RTX 2070 vs GTX 1080
The complete head-to-head: full specs, real per-game FPS with 1% lows, ray tracing, a live CPU-bottleneck check, thermals, power and value, with a straight verdict.
GeForce RTX 2070
GeForce GTX 1080
avg, 1080p
The short answer
The RTX 2070 is clearly the better card, about 19% faster on average, with RT cores, Tensor cores for DLSS, and a much wider memory pipe at 448 GB/s against the 1080's 320 GB/s. What makes this an interesting matchup is that the GTX 1080 actually has more CUDA cores and higher clock speeds on paper, 2,560 cores at up to 1733 MHz against the 2070's 2,304 cores at up to 1620 MHz. Turing's architectural gains simply outweigh the raw spec sheet. Both cards carry 8 GB of VRAM, so this one isn't close, the 2070 wins on nearly every axis that matters.
This is a case where the spec sheet lies a little. The GTX 1080 has more CUDA cores and runs them faster, yet the RTX 2070 beats it by a clear 19% on average, thanks to Turing's per-core efficiency gains over Pascal. Both cards carry the same 8 GB of VRAM, so unlike a lot of matchups on this site, memory capacity isn't the tiebreaker here, it's a genuine speed and feature win for the newer card.
From here the page runs the full head-to-head: specs, ten-game frame rates with 1% lows, Turing's ray tracing and DLSS against Pascal's raw raster, power, pricing over time, and the verdict.
At a glance
Seven dimensions, two cards. Bigger area is not automatically better, each card owns different corners.
Strengths radar
Each axis scored 0-100 relative to the pair
Performance vs price
Up and to the left is better value
Grey dots are neighbouring cards for context.
Which should you buy?
Pick your use case, the winner changes with what you actually do.
Faster across the board, plus DLSS gives it more headroom to grow into demanding titles.
Same 8 GB as the 1080 but with far more bandwidth to feed it, 448 GB/s against 320 GB/s.
Tensor cores accelerate AI workloads directly; the 1080 has no equivalent hardware at all.
Turing's NVENC adds HEVC B-frame support that Pascal's encoder in the 1080 lacks.
175 W against 180 W is close, but the 2070 edges it and adds real feature support on top.
Both carry 8 GB, so the 2070's much higher bandwidth is the deciding factor at high internal resolutions.
Already own a GTX 1080? A 2070 is a worthwhile step up.
Short answer: probably not
A 19% average gain plus RT cores and working DLSS is a real upgrade, not just a sidegrade, and you keep the same 8 GB of VRAM you're used to. If you already stream or edit video, the newer NVENC encoder is worth having on its own.
If you'd rather leapfrog Turing entirely, a used RTX 3060 gets you 12 GB of VRAM and stronger DLSS 2 performance for similar money on the used market. But if you specifically want to stay in this price bracket, the 2070 is the clean upgrade from a 1080.
Full specification comparison
Every spec side by side. The highlighted figure wins each row.
| Specification | RTX 2070 | GTX 1080 |
|---|---|---|
| Architecture | ||
| GPU / architecture | TU106-400 · Turing | GP104-400 · Pascal |
| Process node | TSMC 12nm FFN ▲ | TSMC 16nm FinFET |
| Release date | Oct 17, 2018 | May 27, 2016 |
| Launch MSRP | $499 ▲ | $599 |
| Compute | ||
| CUDA cores | 2,304 | 2,560 ▲ |
| RT cores | 36 · 1st gen ▲ | None |
| Tensor cores | 288 · 1st gen ▲ | None |
| Base / boost clock | 1410 / 1620 MHz | 1607 / 1733 MHz ▲ |
| L2 cache | 4 MB ▲ | 2 MB |
| Memory | ||
| VRAM | 8 GB GDDR6 | 8 GB GDDR5X |
| Memory bus | 256-bit | 256-bit |
| Bandwidth | 448 GB/s ▲ | 320 GB/s |
| Power & form factor | ||
| TDP | 175 W ▲ | 180 W |
| Power connector | 1× 8-pin | 1× 8-pin |
| Suggested PSU | 550 W | 550 W |
| PCIe interface | 3.0 ×16 | 3.0 ×16 |
| Features | ||
| Upscaling | DLSS 2 / Super Res + FSR ▲ | FSR / NIS only, no DLSS |
| AV1 encode (NVENC) | No | No |
| Display outputs | 3×DP1.4 · HDMI 2.0b | 3×DP1.4 · HDMI 2.0b · DVI-D |
Average FPS by game
Switch resolution, metric, ray tracing and upscaling, the bars update live. Hover a bar for the avg / 1% / 0.1% breakdown.
Average FPS · 1080p · Rasterization. Representative data, hover a bar for avg / 1% / 0.1%. Note the GTX 1080's 1% lows fall off at 1440p and up from its 8 GB VRAM.
Smoothness & frametimes
Average FPS hides stutter. This is one run plotted as frametime, a flat line is smooth, spikes are hitches. Watch what the GTX 1080's 8 GB VRAM does in a texture-heavy game.
Frametime: Starfield, 1080p Ultra
Lower and flatter is better · 16.7 ms = 60 fps
Counterintuitive but real: the slower RTX 2070 delivers a smoother line here because 12 GB holds the texture pool the 8 GB GTX 1080 keeps evicting, those spikes are VRAM stutter, not raw speed.
Lab scores
Clean apples-to-apples tests, useful, but weaker at predicting real games than the FPS above.
Ray tracing & upscaling
Ray tracing tanks both cards at native resolution. What makes it playable is upscaling, and here the GTX 1080 has a real edge.
Is frame generation "real" performance?
Neither card supports DLSS 3 Frame Generation, which needs the Optical Flow Accelerator hardware from RTX 40-series cards. The RTX 2070's DLSS is limited to Super Resolution upscaling, and the GTX 1080 predates Tensor cores entirely, so it has no DLSS at all.
AMD's FSR 3 Frame Generation works on both since it isn't hardware-locked. It's the only frame-gen path either card has, best used in single-player games and skipped in competitive titles or below roughly 50 base fps.
Will your CPU bottleneck these cards?
A faster GPU only helps if your CPU keeps up. Pick your CPU and resolution to see expected FPS and bottleneck for each card.
CPU + GPU pairing check
The willitbottleneck engine, built right into the comparison.
Estimated from a 43-CPU model across 1440p. A weak CPU mostly bites at 1080p; at higher resolutions the GPU becomes the limiter and the bottleneck shrinks.
Thermals & noise
The efficiency gap shows up here: the newer card runs cooler and quieter, day in and day out.
Real draw, efficiency & PSU
TDP is the rated figure; what matters is real gaming draw and how many frames you get per watt.
Both use a single 8-pin connector. If your PSU already ran the RTX 2070, it will run the GTX 1080 easily. Leave about 30% headroom for spikes.
Will it fit, and what to pair it with?
Size to scale
Card length matters for small cases
What monitor to pair
GTX 1080: 1080p at 100-144 Hz, or 1440p at 60 Hz on most titles. RTX 2070: 1080p at 144-165 Hz, or 1440p at 75-90 Hz with DLSS in supported games. Treat 4K as out of reach for either.
Running an older PCIe 3.0 board?
Both run a full PCIe 3.0 ×16 link, identical interfaces, so there's no difference to weigh here. Either card also works fine in a modern PCIe 4.0 or 5.0 board without penalty, since neither approaches the bandwidth ceiling of even a 3.0 ×8 link.
Price history & dollars-per-frame
Raw price is only half the story. Cost per frame is the number that decides it, and it keeps moving as prices fall.
New price over time
Street price of new cards (USD)
The 8 GB vs 8 GB question
RTX 2070 · 8 GB
- Same 8 GB as the GTX 1080, but fed by 448 GB/s of bandwidth against the 1080's 320 GB/s.
- Tensor cores make local AI tools noticeably faster than the 1080's raw compute-only approach.
- 8 GB is still 8 GB, this card won't outlast a heavier VRAM requirement any better than the 1080 will.
- 1st-gen RT cores are genuinely slow for real-time ray tracing, useful mainly as a checkbox feature.
GTX 1080 · 8 GB
- 8 GB on a 256-bit bus is a solid, proven amount for 1080p and reasonable 1440p gaming.
- Simpler architecture with fewer bottlenecks elsewhere on the chip, so the memory system isn't fighting for die space with RT or Tensor hardware.
- 320 GB/s of bandwidth is meaningfully behind the 2070's 448 GB/s, showing up first in emulation and texture-heavy scenes.
- No Tensor cores at all, so there's no DLSS fallback and AI workloads run on general compute only.
Emulation & memory bandwidth
The narrower bus is a real downgrade for emulation
This one is straightforward. Both cards carry the same 8 GB, so capacity isn't the differentiator it usually is in these matchups. What decides it is bandwidth, and the RTX 2070's 448 GB/s crushes the GTX 1080's 320 GB/s. At high internal resolutions in RPCS3 or demanding Switch emulator cores, that gap shows up directly as steadier frame pacing on the 2070.
Creator, AI & streaming
The newer card is quicker per task, but the older card's extra VRAM unlocks larger jobs it can't fit at all.
Streaming: the AV1 encoder is the differentiator
The RTX 2070 has the newer NVENC implementation, and Turing added HEVC B-frame support that Pascal's encoder in the GTX 1080 lacks, a real, visible gain in quality at the same bitrate. Neither card has an AV1 encoder, that arrived with Ada Lovelace.
Overclocking & undervolting
Pascal's GTX 1080 has real overclocking headroom, often 10-15% with a decent cooler, enough to meaningfully narrow the gap to a stock 2070. The 2070 has less room to give, typically 5-8%, so undervolting is the better use of time there.
Pros & cons
RTX 2070
- RT cores and Tensor cores for DLSS, the 1080 has neither
- About 19% faster on average, with 40% more memory bandwidth
- Newer NVENC encoder with HEVC B-frame support
- Fewer CUDA cores and lower clocks on paper than the 1080
- Costs a bit more on the used market
GTX 1080
- 8 GB VRAM on a wide 256-bit bus, same capacity as the 2070
- More CUDA cores and higher clocks out of the box
- Strong overclocking headroom, often 10-15%
- No RT cores or DLSS, stuck with FSR/NIS upscaling
- About 19% slower despite the higher core count
- Older NVENC without HEVC B-frame support
Buying pitfalls
RTX 2070 vs RTX 2070 Super, don't confuse them
The 2070 Super uses a larger die with 2,560 CUDA cores and faster GDDR6, a real step up from the plain 2070's 2,304 cores. Used listings mix them up constantly, confirm the exact model, not just '2070', before you buy.
GTX 1080 vs GTX 1080 Ti, they're a tier apart
The 1080 Ti packs 3,584 CUDA cores and 11 GB of GDDR5X, well above the plain 1080's 2,560 cores and 8 GB. They get lumped together in used listings, confirm the exact model before buying.
RTX 2070 vs GTX 1080: your questions answered
In every game tested here, yes. The margin moves a bit, from about 16% in Alan Wake 2 up to nearly 21% in Elden Ring, but the 2070 leads across the whole ten-game spread, not just in the average.
The RTX 2070, unless the GTX 1080 is priced noticeably lower where you're shopping. It's faster, adds working DLSS and RT cores, and matches the 1080's 8 GB of VRAM, so there's little upside to picking the older card at a similar price.
Both still handle 1080p high settings without much drama. The RTX 2070 has more room left for current AAA games and can lean on DLSS in supported titles, something the GTX 1080 simply can't do.
Yes, this is one of the more worthwhile upgrades covered on this site. You keep the same 8 GB of VRAM you already have, pick up roughly 19% more speed, and gain RT cores and working DLSS that the 1080 never had.
Turing simply does more work per core than Pascal does. Changes like concurrent integer and floating-point execution and a bigger, unified L2 cache mean the RTX 2070 outruns the GTX 1080 despite trailing it on paper in both core count and clock speed.
The RTX 2070, no contest, its 288 Tensor cores accelerate AI workloads in hardware. The GTX 1080 predates Tensor cores entirely and has to fall back on slower general-purpose compute for the same jobs.
The RTX 2070. Turing's NVENC block added HEVC B-frame support that the GTX 1080's older Pascal encoder lacks, which is a real, visible quality gain at matching bitrates, not just a spec-sheet difference.
None. Only the RTX 2070 has the hardware for it, 36 RT cores plus 288 Tensor cores driving DLSS Super Resolution. The GTX 1080 launched two years before either technology existed.
No, that feature is locked to RTX 40-series and newer cards with an Optical Flow Accelerator. AMD's FSR 3 Frame Generation isn't hardware-locked though, so it runs on both the 2070 and the 1080.
Not much, honestly. A Ryzen 5 3600 or Core i5-10400F and up keeps either card fed at 1080p and through most of 1440p, use the pairing checker above if you want a number specific to your CPU.
They're close on the spec sheet, 175 W for the 2070 against 180 W for the 1080, but in practice the 2070 runs noticeably cooler and quieter thanks to Turing's more efficient design.
A solid 550 W unit handles either card with room to spare. If you want headroom for a future upgrade, step up to 600 W, but 550 W is enough for either as they stand.
Not quite, the RTX 2070 is a compact 229 mm card while the reference GTX 1080 runs longer at 267 mm, so double-check clearance if you're eyeing the 1080. Both use PCIe 3.0 ×16, so there's no interface difference to worry about either way.
Yes, both do, but the mobile RTX 2070 and mobile GTX 1080 run at lower clocks and power limits than the desktop cards covered on this page, so don't expect laptop numbers to match these directly.
Not for current AAA titles. Both are built for 1080p, with the RTX 2070 stretching further into 1440p than the GTX 1080 can. Neither belongs on a 4K shopping list in 2026.
Used is really the only option left for both. A used RTX 2070 around $100 delivers better value per frame than the cheaper GTX 1080 despite costing a bit more, and it throws in real DLSS and RT support besides.
Both still sit on NVIDIA's Game Ready driver branch today. The Turing-based RTX 2070 is likely to stay supported longer simply because it's the newer architecture of the two, though neither will pick up new architecture-level features going forward.