NVIDIA GeForce GTX 1060 3 GB vs NVIDIA GeForce RTX 2070 Comparison

The data shows two NVIDIA GPUs from different generations: the GeForce GTX 1060 3 GB, a Pascal-era card from 2016, and the GeForce RTX 2070, a Turing-based card from 2018. While their overall percentile rankings are close—64th and 63rd respectively—their performance profiles are drastically different, with the RTX 2070 delivering nearly triple the compute throughput in several key tests. The GTX 1060 3 GB remains a capable entry-level option, but the RTX 2070 is in a different performance class entirely.

The Verdict

The RTX 2070 is the clear performance winner, and the data indicates it is the appropriate choice for users who need substantial compute power. In the Geekbench OpenCL test, it scores 79,966 versus the GTX 1060 3 GB’s 27,001, a margin of 66.2%. The Vulkan test shows an even larger gap: 82,521 versus 17,475, meaning the RTX 2070 leads by 78.8%. These are not incremental gains; they represent a generational leap.

However, the GTX 1060 3 GB is not without merit. Its average benchmark score of 19,334 is actually higher than the RTX 2070’s 18,789, despite the latter’s dominance in the head-to-head tests. This is because the GTX 1060 3 GB’s benchmark suite (Geekbench Metal, OpenCL, Vulkan) differs from the RTX 2070’s (which includes Passmark tests). The GTX 1060 3 GB also sits in a tight cluster of rivals—it is within 1.3% of the AMD Radeon RX 7900 XTX, NVIDIA TITAN Xp, GeForce GTX 780, and Tesla K20m—suggesting it is a well-balanced card for its era. The RTX 2070, by contrast, is surrounded by professional and workstation-class parts like the Tesla K80 and RTX 2000 Ada Generation, indicating its compute orientation.

Pick the RTX 2070 if you need raw performance, modern API support (DirectX 12 Ultimate), and ray tracing capabilities. Pick the GTX 1060 3 GB if you are looking for a legacy card that still competes with much newer hardware in aggregate scores, but be aware that its individual compute results are far behind.

Where Each One Wins

The RTX 2070 wins decisively in every direct head-to-head benchmark. In Geekbench OpenCL, it achieves nearly three times the score of the GTX 1060 3 GB (79,966 vs 27,001). In Geekbench Vulkan, the advantage is even more pronounced, with the RTX 2070 scoring 82,521 against 17,475. These are the only two tests where both cards were measured against each other, and the RTX 2070 won both.

The GTX 1060 3 GB’s wins are less direct. Its average benchmark score of 19,334 is 2.9% higher than the RTX 2070’s 18,789. It also performs competitively against its own nearest rivals, with a delta of just -0.4% versus the AMD Radeon RX 7900 XTX and +0.8% versus the NVIDIA TITAN Xp. This suggests that, in the specific workloads captured by its benchmark set (including Geekbench Metal, where it scores 13,527), the GTX 1060 3 GB punches above its weight class. However, none of these tests overlap with the RTX 2070’s suite, so this is not a direct comparison.

For gaming and general compute, the RTX 2070 is the superior tool. For legacy DirectX 9/10/11 workloads, the RTX 2070’s Passmark scores (211 for DirectX 9, 111 for DirectX 10, 129 for DirectX 11) indicate strong compatibility, though no comparable GTX 1060 3 GB data exists in the pack.

Architecture Differences

The two cards are built on fundamentally different architectures. The GTX 1060 3 GB uses the GP106 chip on NVIDIA’s Pascal architecture, fabricated on TSMC’s 16 nm process. It packs 4,400 million transistors onto a 200 mm² die, yielding a transistor density of 22.0M per mm². The RTX 2070 uses the TU106 chip on the Turing architecture, built on TSMC’s 12 nm process. It contains 10,800 million transistors on a 445 mm² die, with a density of 24.3M per mm². The RTX 2070 has 2.45 times the transistor count and a 2.22 times larger die.

The memory subsystems differ significantly. The GTX 1060 3 GB has 3 GB of GDDR5 on a 192-bit bus, providing 192.2 GB/s of bandwidth. The RTX 2070 has 8 GB of GDDR6 on a 256-bit bus, delivering 448.0 GB/s—more than double the bandwidth. Clock speeds tell a nuanced story: the GTX 1060 3 GB has a higher base clock (1506 MHz vs 1410 MHz) and boost clock (1708 MHz vs 1620 MHz), but the RTX 2070’s memory runs at 14 Gbps effective versus 8 Gbps.

The compute resources are heavily skewed toward the RTX 2070. It has 2,304 shading units, 144 TMUs, and 64 ROPs, versus the GTX 1060 3 GB’s 1,152 shading units, 72 TMUs, and 48 ROPs. The RTX 2070 also introduces 36 ray tracing cores and 288 tensor cores, which the GTX 1060 3 GB lacks entirely. This translates into 7.465 TFLOPS FP32 performance for the RTX 2070 versus 3.935 TFLOPS for the GTX 1060 3 GB. The FP16 gap is enormous: 14.93 TFLOPS (2:1 ratio) versus 61.49 GFLOPS (1:64 ratio).

Power and physical characteristics also differ. The RTX 2070 has a 175 W TDP and requires a 450 W PSU, while the GTX 1060 3 GB has a 120 W TDP and a 300 W PSU suggestion. The RTX 2070 uses a single 8-pin power connector, while the GTX 1060 3 GB uses a 6-pin. The RTX 2070 is shorter (229 mm vs 250 mm) but adds a USB Type-C output and supports DirectX 12 Ultimate (12_2), whereas the GTX 1060 3 GB is limited to DirectX 12 (12_1).

FAQ

Q: Which card has better overall benchmark performance?

A: It depends on the test. The RTX 2070 wins both direct head-to-head tests by large margins (66.2% in OpenCL, 78.8% in Vulkan). However, the GTX 1060 3 GB has a higher average benchmark score (19,334 vs 18,789), though these averages come from different test suites.

Q: Does the RTX 2070 support ray tracing?

A: Yes. The RTX 2070 has 36 ray tracing cores, a feature completely absent from the GTX 1060 3 GB, which has no RT cores.

Q: Which card has more memory bandwidth?

A: The RTX 2070 has 448.0 GB/s of bandwidth from its 8 GB GDDR6 memory on a 256-bit bus. The GTX 1060 3 GB offers 192.2 GB/s from 3 GB GDDR5 on a 192-bit bus.

Q: Are these cards the same physical size?

A: No. The GTX 1060 3 GB is 250 mm (9.8 inches) long, while the RTX 2070 is 229 mm (9 inches) long. The RTX 2070 also has listed dimensions of 113 mm height and 35 mm width, while the GTX 1060 3 GB’s height and width are not specified.

Q: Which card has a higher boost clock?

A: The GTX 1060 3 GB has a boost clock of 1708 MHz, which is higher than the RTX 2070’s 1620 MHz. However, the RTX 2070 has a higher base clock in terms of raw performance due to its superior architecture and compute units.

Q: What are the power requirements?

A: The RTX 2070 has a 175 W TDP and recommends a 450 W PSU, while the GTX 1060 3 GB has a 120 W TDP and suggests a 300 W PSU. The RTX 2070 uses a single 8-pin connector, while the GTX 1060 3 GB uses a 6-pin.

Head-to-Head Benchmarks

The RTX 2070 dominates the two direct comparisons available. In Geekbench OpenCL, the RTX 2070 scores 79,966 against the GTX 1060 3 GB’s 27,001. This represents a 66.2% delta, meaning the RTX 2070 is roughly three times faster. The Geekbench Vulkan test shows an even starker contrast: 82,521 for the RTX 2070 versus 17,475 for the GTX 1060 3 GB, a 78.8% margin. In this test, the RTX 2070 achieves nearly five times the score.

These results align with the architectural disparity. The RTX 2070’s 7.465 TFLOPS FP32 throughput is 89.7% higher than the GTX 1060 3 GB’s 3.935 TFLOPS. The texture rate tells a similar story: 233.3 GTexel/s versus 123.0 GTexel/s. The pixel rate gap is smaller but still significant: 103.7 GPixel/s versus 81.98 GPixel/s.

The GTX 1060 3 GB’s only statistical advantage is its aggregate average benchmark score (19,334 vs 18,789), which is driven by its strong Geekbench Metal result of 13,527 and its OpenCL score of 27,001. However, these numbers cannot be directly compared to the RTX 2070’s Passmark suite, where the RTX 2070 scores 16,094 in G3D and 6,411 in GPU Compute. The RTX 2070’s Passmark DirectX 9 score of 211, DirectX 10 score of 111, and DirectX 11 score of 129 confirm its backward compatibility, while its DirectX 12 score of 60 reflects the demanding nature of modern APIs.

In summary, every measurable head-to-head metric favors the RTX 2070, with margins ranging from 66.2% to 78.8%. The GTX 1060 3 GB’s higher average score is an artifact of differing test methodologies, not a sign of competitive performance. For users seeking maximum throughput, the RTX 2070 is the only choice.