AMD Radeon RX 9070 XT vs NVIDIA GeForce GTX 1660 Comparison

Head-to-Head Benchmarks

The benchmark data presents a clear overall picture: the AMD Radeon RX 9070 XT wins 9 of the 10 recorded head-to-head comparisons, with the NVIDIA GeForce GTX 1660 taking a single, notable victory. The most decisive margin appears in the 3DMark Steel Nomad DX12 test, where the RX 9070 XT scores 7260 against the GTX 1660’s 1065, a delta of 581.7%. That result is not just a win; it is a generational gap expressed in raw performance terms. Similarly, the Passmark DirectX 11 test shows the RX 9070 XT at 292 versus 79, a 269.6% advantage, and the Passmark GPU Compute test shows 15864 against 4963, a 219.6% lead. These are not marginal improvements; they are multiples of performance.

The GTX 1660’s sole win comes in Geekbench OpenCL, where it posts 47850 against the RX 9070 XT’s 17428, a delta of -63.6% in favor of the older NVIDIA card. This is a striking anomaly. The RX 9070 XT dominates in Vulkan, scoring 65879 to 50137, a 31.4% lead, yet the OpenCL result flips the script entirely. The data suggests that the GTX 1660’s compute path in OpenCL is disproportionately strong relative to its other scores, or that the RX 9070 XT’s OpenCL driver implementation lags its Vulkan and DirectX performance. In every other recorded test, the RX 9070 XT leads by margins ranging from 59.2% in Passmark DirectX 12 (78 vs 49) to 152.5% in Passmark DirectX 10 (154 vs 61), 130.1% in Passmark G3D (26795 vs 11646), 99.4% in Passmark DirectX 9 (353 vs 177), and 71.1% in Passmark G2D (1328 vs 776).

The average benchmark score reflects this asymmetry: the RX 9070 XT averages 13543, while the GTX 1660 averages 11680. That is a 15.9% gap in average scores, but the head-to-head deltas are far larger in most individual tests, indicating that the average is pulled down by the OpenCL outlier. The percentile ranks are closer than the raw scores suggest: the RX 9070 XT sits at the 55th percentile among all GPUs, while the GTX 1660 sits at the 51st. This implies that the RX 9070 XT’s 7260 Steel Nomad score is exceptional, but its overall benchmark profile places it only modestly above the GTX 1660 in percentile terms. The data does not support the idea that the RX 9070 XT is a universal powerhouse; rather, it is a specialist in DirectX and Vulkan workloads, with a notable weakness in OpenCL compute.

Architecture Differences

The two cards come from different eras and different design philosophies. The AMD Radeon RX 9070 XT uses the Navi 48 chip built on RDNA 4.0 architecture, fabricated on a 4 nm process at TSMC. It packs 53,900 million transistors into a 357 mm² die, yielding a transistor density of 151.0M per mm². The NVIDIA GeForce GTX 1660 uses the TU116 chip on Turing architecture, built on a 12 nm process, also at TSMC, with 6,600 million transistors on a 284 mm² die, giving a density of 23.2M per mm². The RX 9070 XT is not just newer; it is an order of magnitude more complex, with roughly eight times the transistor count.

Memory configuration diverges sharply. The RX 9070 XT has 16 GB of GDDR6 on a 256-bit bus, delivering 644.6 GB/s of bandwidth. The GTX 1660 has 6 GB of GDDR5 on a 192-bit bus, with 192.1 GB/s. That is a 3.4x difference in bandwidth and a 2.7x difference in capacity. The RX 9070 XT’s memory clock runs at 2518 MHz (20.1 Gbps effective), while the GTX 1660’s memory clock is 2001 MHz (8 Gbps effective). The RX 9070 XT also has far more execution resources: 4096 shading units, 256 texture mapping units, and 128 ROPs, against the GTX 1660’s 1408 shading units, 88 TMUs, and 48 ROPs. The RX 9070 XT includes 64 ray tracing cores; the GTX 1660 has none. Neither card has tensor cores, but the RX 9070 XT’s FP32 throughput is 48.66 TFLOPS, while the GTX 1660’s is 5.027 TFLOPS. The RX 9070 XT’s FP16 is 48.66 TFLOPS (1:1), while the GTX 1660’s is 10.05 TFLOPS (2:1).

Pixel and texture rates follow the same pattern. The RX 9070 XT posts 380.2 GPixel/s and 760.3 GTexel/s, versus 85.68 GPixel/s and 157.1 GTexel/s for the GTX 1660. The RX 9070 XT uses PCIe 5.0 x16, while the GTX 1660 uses PCIe 3.0 x16. Display outputs differ: the RX 9070 XT has 1x HDMI 2.1b and 3x DisplayPort 2.1a, while the GTX 1660 has 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a. The RX 9070 XT supports DirectX 12 Ultimate (12_2), while the GTX 1660 supports DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4. The RX 9070 XT’s TDP is 304 W, requiring a 700 W PSU and 2x 8-pin connectors, while the GTX 1660’s TDP is 120 W, needing a 300 W PSU and 1x 8-pin connector. The RX 9070 XT is marked as Active production; the GTX 1660 is End-of-life.

FAQ

Q: Which card wins in 3DMark Steel Nomad DX12?

A: The AMD Radeon RX 9070 XT scores 7260, against the NVIDIA GeForce GTX 1660’s 1065, a 581.7% advantage.

Q: Is there any test where the GTX 1660 beats the RX 9070 XT?

A: Yes, in Geekbench OpenCL, the GTX 1660 scores 47850 versus 17428, a 63.6% margin in its favor.

Q: How much faster is the RX 9070 XT in Vulkan?

A: The RX 9070 XT scores 65879 in Geekbench Vulkan, versus 50137 for the GTX 1660, a 31.4% lead.

Q: What is the memory bandwidth difference?

A: The RX 9070 XT has 644.6 GB/s of bandwidth, while the GTX 1660 has 192.1 GB/s, a 3.4x gap.

Q: Does the GTX 1660 have ray tracing cores?

A: No, the GTX 1660 has no ray tracing cores; the RX 9070 XT has 64.

Q: What are the average benchmark scores?

A: The RX 9070 XT averages 13543, while the GTX 1660 averages 11680.

The Verdict

The data indicates that the AMD Radeon RX 9070 XT is the superior card for almost all recorded workloads. It wins 9 of 10 head-to-head tests, with margins that are often several hundred percent. The 3DMark Steel Nomad DX12 result, the Passmark DirectX 11 result, and the Passmark GPU Compute result are the strongest evidence of its dominance. For users running DirectX 9, 10, 11, or 12 titles, or Vulkan applications, the RX 9070 XT is the clear choice. Its 16 GB memory and 644.6 GB/s bandwidth also give it a substantial capacity and throughput advantage over the GTX 1660’s 6 GB and 192.1 GB/s.

The GTX 1660 is not without merit, but its only recorded win is in Geekbench OpenCL, where it outperforms by 63.6%. That is a significant margin, and it suggests that for OpenCL compute workloads, the older NVIDIA card may be preferable. However, the GTX 1660’s overall percentile rank (51st) is only slightly below the RX 9070 XT’s (55th), meaning that in the broader GPU landscape, both cards occupy a similar mid-tier position. The RX 9070 XT’s average score is higher, but its OpenCL weakness drags its average down relative to its DirectX and Vulkan peaks. The GTX 1660 is End-of-life, while the RX 9070 XT is Active, which matters for long-term driver support and availability. The RX 9070 XT requires a 700 W PSU and 2x 8-pin connectors, versus 300 W and 1x 8-pin for the GTX 1660, so the RX 9070 XT demands a more substantial power supply.

Specification Differences

| Field | AMD Radeon RX 9070 XT | NVIDIA GeForce GTX 1660 |

|-------|----------------------|--------------------------|

| Process Node | 4 nm | 12 nm |

| Transistors | 53,900 million | 6,600 million |

| Die Size | 357 mm² | 284 mm² |

| Transistor Density | 151.0M / mm² | 23.2M / mm² |

| Base Clock | 1660 MHz | 1530 MHz |

| Boost Clock | 2970 MHz | 1785 MHz |

| Game Clock | 2400 MHz | N/A |

| Memory Clock | 2518 MHz (20.1 Gbps effective) | 2001 MHz (8 Gbps effective) |

| Memory Size | 16 GB | 6 GB |

| Memory Type | GDDR6 | GDDR5 |

| Memory Bus Width | 256 bit | 192 bit |

| Memory Bandwidth | 644.6 GB/s | 192.1 GB/s |

| Shading Units | 4096 | 1408 |

| TMUs | 256 | 88 |

| ROPs | 128 | 48 |

| RT Cores | 64 | None |

| Pixel Rate | 380.2 GPixel/s | 85.68 GPixel/s |

| Texture Rate | 760.3 GTexel/s | 157.1 GTexel/s |

| FP32 | 48.66 TFLOPS | 5.027 TFLOPS |

| FP16 | 48.66 TFLOPS (1:1) | 10.05 TFLOPS (2:1) |

| TDP | 304 W | 120 W |

| Power Connectors | 2x 8-pin | 1x 8-pin |

| Suggested PSU | 700 W | 300 W |

| Bus Interface | PCIe 5.0 x16 | PCIe 3.0 x16 |

| Display Outputs | 1x HDMI 2.1b, 3x DisplayPort 2.1a | 1x DVI, 1x HDMI 2.0, 1x DisplayPort 1.4a |

| DirectX | 12 Ultimate (12_2) | 12 (12_1) |

| Production Status | Active | End-of-life |

| Release Date | 2025-03-05 | 2019-03-13 |

| Launch MSRP | 599 USD | 219 USD |

Where Each One Wins

The AMD Radeon RX 9070 XT wins in every DirectX generation tested, including DirectX 9 (353 vs 177), DirectX 10 (154 vs 61), DirectX 11 (292 vs 79), and DirectX 12 (78 vs 49). It also wins in Vulkan (65879 vs 50137), 3DMark Steel Nomad DX12 (7260 vs 1065), Passmark G2D (1328 vs 776), Passmark G3D (26795 vs 11646), and Passmark GPU Compute (15864 vs 4963). The RX 9070 XT is the clear winner for gaming, especially in modern DirectX 12 and Vulkan titles, and for general GPU compute tasks that use those APIs.

The NVIDIA GeForce GTX 1660 wins in Geekbench OpenCL (47850 vs 17428). That is its only recorded victory, but it is substantial. For workloads that rely specifically on OpenCL compute, such as certain scientific simulations, video processing, or legacy applications, the GTX 1660 may be the better choice. The GTX 1660 also has a lower TDP (120 W versus 304 W), so it is more suited to systems with modest power supplies or thermal constraints. Its smaller physical footprint (229 mm length, 111 mm height, 35 mm width) may also fit in compact cases where the RX 9070 XT’s dimensions are not specified. The GTX 1660 is End-of-life, so it is not recommended for new builds, but for existing systems or specific OpenCL tasks, it retains a niche advantage.