GeForce GTX 1070 vs GeForce GTX 1660: specs and benchmark comparison
Short verdict
NVIDIA GeForce GTX 1660 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 +35.4%NVIDIA GeForce GTX 1660 leads in this scenario (up to 35.4% 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 GTX 1070 is better2
Video memory 2 GB more8 GB vs 6 GB
Memory bus 64 bit more256 bit vs 192 bit
Where GeForce GTX 1660 is better5
Blender Benchmark 2.90 6% lower344 s vs 365 s
Blender Benchmark 2.90 8.3% lower201 s vs 217 s
Blender Benchmark 2.90 9.2% lower329 s vs 360 s
Blender Benchmark 4.5.0 29.8% faster680 samples/min vs 524 samples/min
Launch price $160 lower$219 vs $379
Benchmarks
Results are compared within one benchmark, one version and the same conditions.
- 2.3% difference, within the threshold
Blender Benchmark 2.91.2
GPU, bmw27 (CUDA) · s, lower is better
GeForce GTX 1070105.4 sGeForce GTX 1660103 sPosition among tested models (9, CUDA)
B · 103A · 105050100150Each dot is one model’s median in this benchmark. The scale starts at zero, s. The 2.3% difference is below the noticeable threshold (5%).
About the measurement and source
- A · NVIDIA GeForce GTX 1070
105.4 s · community measurement, sample 21 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 21 submissions (2 outliers excluded), interquartile range 103.05–107.6; submissions 2021-02-01 to 2021-11-14; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce GTX 1660
103 s · community measurement, sample 21 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 21 submissions (3 outliers excluded), interquartile range 100.54–103.45; submissions 2021-02-13 to 2021-03-11; 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 6%
Blender Benchmark 2.90
GPU, classroom (CUDA) · s, lower is better
GeForce GTX 1070365.3 sGeForce GTX 1660344.5 sPosition among tested models (25, CUDA)
B · 344A · 3650200400600800Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce GTX 1660 is ahead by 6% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce GTX 1070
365.3 s · community measurement, sample 179 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 179 submissions (23 outliers excluded), interquartile range 355.81–376.18; submissions 2020-09-02 to 2021-10-14; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce GTX 1660
344.5 s · community measurement, sample 45 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 45 submissions (8 outliers excluded), interquartile range 341.82–356.75; submissions 2020-09-03 to 2021-01-20; 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 8.3%
Blender Benchmark 2.90
GPU, fishy_cat (CUDA) · s, lower is better
GeForce GTX 1070217.2 sGeForce GTX 1660200.5 sPosition among tested models (22, CUDA)
B · 201A · 2170200400600Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce GTX 1660 is ahead by 8.3% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce GTX 1070
217.2 s · community measurement, sample 159 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 159 submissions (15 outliers excluded), interquartile range 212.58–221.75; submissions 2020-09-02 to 2021-10-14; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce GTX 1660
200.5 s · community measurement, sample 40 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 40 submissions (6 outliers excluded), interquartile range 199.05–204.92; submissions 2020-09-03 to 2021-01-09; 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 9.2%
Blender Benchmark 2.90
GPU, koro (CUDA) · s, lower is better
GeForce GTX 1070359.7 sGeForce GTX 1660329.3 sPosition among tested models (22, CUDA)
B · 329A · 3600200400600800Each dot is one model’s median in this benchmark. The scale starts at zero, s. NVIDIA GeForce GTX 1660 is ahead by 9.2% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce GTX 1070
359.7 s · community measurement, sample 157 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 157 submissions (10 outliers excluded), interquartile range 350.81–370.44; submissions 2020-09-02 to 2021-10-14; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce GTX 1660
329.3 s · community measurement, sample 38 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 38 submissions (7 outliers excluded), interquartile range 326.4–336.55; submissions 2020-09-03 to 2021-01-09; 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.
- 0.4% difference, within the threshold
Blender Benchmark 2.90
GPU, pavillon_barcelona (CUDA) · s, lower is better
GeForce GTX 1070604.4 sGeForce GTX 1660606.6 sPosition among tested models (23, CUDA)
A · 604B · 60705001,0001,500Each dot is one model’s median in this benchmark. The scale starts at zero, s. The 0.4% difference is below the noticeable threshold (5%).
About the measurement and source
- A · NVIDIA GeForce GTX 1070
604.4 s · community measurement, sample 154 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 154 submissions (14 outliers excluded), interquartile range 591.84–621.3; submissions 2020-09-02 to 2021-10-14; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce GTX 1660
606.6 s · community measurement, sample 38 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 38 submissions (6 outliers excluded), interquartile range 602.24–614.64; submissions 2020-09-03 to 2021-01-09; 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.
- 2.4% difference, within the threshold
Blender Benchmark 2.90
GPU, victor (CUDA) · s, lower is better
GeForce GTX 1070800.4 sGeForce GTX 1660781.7 sPosition among tested models (20, CUDA)
B · 782A · 80005001,0001,500Each dot is one model’s median in this benchmark. The scale starts at zero, s. The 2.4% difference is below the noticeable threshold (5%).
About the measurement and source
- A · NVIDIA GeForce GTX 1070
800.4 s · community measurement, sample 138 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 138 submissions (22 outliers excluded), interquartile range 784.18–821.9; submissions 2020-09-02 to 2021-10-14; CUDA backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce GTX 1660
781.7 s · community measurement, sample 36 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 36 submissions (7 outliers excluded), interquartile range 768.8–797.89; submissions 2020-09-03 to 2021-01-09; 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 29.8%
Blender Benchmark 4.5.0
GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) · samples/min, higher is better
GeForce GTX 1070523.6 samples/minGeForce GTX 1660679.7 samples/minPosition among tested models (67, OptiX)
A · 524B · 68005,00010,00015,000Each dot is one model’s median in this benchmark. The scale starts at zero, samples/min. NVIDIA GeForce GTX 1660 is ahead by 29.8% in this benchmark.
About the measurement and source
- A · NVIDIA GeForce GTX 1070
523.6 samples/min · community measurement, sample 72 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 72 submissions (6 outliers excluded), interquartile range 509.18–537.94; submissions 2025-07-25 to 2026-03-05; OPTIX backend. Snapshot 2026-11-14.
- B · NVIDIA GeForce GTX 1660
679.7 samples/min · community measurement, sample 22 · November 14, 2026
Blender Open Data (Blender Foundation)
Median of 22 submissions (0 outliers excluded), interquartile range 623.88–697.87; submissions 2025-07-17 to 2026-03-26; 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 (17)
| Benchmark | A | B | Difference |
|---|---|---|---|
| Blender Benchmark 2.82GPU, bmw27 (CUDA) | 102.9 | 97.7 | B +5.4% |
| Blender Benchmark 2.83GPU, bmw27 (CUDA) | 105.6 | 98.2 | B +7.5% |
| Blender Benchmark 2.90GPU, bmw27 (CUDA) | 112.1 | 103.8 | B +7.9% |
| Blender Benchmark 2.82GPU, classroom (CUDA) | 353.4 | 337.8 | B +4.6% |
| Blender Benchmark 2.83GPU, classroom (CUDA) | 363.8 | 340.7 | B +6.8% |
| Blender Benchmark 2.83GPU, fishy_cat (CUDA) | 213 | 194 | B +9.8% |
| Blender Benchmark 2.82GPU, koro (CUDA) | 401 | 316.7 | B +26.6% |
| Blender Benchmark 2.83GPU, koro (CUDA) | 393.7 | 330.3 | B +19.2% |
| Blender Benchmark 2.82GPU, pavillon_barcelona (CUDA) | 575.1 | 582.5 | A +1.3% |
| Blender Benchmark 2.83GPU, pavillon_barcelona (CUDA) | 588.2 | 580.9 | B +1.3% |
| Blender Benchmark 2.83GPU, victor (CUDA) | 807.6 | 800.2 | B +0.9% |
| Blender Benchmark 3.1.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 567.2 | 744.3 | B +31.2% |
| Blender Benchmark 3.3.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 567.4 | 724.3 | B +27.6% |
| Blender Benchmark 3.4.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 567.6 | 726.6 | B +28% |
| Blender Benchmark 3.6.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 574 | 777.3 | B +35.4% |
| Blender Benchmark 4.0.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 540.2 | 700 | B +29.6% |
| Blender Benchmark 4.1.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 548 | 700.7 | B +27.9% |
Each benchmark version has its own scale: numbers are compared only within a row.
Reference results (91)
| Benchmark | A | B | Difference |
|---|---|---|---|
| Blender Benchmark 2.81GPU, barbershop_interior (CUDA) | 988.7 | 1,138.5 | reference |
| Blender Benchmark 2.81aGPU, barbershop_interior (CUDA) | 1,012.7 | 1,181.5 | reference |
| Blender Benchmark 2.81GPU, bmw27 (CUDA) | 103.4 | 96.3 | reference |
| Blender Benchmark 2.81aGPU, bmw27 (CUDA) | 104.1 | 95.2 | reference |
| Blender Benchmark 2.92GPU, bmw27 (CUDA) | 104.2 | 103.8 | reference |
| Blender Benchmark 2.93.1GPU, bmw27 (CUDA) | 90.2 | 88.5 | reference |
| Blender Benchmark 2.90GPU, bmw27 (OPTIX) | 108.5 | 82.1 | reference |
| Blender Benchmark 2.90.1GPU, bmw27 (OPTIX) | 162.3 | — | no pair |
| Blender Benchmark 2.91.2GPU, bmw27 (OPTIX) | 107.8 | 74.3 | reference |
| Blender Benchmark 2.92GPU, bmw27 (OPTIX) | 109.1 | 79.2 | reference |
| Blender Benchmark 2.93.1GPU, bmw27 (OPTIX) | 108.3 | 86 | reference |
| Blender Benchmark 2.81GPU, classroom (CUDA) | 351.4 | 342 | reference |
| Blender Benchmark 2.81aGPU, classroom (CUDA) | 361.2 | 345.9 | reference |
| Blender Benchmark 2.91.2GPU, classroom (CUDA) | 329.9 | 347.4 | reference |
| Blender Benchmark 2.92GPU, classroom (CUDA) | 351.7 | 361.4 | reference |
| Blender Benchmark 2.93.1GPU, classroom (CUDA) | 317.9 | 316.4 | reference |
| Blender Benchmark 2.90GPU, classroom (OPTIX) | 396.6 | 319.1 | reference |
| Blender Benchmark 2.90.1GPU, classroom (OPTIX) | 630.3 | — | no pair |
| Blender Benchmark 2.91.2GPU, classroom (OPTIX) | 389.5 | 313.8 | reference |
| Blender Benchmark 2.92GPU, classroom (OPTIX) | 386.7 | — | no pair |
| Blender Benchmark 2.93.1GPU, classroom (OPTIX) | 412.9 | 328 | reference |
| Blender Benchmark 2.81GPU, fishy_cat (CUDA) | 205.9 | 186.7 | reference |
| Blender Benchmark 2.81aGPU, fishy_cat (CUDA) | 208.3 | 186.4 | reference |
| Blender Benchmark 2.82GPU, fishy_cat (CUDA) | 211.7 | 188.1 | reference |
| Blender Benchmark 2.91.2GPU, fishy_cat (CUDA) | 204 | 188.6 | reference |
| Blender Benchmark 2.92GPU, fishy_cat (CUDA) | 208.4 | 201.5 | reference |
| Blender Benchmark 2.93.1GPU, fishy_cat (CUDA) | 211.5 | 198.1 | reference |
| Blender Benchmark 2.90GPU, fishy_cat (OPTIX) | 205.8 | 159.5 | reference |
| Blender Benchmark 2.90.1GPU, fishy_cat (OPTIX) | 322.6 | — | no pair |
| Blender Benchmark 2.91.2GPU, fishy_cat (OPTIX) | 200.3 | 159.4 | reference |
| Blender Benchmark 2.92GPU, fishy_cat (OPTIX) | 204.8 | — | no pair |
| Blender Benchmark 2.93.1GPU, fishy_cat (OPTIX) | 213.2 | 170.1 | reference |
| Blender Benchmark 2.81GPU, koro (CUDA) | 401.4 | 320.4 | reference |
| Blender Benchmark 2.81aGPU, koro (CUDA) | 408.8 | 316.5 | reference |
| Blender Benchmark 2.91.2GPU, koro (CUDA) | 345.7 | 286.5 | reference |
| Blender Benchmark 2.92GPU, koro (CUDA) | 362.2 | 313.5 | reference |
| Blender Benchmark 2.93.1GPU, koro (CUDA) | 369 | 334.9 | reference |
| Blender Benchmark 2.90GPU, koro (OPTIX) | 239.7 | 199.4 | reference |
| Blender Benchmark 2.90.1GPU, koro (OPTIX) | 377.3 | — | no pair |
| Blender Benchmark 2.91.2GPU, koro (OPTIX) | 228.9 | 196.8 | reference |
| Blender Benchmark 2.92GPU, koro (OPTIX) | 316.7 | — | no pair |
| Blender Benchmark 2.93.1GPU, koro (OPTIX) | 247.4 | 211.1 | reference |
| Blender Benchmark 2.81GPU, pavillon_barcelona (CUDA) | 575.7 | 580.6 | reference |
| Blender Benchmark 2.81aGPU, pavillon_barcelona (CUDA) | 580.5 | 584.7 | reference |
| Blender Benchmark 2.91.2GPU, pavillon_barcelona (CUDA) | 558.8 | 551.6 | reference |
| Blender Benchmark 2.92GPU, pavillon_barcelona (CUDA) | 582.9 | 592.3 | reference |
| Blender Benchmark 2.93.1GPU, pavillon_barcelona (CUDA) | 550.6 | 518.7 | reference |
| Blender Benchmark 2.90GPU, pavillon_barcelona (OPTIX) | 579 | 458.8 | reference |
| Blender Benchmark 2.90.1GPU, pavillon_barcelona (OPTIX) | 873.9 | — | no pair |
| Blender Benchmark 2.91.2GPU, pavillon_barcelona (OPTIX) | 560.3 | 455.5 | reference |
| Blender Benchmark 2.92GPU, pavillon_barcelona (OPTIX) | 558.4 | — | no pair |
| Blender Benchmark 2.93.1GPU, pavillon_barcelona (OPTIX) | 592.5 | 471.8 | reference |
| Blender Benchmark 2.81GPU, victor (CUDA) | 836.2 | 806.9 | reference |
| Blender Benchmark 2.81aGPU, victor (CUDA) | 906.2 | 798.7 | reference |
| Blender Benchmark 2.82GPU, victor (CUDA) | 878.9 | 789.6 | reference |
| Blender Benchmark 2.91.2GPU, victor (CUDA) | 763.4 | 764.3 | reference |
| Blender Benchmark 2.92GPU, victor (CUDA) | 867.1 | 919.3 | reference |
| Blender Benchmark 2.93.1GPU, victor (CUDA) | 850.4 | 860.9 | reference |
| Blender Benchmark 2.90GPU, victor (OPTIX) | 778.6 | 646.9 | reference |
| Blender Benchmark 2.90.1GPU, victor (OPTIX) | 1,202.1 | — | no pair |
| Blender Benchmark 2.91.2GPU, victor (OPTIX) | 752.1 | 688.9 | reference |
| Blender Benchmark 2.92GPU, victor (OPTIX) | 818.9 | — | no pair |
| Blender Benchmark 2.93.1GPU, victor (OPTIX) | 848.4 | 702.8 | reference |
| Blender Benchmark 3.0.1GPU, sum of scenes monster + junkshop + classroom (CUDA) | 705.2 | 808.8 | reference |
| Blender Benchmark 3.1.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 639.5 | 707.9 | reference |
| Blender Benchmark 3.2.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 642.2 | — | no pair |
| Blender Benchmark 3.2.1GPU, sum of scenes monster + junkshop + classroom (CUDA) | 653.8 | 764.5 | reference |
| Blender Benchmark 3.3.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 619.2 | — | no pair |
| Blender Benchmark 3.4.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 647.6 | — | no pair |
| Blender Benchmark 3.5.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 631.4 | 791.7 | reference |
| Blender Benchmark 3.6.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 653 | 616.3 | reference |
| Blender Benchmark 4.0.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 610.5 | 754.9 | reference |
| Blender Benchmark 4.1.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 662.3 | 744.5 | reference |
| Blender Benchmark 4.2.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 606.7 | — | no pair |
| Blender Benchmark 4.3.0GPU, sum of scenes monster + junkshop + classroom (CUDA) | 628.7 | 775.6 | reference |
| Blender Benchmark 4.2.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 534.5 | 689.2 | reference |
| Blender Benchmark 5.1.1GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 449.5 | 595.5 | reference |
| Blender Benchmark 5.2.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 474.2 | 617.6 | reference |
| Blender Benchmark 3.0.1GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 624.6 | 798.6 | reference |
| Blender Benchmark 3.2.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 567.1 | 644.9 | reference |
| Blender Benchmark 3.2.1GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 572.3 | 753.3 | reference |
| Blender Benchmark 3.5.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 573.3 | 754 | reference |
| Blender Benchmark 4.3.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 522.5 | 686.4 | reference |
| Blender Benchmark 4.4.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 527.1 | 676 | reference |
| Blender Benchmark 5.1.0GPU, sum of scenes monster + junkshop + classroom (OptiX / HIP / oneAPI / Metal) | 461.2 | 490.6 | reference |
| Blender Benchmark 2.90.1GPU, bmw27 (CUDA) | — | 111.6 | no pair |
| Blender Benchmark 2.90.1GPU, classroom (CUDA) | — | 385.9 | no pair |
| Blender Benchmark 2.90.1GPU, fishy_cat (CUDA) | — | 261.8 | no pair |
| Blender Benchmark 2.90.1GPU, koro (CUDA) | — | 402.1 | no pair |
| Blender Benchmark 2.90.1GPU, pavillon_barcelona (CUDA) | — | 874.8 | no pair |
| Blender Benchmark 2.90.1GPU, victor (CUDA) | — | 980.8 | no pair |
Each benchmark version has its own scale: numbers are compared only within a row.
Specifications
| Specification | A · GeForce GTX 1070 | B · GeForce GTX 1660 | Difference |
|---|---|---|---|
| Chip | |||
| GPU chip | GP104-200-A1 | no data | no data for B |
| Architecture | Pascal | Turing | differs |
| Shader unitsShader unit counts are comparable only within one architecture. | 1,920 (CUDA cores) | 1,408 (CUDA cores) | not directly comparable |
| Clocks | |||
| Base clock | 1,506 MHz | 1,530 MHz | B: 24 MHz more |
| Boost clockClocks of different architectures do not reflect performance directly. For laptop chips, the clock depends on the laptop’s power limit. | 1,683 MHz | 1,785 MHz | B: 102 MHz more |
| Memory | |||
| Video memory | 8 GB | 6 GB | A: 2 GB more |
| Memory type | GDDR5 | GDDR5 | equal |
| Memory bus | 256 bit | 192 bit | A: 64 bit more |
| Memory bandwidthExcludes cache (Infinity Cache / L2), which affects effective bandwidth differently. | 256.3 GB/s | no data | no data for B |
| 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. | 150 W (TDP) | 120 W (board power) | not directly comparable |
| Recommended PSU | no data | 450 W | no data for A |
| Interface | PCIe 3.0 x16 | PCIe 3.0 | differs |
| Length (reference / Founders Edition)Partner cards have different dimensions; check the specific SKU. | no data | 145 mm | no data for A |
| Thickness (slots) | no data | 2 | no data for A |
| Power connectors (reference) | no data | 1x 8-pin | no data for A |
Chip
GPU chip
AGP104-200-A1
Bno data
no data for BArchitecture
APascal
BTuring
differsShader units
A1,920 (CUDA cores)
B1,408 (CUDA cores)
not directly comparableShader unit counts are comparable only within one architecture.
Clocks
Base clock
A1,506 MHz
B1,530 MHz
B: 24 MHz moreBoost clock
A1,683 MHz
B1,785 MHz
B: 102 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
B6 GB
A: 2 GB moreMemory type
AGDDR5
BGDDR5
equalMemory bus
A256 bit
B192 bit
A: 64 bit moreMemory bandwidth
A256.3 GB/s
Bno data
no data for BExcludes cache (Infinity Cache / L2), which affects effective bandwidth differently.
Power and dimensions
Board power
A150 W (TDP)
B120 W (board power)
not directly comparableTGP (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
Ano data
B450 W
no data for AInterface
APCIe 3.0 x16
BPCIe 3.0
differsLength (reference / Founders Edition)
Ano data
B145 mm
no data for APartner cards have different dimensions; check the specific SKU.
Thickness (slots)
Ano data
B2
no data for APower connectors (reference)
Ano data
B1x 8-pin
no data for A
“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 GTX 1070 and NVIDIA GeForce GTX 1660 have?
NVIDIA GeForce GTX 1070: 8 GB GDDR5, 256-bit bus. NVIDIA GeForce GTX 1660: 6 GB GDDR5, 192-bit bus.
Which PCIe interface do the cards use?
NVIDIA GeForce GTX 1070: PCIe 3.0 x16. NVIDIA GeForce GTX 1660: PCIe 3.0.
Which is faster: NVIDIA GeForce GTX 1070 or NVIDIA GeForce GTX 1660?
NVIDIA GeForce GTX 1660 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)
- Wikipedia (English) — specifications of NVIDIA GeForce GTX 1070, release date and launch pricechecked November 14, 2026
- NVIDIA — specifications of NVIDIA GeForce GTX 1660checked November 7, 2026
- Blender Open Data (Blender Foundation) — benchmark Blender Benchmark 2.81, benchmark Blender Benchmark 2.81a, benchmark Blender Benchmark 2.82, benchmark Blender Benchmark 2.83, benchmark Blender Benchmark 2.90, benchmark Blender Benchmark 2.91.2, benchmark Blender Benchmark 2.92, benchmark Blender Benchmark 2.93.1, benchmark Blender Benchmark 2.90.1, 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 5.1.1, benchmark Blender Benchmark 5.2.0, benchmark Blender Benchmark 4.4.0, benchmark Blender Benchmark 5.1.0checked November 14, 2026
How we compare: methodology. Where the data comes from: sources.