NVIDIA is reportedly preparing a new graphics card, the GeForce RTX 5090 SE, that would slot into its lineup between the existing RTX 5080 and the top-tier RTX 5090, according to TweakTown.
The card is said to be built on NVIDIA's Blackwell graphics architecture. According to TechSpot, it would use a cut-down version of the GB202 die — the same underlying silicon found in the flagship RTX 5090. In practice, that means a chip closely related to NVIDIA's most powerful consumer GPU, but with some of its capabilities trimmed back.
The headline appeal is price. TechSpot reports the RTX 5090 SE would promise "flagship power for $500 less" than the full RTX 5090 — a meaningful gap for buyers who want near-top-tier performance without paying the premium for NVIDIA's most expensive card.
There's a notable catch. TechSpot frames the deal as conditional on availability, with the pointed qualifier "if you can find one." That echoes a recurring frustration in the GPU market, where high-demand cards can be hard to buy at their listed prices.
It's important to stress that these details remain unconfirmed. Both reports describe the RTX 5090 SE as rumored and "reportedly" in the works; NVIDIA has not, based on these sources, officially announced the card, its final specifications, or a price.
Why it matters: for anyone shopping for a high-end graphics card — gamers, creators and increasingly AI enthusiasts — a cheaper card built on the same silicon as NVIDIA's flagship could reshape the value calculation at the top of the market, provided the rumored pricing and supply hold up.