Nvidia RTX 4090 review: great 4K performance, DLSS 3 transforms frame rates, and 24GB of VRAM, but a 450W power draw, a dongle adapter, huge, and expensive
Tom Warren / The Verge :
Context & Ripple Effects
Nvidia’s preceding RTX 3080 Ti review had already established a high-end formula of strong 1440p/4K gaming, DLSS benefits, premium pricing, heat and an awkward adapter. The RTX 4090 advances the performance side with DLSS 3 while making the physical and power costs of that tier more explicit.
The subsequent RTX 4080 review repeats the combination of 4K capability, DLSS 3, large hardware and an annoying adapter, indicating that these were stack-wide trade-offs rather than isolated 4090 quirks.
First-order effects
- High-end PC buyers gain substantially stronger 4K performance and DLSS 3 frame generation, but must accommodate Nvidia’s 450W draw, oversized card and adapter.
- Nvidia positions 24GB of VRAM and DLSS 3 as differentiators for its flagship while accepting a narrower audience willing to pay for and house the hardware.
Second-order effects
- Nvidia’s RTX 4080 inherits the same DLSS 3 and adapter-led proposition, leaving buyers to distinguish the cards primarily by performance and memory rather than by installation convenience.
- The later RTX 4070’s $599 1440p positioning creates a clearer alternative for buyers who want the RTX 40-series feature set without making 4K performance the purchase criterion.
Third-order effects
- Nvidia’s consumer GPU range is moving toward sharper tiering: flagship 4K performance and larger memory at the top, versus more accessible cards optimized around 1440p.
- If DLSS continues to improve across generations, software-assisted frame rates become a more durable product differentiator alongside raw GPU performance, memory and power limits.
The trend: High-end graphics cards are increasingly sold as a trade-off between AI-assisted performance gains and the cost, memory, power and physical demands of flagship hardware.