RTX 3090 vs RTX 4080
The RTX 3090 and the RTX 4080 sit side by side in the same lineup, built on a shared foundation but aimed at slightly different buyers. By our tally, the RTX 4080 leads 2 of 3 areas. The gpu & performance section is where it pulls furthest ahead of the RTX 3090. It also matters that $300 separates their launch prices, with the RTX 3090 at the top.
At a glance
- GPU & Performance RTX 4080 wins
- Memory RTX 3090 wins
- Power & Thermals RTX 4080 wins
- Price RTX 4080 wins
- GPU & Performance2.4
- Memory4.8
- Power & Thermals7.0
- GPU & Performance3.9
- Memory1.5
- Power & Thermals7.6
Spec scores (0 to 10) are calculated from manufacturer specs against other models in the same class. See how the formula works. They are not reviews or user ratings.
Key Differences
-
The RTX 3090 has 8 GB more VRAM
The RTX 3090 clearly leads on vram, and 24 GB versus 16 GB is a difference that reaches well beyond the spec sheet. At high resolutions or with texture packs, running out of VRAM can hurt more than a small speed gap.
-
The RTX 4080 has 37% more theoretical FP32 compute
On fp32 compute, the RTX 4080 posts 48.7 TFLOPS against 35.6 TFLOPS for the RTX 3090, a gap big enough to feel in everyday use. Architectures use their compute differently, so treat the gap as a guide and look at game benchmarks for the details.
-
The RTX 3090 has 31% more memory bandwidth
This is one of the widest gaps in the whole comparison: 936 GB/s for the RTX 3090 versus 717 GB/s for the RTX 4080. The RTX 3090 has more bandwidth to work with, helping it hold up in memory-heavy games.
-
The RTX 4080 has 30 W lower board power
The RTX 4080 has a modest edge on total board power: 320 W against 350 W for the RTX 3090. Higher board power often comes with higher performance, so the RTX 3090’s appetite may be part of the price of its speed.
-
The RTX 3090 has 7.9% more shader units
It’s a small lead (10,496 versus 9,728), but it’s real, and it goes to the RTX 3090. Shader units do most of the heavy lifting in games, so more of them usually means more frames per second within the same architecture.
Design & Size
- RTX 3090 12.32 in × 3 slots
- RTX 4080 11.97 in × 3 slots
The RTX 3090 is physically larger, so the RTX 4080 will be the easier fit wherever space is limited.
Size & Connectivity
| Spec | RTX 3090 | RTX 4080 |
|---|---|---|
| Reference card length | 12.32 in | 11.97 in |
| Slot width | 3 slots | 3 slots |
| Display outputs | 3x DisplayPort 1.4a, 1x HDMI 2.1 | 3x DisplayPort 1.4a, 1x HDMI 2.1 |
GPU & Performance
RTX 4080 leads
Few areas are as clear-cut as this one: the RTX 4080 leads the gpu & performance section with room to spare. The RTX 4080 takes this one with 37% more theoretical FP32 compute. FP32 compute is a theoretical peak that combines shader count and clock speed, a quick way to size up raw power. The RTX 3090 has the edge with 7.9% more shader units. For gpu & performance, the RTX 4080 is the one to beat.
FP32 compute
RTX 3090: 35.6 TFLOPS RTX 4080: 48.7 TFLOPS The RTX 4080 has more theoretical compute, which often, but not always, translates into higher frame rates.
Shader units
RTX 3090: 10,496 RTX 4080: 9,728 The RTX 3090’s extra shaders give it more raw muscle, although comparisons across different architectures aren’t one-to-one.
Memory
RTX 3090 leads
When it comes to memory, the RTX 3090 is comfortably ahead. The RTX 3090 adds 8 GB more VRAM. More VRAM lets the card hold higher-resolution textures and bigger scenes without stutter. If memory is your priority, the RTX 3090 is the card to pick.
VRAM
RTX 3090: 24 GB RTX 4080: 16 GB Newer games tend to ask for more video memory, so the RTX 3090’s extra VRAM helps the card age better.
Memory bandwidth
RTX 3090: 936 GB/s RTX 4080: 717 GB/s Bandwidth needs vary by architecture, so compare it alongside VRAM and cache rather than on its own.
Features
The features entries below are descriptive rather than numeric. Read them side by side to see where the RTX 3090 and the RTX 4080 part ways.
Power & Thermals
RTX 4080 leads
On power & thermals, the difference is modest, and it favors the RTX 4080. On the RTX 4080, you get 30 W lower board power. The RTX 4080 draws less power, which makes it easier to cool in compact cases. On power & thermals alone, the RTX 4080 gets the nod.
Total board power
RTX 3090: 350 W RTX 4080: 320 W Lower board power means less heat, quieter cooling and a smaller impact on your electricity bill.
Full Specifications
GPU & Performance
| Spec | RTX 3090 | RTX 4080 |
|---|---|---|
| Architecture | Ampere | Ada Lovelace |
| GPU die | GA102 | AD103 |
| Process | Samsung 8N | TSMC 4N |
| Shader units | 10,496 (winner) Fewest shader units | 9,728 |
| Ray tracing units | 82 (winner) Fewest ray tracing units | 76 |
| Base clock | 1,395 MHz Lowest base clock | 2,205 MHz (winner) |
| Boost clock | 1,695 MHz Lowest boost clock | 2,505 MHz (winner) |
| FP32 computeTheoretical peak; AMD counts dual-issue, so compare within a brand only | 35.6 TFLOPS Least compute in class | 48.7 TFLOPS (winner) |
Memory
| Spec | RTX 3090 | RTX 4080 |
|---|---|---|
| VRAM | 24 GB (winner) | 16 GB |
| Memory type | GDDR6X | GDDR6X |
| Memory bus | 384-bit (winner) | 256-bit |
| Memory bandwidth | 936 GB/s (winner) Least memory bandwidth | 717 GB/s |
Features
| Spec | RTX 3090 | RTX 4080 |
|---|---|---|
| Upscaling | DLSS Super Resolution | DLSS Super Resolution |
| Frame generation | No DLSS Frame Generation | DLSS Frame Generation (2x) |
| PCIe interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
Power & Thermals
| Spec | RTX 3090 | RTX 4080 |
|---|---|---|
| Total board power | 350 W Lowest power draw in class | 320 W (winner) |
| Recommended PSU | 750 W Smallest PSU requirement | 750 W |
| Power connector | 1x 12-pin | 1x 16-pin (12VHPWR) |
Size & Connectivity
| Spec | RTX 3090 | RTX 4080 |
|---|---|---|
| Reference card length | 12.32 in | 11.97 in |
| Slot width | 3 slots | 3 slots |
| Display outputs | 3x DisplayPort 1.4a, 1x HDMI 2.1 | 3x DisplayPort 1.4a, 1x HDMI 2.1 |
Price
| Spec | RTX 3090 | RTX 4080 |
|---|---|---|
| Launch price (MSRP) | $1,499 Lowest price in class | $1,199 (winner) |
“N/A” means the manufacturer does not specify the value or we could not confirm it. Figures marked “manufacturer rating” are the maker’s own claims, not our measurements.
Price & Value
At launch, the RTX 3090 cost $300 more than the RTX 4080. In percentage terms, that’s 25%. The RTX 3090’s extra $300 is hard to justify when the RTX 4080 leads in more areas.
Launch prices (US) as of November 2022. Discounts and bundles come and go, so compare current prices before buying.
Which One Should You Buy?
Choose the RTX 3090 if…
- you want more video memory for the years ahead (24 GB vs 16 GB)
- you want more memory bandwidth (936 GB/s vs 717 GB/s)
- you want more shader units working on every frame (10,496 vs 9,728)
- you push high resolutions where memory width helps (384-bit vs 256-bit bus)
Choose the RTX 4080 if…
- you run compute-heavy work alongside gaming (48.7 TFLOPS vs 35.6 TFLOPS FP32)
- you care about your power bill and a quieter case (320 W vs 350 W)
- you care about peak clock speed (2,505 MHz vs 1,695 MHz boost)
- you care about clock speed under sustained load (2,205 MHz vs 1,395 MHz base)
- you want the lower price and can live with the trade-offs ($300 less)
Also Compared
Frequently Asked Questions
Which is better for most people, the RTX 3090 or the RTX 4080?
We’d point most buyers to the RTX 4080. The main exception is if having 8 GB more VRAM matters more to you than the rest.
Should I upgrade from the RTX 3090 to the RTX 4080?
For most owners, it’s a meaningful step. The RTX 4080 brings 37% more theoretical FP32 compute, 30 W lower board power and a 48% higher boost clock over the RTX 3090.
Which is larger, the RTX 3090 or the RTX 4080?
The RTX 4080 is actually the more compact one. It measures 11.97 in, while the RTX 3090 comes in at 12.32 in.
Which is faster, the RTX 3090 or the RTX 4080?
The RTX 4080 has the edge in performance thanks to 37% more theoretical FP32 compute and a 48% higher boost clock.
Which costs less, the RTX 3090 or the RTX 4080?
The RTX 3090 launched at $300 more than the RTX 4080, based on prices as of November 2022.
Executive Summary
The RTX 3090 and the RTX 4080 come from the same family, and honestly, that makes this one harder than a cross-brand fight. For me, the deciding factor is that only the RTX 4080 gives you 37% more theoretical FP32 compute. It also helps that the RTX 4080 has 30 W lower board power. On the other side, the RTX 3090 brings 8 GB more VRAM, and I wouldn’t dismiss that.
Put simply, the RTX 3090 suits people who care about memory, while the RTX 4080 suits people who put gpu & performance first.
My pick: the RTX 4080, unless you’d rather have 8 GB more VRAM on the RTX 3090, in which case the RTX 3090 earns its place.
Personal opinion based on the data above. Your priorities may differ.