GeForce RTX 3070 vs GeForce RTX 3090: specs and benchmark comparison
BNVIDIA · Ampere
NVIDIA GeForce RTX 3090
- 24 GB GDDR6X
- 350 W
- PSU 750 W+
September 24, 2020 · $1,499 at launch
Short verdict
NVIDIA GeForce RTX 3090 leads in every scenario with comparable benchmarks: rendering and compute. This does not imply an advantage in tasks without benchmarks.
Compare with AIRendering and compute
B +95.2%NVIDIA GeForce RTX 3090 leads in this scenario (up to 95.2% in comparable benchmarks).
No comparable benchmarks yet: gaming: rasterization, gaming: ray tracing, power efficiency.
Advantages
Only measured and rated values: a noticeable benchmark gap and specs where more is better.
Where GeForce RTX 3070 is better2
Board power 130 W lower220 W (board power) vs 350 W (board power)
Launch price $1,000 lower$499 vs $1,499
Where GeForce RTX 3090 is better16
Blender Benchmark 2.93.1 56.6% lower17 s vs 27 s
Blender Benchmark 2.93.1 76.3% lower9 s vs 16 s
Blender Benchmark 2.93.1 43.5% lower80 s vs 114 s
Blender Benchmark 2.93.1 59.4% lower33 s vs 53 s
Blender Benchmark 2.93.1 62.8% lower34 s vs 56 s
Blender Benchmark 2.93.1 81.9% lower18 s vs 33 s
Blender Benchmark 2.93.1 69.8% lower55 s vs 93 s
Blender Benchmark 2.93.1 76.6% lower38 s vs 67 s
Blender Benchmark 2.93.1 55.3% lower99 s vs 154 s
Blender Benchmark 2.93.1 65.1% lower45 s vs 74 s
Blender Benchmark 2.93.1 63.2% lower139 s vs 226 s
Blender Benchmark 2.93.1 68.3% lower60 s vs 102 s
Blender Benchmark 4.5.0 69.6% faster5,424 samples/min vs 3,198 samples/min
Shader units 4,608 more10,496 (CUDA cores) vs 5,888 (CUDA cores)
Video memory 16 GB more24 GB vs 8 GB
Memory bus 128 bit more384 bit vs 256 bit
Benchmarks
Results are compared within one benchmark, one version and the same conditions.
- B ahead by 56.6%
Blender Benchmark 2.93.1
GPU, bmw27 (CUDA) · s, lower is better
GeForce RTX 307026.5 sGeForce RTX 309016.9 sPosition among tested models (31, CUDA)
B · 17A · 270100200300Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 56.6% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
26.5 s · community measurement, sample 151 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 151 submissions (37 outliers excluded), interquartile range 26.29–26.91; submissions 2021-07-11 to 2022-02-26; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
16.9 s · community measurement, sample 279 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 279 submissions (58 outliers excluded), interquartile range 16.76–17.19; submissions 2021-07-09 to 2022-03-01; CUDA backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 76.3%
Blender Benchmark 2.93.1
GPU, bmw27 (OPTIX) · s, lower is better
GeForce RTX 307016.4 sGeForce RTX 30909.3 sPosition among tested models (20, OptiX)
B · 9A · 16050100150Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 76.3% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
16.4 s · community measurement, sample 98 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 98 submissions (13 outliers excluded), interquartile range 16.23–16.71; submissions 2021-07-19 to 2022-03-01; OPTIX backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
9.3 s · community measurement, sample 244 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 244 submissions (56 outliers excluded), interquartile range 9.17–9.43; submissions 2021-07-10 to 2022-02-28; OPTIX backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 43.5%
Blender Benchmark 2.93.1
GPU, classroom (CUDA) · s, lower is better
GeForce RTX 3070114.1 sGeForce RTX 309079.5 sPosition among tested models (26, CUDA)
B · 80A · 1140200400600800Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 43.5% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
114.1 s · community measurement, sample 134 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 134 submissions (26 outliers excluded), interquartile range 113.03–116.56; submissions 2021-07-11 to 2022-02-25; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
79.5 s · community measurement, sample 260 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 260 submissions (43 outliers excluded), interquartile range 78.58–81; submissions 2021-07-09 to 2022-03-01; CUDA backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 59.4%
Blender Benchmark 2.93.1
GPU, classroom (OPTIX) · s, lower is better
GeForce RTX 307053.4 sGeForce RTX 309033.5 sPosition among tested models (14, OptiX)
B · 33A · 53050100150200Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 59.4% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
53.4 s · community measurement, sample 101 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 101 submissions (8 outliers excluded), interquartile range 52.53–54.52; submissions 2021-07-22 to 2022-03-01; OPTIX backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
33.5 s · community measurement, sample 182 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 182 submissions (15 outliers excluded), interquartile range 32.84–34.67; submissions 2021-07-10 to 2022-02-28; OPTIX backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 62.8%
Blender Benchmark 2.93.1
GPU, fishy_cat (CUDA) · s, lower is better
GeForce RTX 307056 sGeForce RTX 309034.4 sPosition among tested models (21, CUDA)
B · 34A · 560100200300Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 62.8% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
56 s · community measurement, sample 112 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 112 submissions (23 outliers excluded), interquartile range 55.47–57.13; submissions 2021-07-11 to 2022-02-25; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
34.4 s · community measurement, sample 213 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 213 submissions (51 outliers excluded), interquartile range 34.13–34.9; submissions 2021-07-09 to 2022-03-01; CUDA backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 81.9%
Blender Benchmark 2.93.1
GPU, fishy_cat (OPTIX) · s, lower is better
GeForce RTX 307032.6 sGeForce RTX 309017.9 sPosition among tested models (13, OptiX)
B · 18A · 33020406080Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 81.9% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
32.6 s · community measurement, sample 87 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 87 submissions (5 outliers excluded), interquartile range 32.13–33.58; submissions 2021-07-22 to 2022-03-01; OPTIX backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
17.9 s · community measurement, sample 141 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 141 submissions (40 outliers excluded), interquartile range 17.77–18.14; submissions 2021-07-10 to 2022-02-28; OPTIX backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 69.8%
Blender Benchmark 2.93.1
GPU, koro (CUDA) · s, lower is better
GeForce RTX 307092.7 sGeForce RTX 309054.6 sPosition among tested models (20, CUDA)
B · 55A · 930200400600Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 69.8% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
92.7 s · community measurement, sample 110 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 110 submissions (23 outliers excluded), interquartile range 91.73–94.24; submissions 2021-07-11 to 2022-02-25; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
54.6 s · community measurement, sample 216 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 216 submissions (47 outliers excluded), interquartile range 54.03–55.35; submissions 2021-07-09 to 2022-03-01; CUDA backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 76.6%
Blender Benchmark 2.93.1
GPU, koro (OPTIX) · s, lower is better
GeForce RTX 307067 sGeForce RTX 309037.9 sPosition among tested models (13, OptiX)
B · 38A · 67050100150200Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 76.6% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
67 s · community measurement, sample 78 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 78 submissions (13 outliers excluded), interquartile range 65.96–68.84; submissions 2021-07-19 to 2022-03-01; OPTIX backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
37.9 s · community measurement, sample 143 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 143 submissions (36 outliers excluded), interquartile range 37.49–38.45; submissions 2021-07-10 to 2022-02-28; OPTIX backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 55.3%
Blender Benchmark 2.93.1
GPU, pavillon_barcelona (CUDA) · s, lower is better
GeForce RTX 3070154.3 sGeForce RTX 309099.3 sPosition among tested models (21, CUDA)
B · 99A · 1540200400600800Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 55.3% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
154.3 s · community measurement, sample 109 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 109 submissions (24 outliers excluded), interquartile range 152.83–157.43; submissions 2021-07-11 to 2022-02-25; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
99.3 s · community measurement, sample 215 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 215 submissions (50 outliers excluded), interquartile range 98.46–100.87; submissions 2021-07-09 to 2022-03-01; CUDA backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 65.1%
Blender Benchmark 2.93.1
GPU, pavillon_barcelona (OPTIX) · s, lower is better
GeForce RTX 307074.4 sGeForce RTX 309045.1 sPosition among tested models (13, OptiX)
B · 45A · 740100200300Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 65.1% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
74.4 s · community measurement, sample 84 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 84 submissions (12 outliers excluded), interquartile range 73.58–75.64; submissions 2021-07-19 to 2022-03-01; OPTIX backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
45.1 s · community measurement, sample 145 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 145 submissions (26 outliers excluded), interquartile range 44.57–45.81; submissions 2021-07-10 to 2022-02-28; OPTIX backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 63.2%
Blender Benchmark 2.93.1
GPU, victor (CUDA) · s, lower is better
GeForce RTX 3070226.4 sGeForce RTX 3090138.7 sPosition among tested models (19, CUDA)
B · 139A · 22602505007501,000Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 63.2% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
226.4 s · community measurement, sample 109 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 109 submissions (20 outliers excluded), interquartile range 223.23–229.38; submissions 2021-07-11 to 2022-02-25; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
138.7 s · community measurement, sample 217 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 217 submissions (43 outliers excluded), interquartile range 137.5–140.78; submissions 2021-07-09 to 2022-03-01; CUDA backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 68.3%
Blender Benchmark 2.93.1
GPU, victor (OPTIX) · s, lower is better
GeForce RTX 3070101.7 sGeForce RTX 309060.4 sPosition among tested models (13, OptiX)
B · 60A · 1020100200300Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce RTX 3090 is ahead by 68.3% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
101.7 s · community measurement, sample 85 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 85 submissions (5 outliers excluded), interquartile range 99.6–104.68; submissions 2021-07-22 to 2022-03-01; OPTIX backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
60.4 s · community measurement, sample 144 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 144 submissions (24 outliers excluded), interquartile range 59.46–61.01; submissions 2021-07-10 to 2022-02-28; OPTIX backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Render time of a single Cycles scene without synchronization. Median of submissions after filtering.
- B ahead by 69.6%
Blender Benchmark 4.5.0
GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) · samples/min, higher is better
GeForce RTX 30703,198.3 samples/minGeForce RTX 30905,423.6 samples/minPosition among tested models (67, OptiX)
A · 3,198B · 5,42405,00010,00015,000Each dot is one model’s median in this benchmark. The scale starts at zero, samples/min. NVIDIA GeForce RTX 3090 is ahead by 69.6% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce RTX 3070
3,198.3 samples/min · community measurement, sample 296 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 296 submissions (15 outliers excluded), interquartile range 3085.66–3284.94; submissions 2025-07-17 to 2026-09-05; OPTIX backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce RTX 3090
5,423.6 samples/min · community measurement, sample 596 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 596 submissions (34 outliers excluded), interquartile range 5240.15–5558.47; submissions 2025-07-17 to 2026-10-19; OPTIX backend. Snapshot 2026-11-14.
difference = |A − B| ÷ the smaller value × 100%. The 5% noticeable threshold is a UX heuristic, not statistical significance.
Cycles render in Blender Benchmark. Median of the sum of samples per minute over three scenes after removing outliers.
Other comparable benchmarks (52)
| Benchmark | A | B | Difference |
|---|---|---|---|
| Blender Benchmark 2.90GPU, bmw27 (CUDA) | 34 | 21.5 | B +57.8% |
| Blender Benchmark 2.91.2GPU, bmw27 (CUDA) | 28.6 | 18.5 | B +54.8% |
| Blender Benchmark 2.92GPU, bmw27 (CUDA) | 28.9 | 18.5 | B +56.5% |
| Blender Benchmark 2.90GPU, bmw27 (OPTIX) | 15.3 | 8.9 | B +71.2% |
| Blender Benchmark 2.91.2GPU, bmw27 (OPTIX) | 15.7 | 9.2 | B +69.8% |
| Blender Benchmark 2.92GPU, bmw27 (OPTIX) | 15.5 | 9.1 | B +70.5% |
| Blender Benchmark 2.90GPU, classroom (CUDA) | 111.8 | 70.6 | B +58.3% |
| Blender Benchmark 2.91.2GPU, classroom (CUDA) | 82.9 | 58.8 | B +41% |
| Blender Benchmark 2.92GPU, classroom (CUDA) | 82.2 | 55.5 | B +48% |
| Blender Benchmark 2.90GPU, classroom (OPTIX) | 52.6 | 35.2 | B +49.2% |
| Blender Benchmark 2.91.2GPU, classroom (OPTIX) | 53.6 | 37.1 | B +44.5% |
| Blender Benchmark 2.92GPU, classroom (OPTIX) | 52.8 | 34.4 | B +53.4% |
| Blender Benchmark 2.90GPU, fishy_cat (CUDA) | 61.5 | 39 | B +57.6% |
| Blender Benchmark 2.91.2GPU, fishy_cat (CUDA) | 53.7 | 33.9 | B +58.2% |
| Blender Benchmark 2.92GPU, fishy_cat (CUDA) | 55.3 | 34.8 | B +59% |
| Blender Benchmark 2.90GPU, fishy_cat (OPTIX) | 29.9 | 16.5 | B +80.7% |
| Blender Benchmark 2.91.2GPU, fishy_cat (OPTIX) | 29.9 | 16.7 | B +79.5% |
| Blender Benchmark 2.92GPU, fishy_cat (OPTIX) | 33.4 | 18.4 | B +82.1% |
| Blender Benchmark 2.90GPU, koro (CUDA) | 95.3 | 58 | B +64.5% |
| Blender Benchmark 2.91.2GPU, koro (CUDA) | 94.3 | 57.5 | B +64.1% |
| Blender Benchmark 2.92GPU, koro (CUDA) | 95.6 | 58 | B +64.6% |
| Blender Benchmark 2.90GPU, koro (OPTIX) | 64.1 | 36.5 | B +75.7% |
| Blender Benchmark 2.91.2GPU, koro (OPTIX) | 63.6 | 36.3 | B +75.1% |
| Blender Benchmark 2.92GPU, koro (OPTIX) | 99.6 | 57.1 | B +74.4% |
| Blender Benchmark 2.90GPU, pavillon_barcelona (CUDA) | 188.2 | 119.6 | B +57.4% |
| Blender Benchmark 2.91.2GPU, pavillon_barcelona (CUDA) | 167.7 | 108.5 | B +54.5% |
| Blender Benchmark 2.92GPU, pavillon_barcelona (CUDA) | 168.6 | 109.3 | B +54.3% |
| Blender Benchmark 2.90GPU, pavillon_barcelona (OPTIX) | 72.4 | 45.4 | B +59.7% |
| Blender Benchmark 2.91.2GPU, pavillon_barcelona (OPTIX) | 73.8 | 46.9 | B +57.2% |
| Blender Benchmark 2.92GPU, pavillon_barcelona (OPTIX) | 72.9 | 45.1 | B +61.9% |
| Blender Benchmark 2.90GPU, victor (CUDA) | 247.8 | 159.1 | B +55.7% |
| Blender Benchmark 2.91.2GPU, victor (CUDA) | 200.6 | 124.3 | B +61.5% |
| Blender Benchmark 2.92GPU, victor (CUDA) | 264.4 | 178.3 | B +48.3% |
| Blender Benchmark 2.90GPU, victor (OPTIX) | 84.1 | 52.8 | B +59.4% |
| Blender Benchmark 2.91.2GPU, victor (OPTIX) | 84.8 | 53.9 | B +57.1% |
| Blender Benchmark 2.92GPU, victor (OPTIX) | 107.1 | 62.5 | B +71.2% |
| Blender Benchmark 4.2.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,148.8 | 5,354.8 | B +70.1% |
| Blender Benchmark 5.1.1GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,087.8 | 5,253.3 | B +70.1% |
| Blender Benchmark 5.2.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 2,988.9 | 5,093.8 | B +70.4% |
| Blender Benchmark 3.0.1GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,536.1 | 5,841.8 | B +65.2% |
| Blender Benchmark 3.1.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,427 | 6,560.4 | B +91.4% |
| Blender Benchmark 3.2.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,363.8 | 6,566.2 | B +95.2% |
| Blender Benchmark 3.2.1GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,406.1 | 5,625.7 | B +65.2% |
| Blender Benchmark 3.3.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,542 | 6,018.2 | B +69.9% |
| Blender Benchmark 3.4.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,590.7 | 6,014.6 | B +67.5% |
| Blender Benchmark 3.5.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,774.5 | 6,296.9 | B +66.8% |
| Blender Benchmark 3.6.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,710.9 | 6,305.5 | B +69.9% |
| Blender Benchmark 4.0.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,257.7 | 5,536.1 | B +69.9% |
| Blender Benchmark 4.1.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,285.4 | 5,536.4 | B +68.5% |
| Blender Benchmark 4.3.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,136 | 5,347.6 | B +70.5% |
| Blender Benchmark 4.4.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,162.6 | 5,333 | B +68.6% |
| Blender Benchmark 5.1.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,166.4 | 5,266.1 | B +66.3% |
Each benchmark version has its own scale: numbers are compared only within a row.
Reference results (29)
| Benchmark | A | B | Difference |
|---|---|---|---|
| Blender Benchmark 2.90.1GPU, bmw27 (CUDA) | 31.6 | 20.5 | reference |
| Blender Benchmark 2.90.1GPU, bmw27 (OPTIX) | 16 | 8.9 | reference |
| Blender Benchmark 2.90.1GPU, classroom (CUDA) | 111.3 | 67.4 | reference |
| Blender Benchmark 2.90.1GPU, classroom (OPTIX) | 54.1 | 38.9 | reference |
| Blender Benchmark 2.90.1GPU, fishy_cat (CUDA) | 57.7 | 37.3 | reference |
| Blender Benchmark 2.90.1GPU, fishy_cat (OPTIX) | 30.9 | 16.8 | reference |
| Blender Benchmark 2.90.1GPU, koro (CUDA) | 89.1 | 57.5 | reference |
| Blender Benchmark 2.90.1GPU, koro (OPTIX) | 66.1 | 37.1 | reference |
| Blender Benchmark 2.90.1GPU, pavillon_barcelona (CUDA) | 169.9 | 108.8 | reference |
| Blender Benchmark 2.90.1GPU, pavillon_barcelona (OPTIX) | 75.4 | 46.6 | reference |
| Blender Benchmark 2.90.1GPU, victor (CUDA) | 232.3 | 148.7 | reference |
| Blender Benchmark 2.90.1GPU, victor (OPTIX) | 88.3 | 57.7 | reference |
| Blender Benchmark 3.0.1GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,148.7 | 3,696.5 | reference |
| Blender Benchmark 3.1.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,016.7 | 3,418.3 | reference |
| Blender Benchmark 3.2.1GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,067.7 | 3,448.2 | reference |
| Blender Benchmark 3.3.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,151.9 | 3,729.7 | reference |
| Blender Benchmark 3.4.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,162.3 | 3,797.9 | reference |
| Blender Benchmark 3.5.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,063.9 | 3,864 | reference |
| Blender Benchmark 3.6.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,342.2 | 4,060.8 | reference |
| Blender Benchmark 4.0.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,153 | 3,724.8 | reference |
| Blender Benchmark 4.1.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 1,900.9 | 3,735.6 | reference |
| Blender Benchmark 4.2.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,015.4 | 3,699.3 | reference |
| Blender Benchmark 4.3.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 2,055.2 | 3,730.8 | reference |
| Blender Benchmark 4.5.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 1,793.9 | 3,798 | reference |
| Blender Benchmark 5.1.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 1,991.9 | — | no pair |
| Blender Benchmark 3.1.2GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 3,376.9 | 5,826.5 | reference |
| Blender Benchmark 5.0.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 2,909.7 | 4,875.8 | reference |
| Blender Benchmark 5.3.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 2,573.3 | 5,143.8 | reference |
| Blender Benchmark 3.2.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | — | 3,138.2 | no pair |
Each benchmark version has its own scale: numbers are compared only within a row.
Specifications
| Specification | A · GeForce RTX 3070 | B · GeForce RTX 3090 | Difference |
|---|---|---|---|
| Chip | |||
| Architecture | Ampere | Ampere | equal |
| Shader unitsShader unit counts are comparable only within one architecture. | 5,888 (CUDA cores) | 10,496 (CUDA cores) | B: 4,608 more |
| Clocks | |||
| Base clock | 1,500 MHz | 1,400 MHz | A: 100 MHz more |
| Boost clockClocks of different architectures do not reflect performance directly. For laptop chips, the clock depends on the laptop’s power limit. | 1,730 MHz | 1,700 MHz | A: 30 MHz more |
| Memory | |||
| Video memory | 8 GB | 24 GB | B: 16 GB more |
| Memory type | GDDR6 | GDDR6X | differs |
| Memory bus | 256 bit | 384 bit | B: 128 bit more |
| Power and dimensions | |||
| Board powerTGP (NVIDIA) and TBP (AMD/Intel) are calculated differently; higher is not better. For laptops, the chip maker sets the range and the laptop maker picks the value, so the same chip behaves differently in different laptops. | 220 W (board power) | 350 W (board power) | B: 130 W higher |
| Recommended PSU | 650 W | 750 W | B: 100 W higher |
| Length (reference / Founders Edition)Partner cards have different dimensions; check the specific SKU. | 242 mm | 313 mm | B: 71 mm higher |
| Thickness (slots) | 2 | 3 | B: 1 higher |
| Power connectors (reference) | 1x 8-pin (adapter to 1x 12-pin included) — FE | 2x 8-pin (adapter to 1x 12-pin included) — FE | differs |
Chip
Architecture
AAmpere
BAmpere
equalShader units
A5,888 (CUDA cores)
B10,496 (CUDA cores)
B: 4,608 moreShader unit counts are comparable only within one architecture.
Clocks
Base clock
A1,500 MHz
B1,400 MHz
A: 100 MHz moreBoost clock
A1,730 MHz
B1,700 MHz
A: 30 MHz moreClocks of different architectures do not reflect performance directly. For laptop chips, the clock depends on the laptop’s power limit.
Memory
Video memory
A8 GB
B24 GB
B: 16 GB moreMemory type
AGDDR6
BGDDR6X
differsMemory bus
A256 bit
B384 bit
B: 128 bit more
Power and dimensions
Board power
A220 W (board power)
B350 W (board power)
B: 130 W higherTGP (NVIDIA) and TBP (AMD/Intel) are calculated differently; higher is not better. For laptops, the chip maker sets the range and the laptop maker picks the value, so the same chip behaves differently in different laptops.
Recommended PSU
A650 W
B750 W
B: 100 W higherLength (reference / Founders Edition)
A242 mm
B313 mm
B: 71 mm higherPartner cards have different dimensions; check the specific SKU.
Thickness (slots)
A2
B3
B: 1 higherPower connectors (reference)
A1x 8-pin (adapter to 1x 12-pin included) — FE
B2x 8-pin (adapter to 1x 12-pin included) — FE
differs
“More” does not always mean “better”: higher power draw or size is a system requirement, not an advantage (such bars are dimmed).
Compare with AI
An AI model reviews the specs and benchmarks of this pair and gives a plain-language verdict on who each model suits.
Questions and answers
How much video memory do NVIDIA GeForce RTX 3070 and NVIDIA GeForce RTX 3090 have?
NVIDIA GeForce RTX 3070: 8 GB GDDR6, 256-bit bus. NVIDIA GeForce RTX 3090: 24 GB GDDR6X, 384-bit bus.
What power supply does the manufacturer recommend?
For NVIDIA GeForce RTX 3070, 650 W or more; for NVIDIA GeForce RTX 3090, 750 W or more. This is a recommendation for a typical system; also check the power connectors of the specific card model.
Which is faster: NVIDIA GeForce RTX 3070 or NVIDIA GeForce RTX 3090?
NVIDIA GeForce RTX 3090 leads in every scenario with comparable benchmarks: rendering and compute. This does not imply an advantage in tasks without benchmarks.
See also
Data sources (3)
- NVIDIA — specifications of NVIDIA GeForce RTX 3070, specifications of NVIDIA GeForce RTX 3090checked November 7, 2026
- Blender Open Data (Blender Foundation) — benchmark Blender Benchmark 2.90, benchmark Blender Benchmark 2.90.1, benchmark Blender Benchmark 2.91.2, benchmark Blender Benchmark 2.92, benchmark Blender Benchmark 2.93.1, benchmark Blender Benchmark 3.0.1, benchmark Blender Benchmark 3.1.0, benchmark Blender Benchmark 3.2.0, benchmark Blender Benchmark 3.2.1, benchmark Blender Benchmark 3.3.0, benchmark Blender Benchmark 3.4.0, benchmark Blender Benchmark 3.5.0, benchmark Blender Benchmark 3.6.0, benchmark Blender Benchmark 4.0.0, benchmark Blender Benchmark 4.1.0, benchmark Blender Benchmark 4.2.0, benchmark Blender Benchmark 4.3.0, benchmark Blender Benchmark 4.5.0, benchmark Blender Benchmark 5.1.0, benchmark Blender Benchmark 5.1.1, benchmark Blender Benchmark 5.2.0, benchmark Blender Benchmark 3.1.2, benchmark Blender Benchmark 4.4.0, benchmark Blender Benchmark 5.0.0, benchmark Blender Benchmark 5.3.0checked November 14, 2026
- Wikipedia (English) — release date and launch pricechecked November 14, 2026
How we compare: methodology. Where the data comes from: sources.