in-board partners the Super refresh was delayed, in some accounts indefinitely. The Super line was expected to use higher-capacity 3GB GDDR7 modules. Those are the parts in shortest supply as memory makers prioritize high-bandwidth memory for AI customers.

Reports about timing disagree. Igor's Lab said the working estimate was Q3 2026 while noting that cancellation was not the plan; other, less reliable outlets claimed 2027. Nvidia has not announced a date or prices, so treat all of this as sourced leaks.

Timeline of RTX 50 Super expectations CES, Jan 2026 Super absent Partners told refresh delayed Q3 2026 earlier estimate (Igor's Lab) 2027? lower-quality reports
Based on reports from Guru3D, TweakTown and Igor's Lab; no dates are official. Illustrative.

Why it matters

A Super refresh typically brings a price reset: new cards arrive and older ones drop. Without it, shoppers are stuck with 8GB and 12GB cards at inflated prices. GamesRadar noted the bigger problem was existing prices, which have risen as board makers pass on memory costs. The RX 9070 reportedly went from $549 to $599.

AMD is not a pressure valve either. Several outlets report it has no new consumer GPU planned for 2026, and Notebookcheck observed that the 16GB RX 9070 XT and RX 9060 XT are about the only affordable 16GB cards. With competition quiet, Nvidia has little reason to rush a launch that would consume scarce memory.

RX 9070 reported price before and after increase Radeon RX 9070 price (USD) Before 549 Reported now 599 +$50 reported by GamesRadar; before-price derived from that difference
Bars start at zero. Source: GamesRadar.

Why 3GB modules matter

The Super refresh was largely a memory story. Current GeForce cards are constrained by how much VRAM fits on a board, and higher-density GDDR7 modules would let Nvidia raise capacities without redesigning the memory bus: a card with a given number of memory chips can hold half as much again if each chip is 3GB instead of 2GB. That would address the most common complaint about mid-range cards, that eight or twelve gigabytes ages poorly in modern games. It also means the refresh depends entirely on a part that memory makers have little incentive to prioritize while data-center customers are paying more for the same wafer capacity.

Nvidia's own incentives point the same way. Every gigabyte of memory allocated to a gaming card is a gigabyte not available for its far more profitable AI products. Reports described the company prioritizing AI during the shortage, and some accounts suggested the next generation, RTX 60, could also slip. Nvidia has not confirmed either claim.

What gamers can do

Buyers face an awkward choice. Waiting for a Super card is a bet on a launch that no one has dated, while the cards on shelves have gone up in price, and board makers have warned of further increases. Several outlets have pointed out that AMD's 16GB Radeon cards are among the few well-priced options with generous memory, although even those are creeping upward. Anyone with a working card from the previous generation may find that holding on is the cheapest strategy, since a used GPU market tends to firm up when new supply is constrained.

The deeper lesson is that GPU launches are now gated by memory supply as much as by silicon. A year ago, a delay would have meant a problem with the chip or with demand for it. Now it can mean that the chip is ready and the memory is not. That shifts how the industry should be read: announcements and leaks about new cards say less about what will be available than the price of DRAM does.

What to watch

  • Any official Nvidia statement on a Super or RTX 60 timeline.
  • DRAM contract prices, which decide whether higher-capacity GDDR7 is viable.
  • AMD's RDNA 5 plans; one preview source pointed to an October date, unconfirmed.
  • Used-GPU prices, a likely pressure release for budget buyers.

Sources

  1. Report: NVIDIA told AIBs RTX 50 SUPER refresh is delayed — Guru3D
  2. GeForce RTX 50 SUPER Series 'delayed indefinitely' as NVIDIA informs its partners — TweakTown
  3. Nvidia GeForce RTX 50 Super allegedly postponed to Q3 2026 — Igor's Lab
  4. Nvidia has reportedly shelved its RTX 50 series Super plans — GamesRadar