💪 GPU Battle #34 RTX 5080 vs RTX 4090 💪
GPU Battle #34 — RTX 5080 vs RTX 4090
The RTX 5080 is NVIDIA’s 2025 upper-high-end Blackwell card — 16 GB GDDR7, DLSS 4, and a $999 MSRP. The RTX 4090 is NVIDIA’s 2022 flagship — 24 GB GDDR6X, DLSS 3, and still the second-fastest consumer GPU ever made. One is new with the latest features and a warranty; the other is discontinued and only available used. The comparison is closer than the tier gap suggests — the 4090 still wins on raw performance, but the 5080 has newer features and lower power draw.
Short version: The RTX 4090 is faster in raster and has more VRAM — it’s the better raw performance card. The RTX 5080 has DLSS 4, is more efficient, and comes with a warranty. Used 4090 prices determine the winner.
Specs
The RTX 5080 is Blackwell on TSMC 4N — 10,752 CUDA cores, 16 GB GDDR7 on a 256-bit bus, 360W TGP, PCIe 5.0. It supports DLSS 4 with multi-frame generation, and its RT hardware is fourth-gen.
The RTX 4090 is Ada Lovelace on TSMC 4N — 16,384 CUDA cores, 24 GB GDDR6X on a 384-bit bus, 450W TGP, PCIe 4.0. It supports DLSS 3 with single-frame generation, and its RT hardware is third-gen.
| Spec | RTX 5080 | RTX 4090 |
|---|---|---|
| Architecture | Blackwell | Ada Lovelace |
| CUDA cores | 10,752 | 16,384 |
| VRAM | 16 GB GDDR7 | 24 GB GDDR6X |
| Bus | 256-bit | 384-bit |
| TGP | 360W | 450W |
| PCIe | 5.0 | 4.0 |
| Upscaling | DLSS 4 (multi-frame) | DLSS 3 (single-frame) |
| RT hardware | 4th gen | 3rd gen |
| Launch price | $999 | $1,599 |
The 4090 has 52% more CUDA cores and 50% more VRAM bandwidth. The 5080 has newer architecture, GDDR7, and DLSS 4. On paper, the 4090 is the more powerful card.
Rasterization
In raster, the RTX 4090 is roughly 20–30% faster than the RTX 5080 at 4K. Its wider core count and higher memory bandwidth give it a clear lead. At 1440p the gap is smaller but still favors the 4090.
The 5080 compensates with architectural improvements and faster GDDR7 memory, but it doesn’t close the gap. For pure raster, the 4090 wins.
Ray tracing
Ray tracing is closer. The 5080’s fourth-gen RT cores are more efficient than Ada’s third-gen, and the 5080 lands within 5–15% of the 4090 in RT-heavy titles. In some games with DLSS 4’s multi-frame generation, the 5080 can feel smoother despite lower raw performance.
The 4090 remains a strong RT card — it was the best when it launched — but Blackwell has caught up significantly.
For raw RT, the 4090 is slightly ahead. For perceived smoothness with DLSS 4, the 5080 can match or exceed it.
VRAM and memory
This is a real difference. The 4090 has 24 GB; the 5080 has 16 GB. At 4K with high textures and ray tracing, the 4090’s extra VRAM provides headroom the 5080 lacks. In a few demanding titles, the 5080 can run into VRAM limits at 4K max settings.
For AI workloads, the 4090’s 24 GB is a significant advantage — it can run larger models. The 5080’s 16 GB is limiting for serious AI work.
The 4090’s memory is also faster in bandwidth — ~1,008 GB/s versus ~960 GB/s on the 5080. Not a huge difference, but the 4090’s wider bus gives it an edge.
Power and thermals
The 5080 draws 360W; the 4090 draws 450W. The 5080 is significantly more efficient, runs cooler, and needs a less powerful PSU. For SFF builds or quieter systems, the 5080 is easier to live with.
Both are high-draw cards and need good cooling, but the 5080 is the more reasonable option.
Features
The 5080 supports DLSS 4 with multi-frame generation, DisplayPort 2.1b, and newer NVENC encoders. DLSS 4 can generate up to three additional frames per rendered frame, dramatically improving perceived smoothness in supported games.
The 4090 supports DLSS 3 with single-frame generation — still excellent, but a generation behind. It lacks DisplayPort 2.1 and newer media features.
For features, the 5080 is ahead.
Price and value
New at $999 (if you can find it at MSRP), the 5080 is the more affordable card. But availability has been tight, and street prices often exceed MSRP.
Used at $1,400–1,700, the 4090 is expensive — often more than a new 5080. Its resale value has stayed high because it’s still one of the fastest cards available and has 24 GB of VRAM.
For pure value, the 5080 wins at MSRP. But at current street prices, the two are close in cost, and the 4090’s extra performance and VRAM may justify the premium for some buyers.
The verdict
The RTX 4090 is the faster card in raw performance — 20–30% ahead in raster, with more VRAM and memory bandwidth. For 4K gaming at max settings, AI workloads, or content creation, it’s the better card.
The RTX 5080 is the better new card — DLSS 4, lower power draw, and a warranty at a lower MSRP. For most gamers, its performance is more than enough, and the newer features are valuable.
The decision comes down to availability and price. At MSRP, the 5080 is the better buy. If the 4090 is available used near MSRP, its extra performance and VRAM make it compelling.
Short version:
Choose the RTX 5080 if you’re buying new, want DLSS 4, lower power draw, and a warranty at MSRP.
Choose the RTX 4090 if you want maximum raster performance, 24 GB of VRAM for AI or 4K, and can find it near MSRP used.
Stop using slow, ad-bloated tool sites! 🤮
🔎 Search “KandZ Tools” on Google to use many professional utilities for free.
KandZ.me is the ultimate minimalist hub for:
✅ Finance (Mortgage, Interest, Inflation)
✅ Tech (Base64, JSON, Dev Suite, IP)
✅ Health (BMI, BMR, TDEE)
✅ Productivity (Timer, Workspace, QR)
⚡️ Fast & Private
🔒 No data leaves your device
💎 100% Free
🔗 Use it now: https://tools.kandz.me
🔖 Bookmark it—you’ll need it later!