Skip to content

NVIDIA set to raise prices: Graphics cards will become significantly more expensive in 2026.

Young man comparing two graphics cards in a computer hardware store with PC components displayed behind him.

PC gamers briefly thought the coast was clear, but that window is closing: manufacturers are warning of another wave of graphics card price rises.

For months, graphics cards stayed surprisingly steady even as laptops, RAM and other hardware became noticeably more expensive. That calm is now fading. NVIDIA has formally briefed its partners that graphics memory costs are increasing - and that shift is pushing the entire market into the next round of price changes.

What NVIDIA has reportedly told partners internally about GDDR6/GDDR7

This latest trigger comes from an internal notice that was leaked to the industry outlet Benchlife. The message focuses on expiring supply agreements for memory chips - specifically GDDR6 and GDDR7 - used on current and upcoming gaming GPUs.

"NVIDIA is keeping the official price of the GPU chips stable, but is budgeting significantly higher purchasing costs for the accompanying graphics-memory population."

That may sound like a minor technical detail, but it feeds straight through to the price you see at retailers. Memory is now one of the most expensive single components on a modern graphics card. The more gigabytes of VRAM a card carries, the more sharply this revised cost base is felt.

What makes it especially sensitive is a long-running internal idea: shipping GPUs more often as “bare” chips. In that scenario, ASUS, MSI, Gigabyte and other board partners would need to source the memory entirely themselves - at market rates and without preferential terms. That would shift even more risk and volatility onto the add-in-board (AIB) partners.

Why memory costs are spiking right now

The underlying cause sits in the wider memory market. DRAM and GDDR manufacturers have tuned output and investment over recent years to match demand. At the same time, workloads such as AI, data centres and high-end gaming are all pushing towards larger capacities and faster memory.

  • Production costs for modern memory chips are rising markedly.
  • Contracts with older, cheaper terms expire at the start of 2026.
  • Manufacturers are optimising capacity towards higher-margin segments such as servers.

As a result, laptops with lots of RAM already cost up to 20 percent more, and standard RAM kits are clearly more expensive than they were a year ago. Graphics cards had so far lagged behind this pattern - stock made with “old” memory was simply sold down over time. That buffer is now running out.

Board partners start turning the price dial

Taiwan’s Commercial Times reports that board partners see very little room left to absorb the increases. MSI already moved first in December with higher pricing on new RTX-50 models. Gigabyte and ASUS, according to the report, are preparing similar adjustments.

"Once the old inventory has been sold, manufacturers will price using the new, more expensive memory baseline - and that shows up immediately on retail price tags."

In Europe, the direction is already visible: cards with 16 GB of VRAM and more are climbing fastest. Many shops are showing mark-ups between 15 and 20 percent. This is not limited to NVIDIA models; AMD’s Radeon line-up is also sliding into the new price band.

For the Radeon RX 9000 range, retailers are reporting jumps of 10 to 18 percent - depending on the model and memory configuration. That makes a cheap entry into the upper tier increasingly hard to find.

NVIDIA cuts memory, AMD pushes full performance: how GPU strategies are changing

These conditions are forcing manufacturers into clearer strategic choices. More memory means higher bill-of-materials costs - and that pushes products into price territory many gamers are unwilling to follow.

NVIDIA partners: less VRAM, lower entry prices

On the NVIDIA side, partners are responding mainly by placing greater emphasis on 8 GB models. In the mainstream bracket - around series such as the RTX 5060 and RTX 5060 Ti 8 GB - that capacity is returning to centre stage.

  • 8 GB cards remain financially within reach for many players.
  • Lower memory requirements reduce the risk of being stuck with costly inventory.
  • Marketing shifts back towards “good enough for 1080p” rather than “maxed out at 4K”.

That can work for casual gamers who mainly play at Full HD. But anyone targeting higher texture quality, ray tracing, and 1440p or 4K will run into 8 GB limits more often - including stutter, streaming hitches and having to drop settings.

AMD: more VRAM as a selling point - with added risk

AMD is taking a different line. The “Reds” are still pushing 16 GB VRAM cards, particularly XT models in the higher-performance segment. On paper, it’s an appealing proposition: more memory, better longevity, and stronger results in memory-hungry titles.

The downside is that these models are hit hardest in percentage terms by rising memory prices. A card with double the memory population doesn’t just cost a bit more to build - it costs materially more, and that ultimately ends up with the buyer.

How hard this will really hit PC gamers

A common question is whether this means every graphics card will become 20 percent more expensive. It’s unlikely to be that severe across every segment. The sharpest increases should concentrate on newly introduced or freshly produced models, especially those with higher VRAM capacities.

Segment Typical VRAM size Current trend
Entry-level 6–8 GB slight to moderate mark-up
Mid-range 8–12 GB clearly noticeable rise, especially on refreshes
Upper mid-range / High-end 16–24 GB in part 15–20 % more, depending on model

If you already own a solid card, there’s no need to panic. Many current GPUs remain well suited to 1080p and 1440p gaming for the next few years. The pinch point is for people who have been saving specifically for an upgrade and are now finding their target card has moved into a distinctly higher price bracket.

When buying still makes sense - and what to watch

The question of the “right” moment to buy becomes more complicated. Traditional advice such as “wait for the next generation” only applies to a limited extent here, because new series will be priced from the outset around higher memory costs.

A few straightforward guidelines can help:

  • Current deals on high-VRAM cards from “older” production may be good value while retailers clear remaining stock.
  • If you mainly play at Full HD, choosing 8 GB models deliberately can save money.
  • For 1440p and 4K, more VRAM remains critical - cutting back here often means paying later through reduced detail or lower frame rates.
  • Track pricing over time: when demand softens, retailers discount even if their own costs have risen.

One more factor is often overlooked: game engines vary widely in memory efficiency. Competitive shooters typically cope with less VRAM, while modern open-world games and demanding single-player AAA releases can quickly fill double-digit gigabyte ranges. If you mostly play esports titles, you can more realistically save on memory.

Why VRAM now plays such a big role in future-proofing

VRAM (video RAM) is the memory located directly on the graphics card where textures, geometry data and intermediate results are stored. Unlike system RAM, it cannot be upgraded later. The capacity is fixed to the card.

The higher the resolution, texture settings and ray tracing load, the more VRAM is required. Once VRAM is full, the GPU is forced onto much slower paths such as system RAM or even the SSD. The practical result is micro-stutter, frame-rate drops and messy streaming artefacts.

Against that backdrop, rising memory prices have a double effect: they make specific models more expensive, and they also push more gamers towards configurations with tighter VRAM limits - meaning systems that will hit those limits sooner.

If you’re planning a new graphics card purchase in the next one to two years, it makes sense to focus less on raw teraflop figures or marketing labels and to scrutinise memory configuration closely. In a period where each extra gigabyte costs noticeably more, VRAM can be the deciding factor between a card that remains enjoyable for three years and one that becomes frustrating much sooner.

Comments

No comments yet. Be the first to comment!

Leave a Comment