AMD Radeon R7 350 vs NVIDIA GeForce GTX 1660 Ti Comparison

The NVIDIA GeForce GTX 1660 Ti and the AMD Radeon R7 350 represent two starkly different eras of GPU design, yet their average benchmark scores place them surprisingly close in the database’s overall hierarchy. The GTX 1660 Ti, built on Turing architecture, posts a significantly higher average benchmark score of 7,723, while the R7 350, based on GCN 1.0, trails at 7,425. However, the data reveals that this overall parity masks dramatic swings in specific workloads, with each card claiming a decisive victory in one of the two head-to-head tests.

FAQ

Q: Which card has the higher average benchmark score?

A: The NVIDIA GeForce GTX 1660 Ti leads with an average benchmark score of 7,723, placing it 3.9% ahead of the AMD Radeon R7 350’s 7,425 average.

Q: How do the two cards compare in Geekbench OpenCL performance?

A: The GTX 1660 Ti dominates this test with a score of 51,029, which is 554.9% higher than the R7 350’s 7,792.

Q: Does the AMD Radeon R7 350 win any benchmark outright?

A: Yes, the R7 350 wins the Geekbench Vulkan test with a score of 7,057, outperforming the GTX 1660 Ti’s 5,827 by 17.4%.

Q: What is the percentile ranking for each GPU against all other GPUs?

A: The GTX 1660 Ti sits at the 41st percentile, while the R7 350 is at the 40th percentile, indicating they are in a similar overall performance tier despite their architectural differences.

Q: What are the closest rivals to the GTX 1660 Ti by average score?

A: The AMD Radeon 540 is the nearest rival with an average score of 7,673 (0.7% slower), followed by the AMD Radeon Pro WX 3100 at 7,580 (1.9% slower) and the AMD Radeon R7 250 at 7,557 (2.2% slower).

Q: What is the closest rival to the R7 350 by average score?

A: The Intel UHD Graphics 750 is the nearest rival with an average score of 7,441, which is actually 0.2% slower than the R7 350’s 7,425, making the R7 350 the faster card in that comparison.

Architecture Differences

The two GPUs are built on fundamentally different fabrication and design philosophies. The NVIDIA GeForce GTX 1660 Ti uses the TU116 chip fabricated on a 12 nm process at TSMC, packing 6,600 million transistors into a 284 mm² die. This yields a transistor density of 23.2M per mm². In contrast, the AMD Radeon R7 350 uses the older Cape Verde chip on a 28 nm process, also from TSMC, with just 1,500 million transistors on a much smaller 123 mm² die, resulting in a lower density of 12.2M per mm².

The GTX 1660 Ti’s Turing architecture brings a substantial increase in raw compute resources. It features 1,536 shading units, 96 texture mapping units (TMUs), and 48 raster output units (ROPs). The R7 350, using GCN 1.0, is configured with 512 shading units, 32 TMUs, and only 16 ROPs. This 3x difference in shading units and TMUs, and a 3x difference in ROPs, explains the massive gap in theoretical throughput rates.

Clock speeds also differ significantly. The GTX 1660 Ti operates with a base clock of 1500 MHz and a boost clock of 1770 MHz, while the R7 350 has no listed base or boost clocks, relying instead on a memory clock of 1125 MHz (4.5 Gbps effective). Memory configurations are worlds apart: the GTX 1660 Ti offers 6 GB of GDDR6 on a 192-bit bus, delivering 288.0 GB/s of bandwidth, whereas the R7 350 provides only 2 GB of GDDR5 on a 128-bit bus, yielding 72.00 GB/s. This four-fold bandwidth advantage for the NVIDIA card is critical for modern texture-heavy workloads.

Feature support also diverges. The GTX 1660 Ti supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The R7 350 supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The newer Vulkan revision on the NVIDIA card may influence API-specific performance, as seen in the benchmark results. Power characteristics are equally distinct: the GTX 1660 Ti has a 120 W TDP and requires a single 8-pin power connector, while the R7 350 sips power at just 55 W TDP with no external power connector needed.

Head-to-Head Benchmarks

The database includes only two direct head-to-head comparisons, and they tell a story of extreme specialization. The Geekbench OpenCL test is a complete rout in favor of the NVIDIA GeForce GTX 1660 Ti. The NVIDIA card scores 51,029 against the AMD card’s 7,792, a delta of 554.9%. This staggering margin reflects the GTX 1660 Ti’s superior FP32 compute throughput of 5.437 TFLOPS compared to the R7 350’s 819.2 GFLOPS. In this test, the GTX 1660 Ti is not just faster; it is in a different performance class entirely, leveraging its 1,536 shading units and 288 GB/s memory bandwidth to process far more parallel operations per second.

The Geekbench Vulkan test flips the script. Here, the AMD Radeon R7 350 takes the win with a score of 7,057, beating the GTX 1660 Ti’s 5,827 by 17.4%. This is a counterintuitive result given the hardware disparity, but it highlights how driver optimization and architecture-specific instruction scheduling can override raw specs. The R7 350’s GCN architecture, despite its age, appears to handle this particular Vulkan workload more efficiently than the Turing-based GTX 1660 Ti. The result is particularly notable because the GTX 1660 Ti lists a higher Vulkan API version (1.4 vs 1.2.170), yet still falls behind in execution.

Looking at the broader benchmark suite, the GTX 1660 Ti shows its strength across legacy and modern DirectX tests. It scores 66 in Passmark DirectX 10, 103 in DirectX 11, 51 in DirectX 12, and 190 in DirectX 9. The R7 350 has no corresponding scores in these tests, making direct comparisons impossible. In synthetic compute and graphics tests, the GTX 1660 Ti posts a Passmark G3D score of 12,871 and a G2D score of 804, with a GPU compute score of 5,226. These numbers, combined with its 3DMark Steel Nomad DX12 score of 1,061, paint a picture of a card that is consistently strong across varied workloads.

The Verdict

The data presents a clear split decision. For users prioritizing raw compute performance, the NVIDIA GeForce GTX 1660 Ti is the unequivocal choice. Its 554.9% lead in OpenCL and its 41st percentile ranking versus the R7 350’s 40th percentile show that it delivers more consistent high-end performance across most metrics. The GTX 1660 Ti’s 6 GB GDDR6 memory and 288 GB/s bandwidth make it suitable for modern applications that demand high texture throughput and large frame buffers.

However, the AMD Radeon R7 350 is not without merit. Its 17.4% victory in the Vulkan test suggests that for specific API workloads, it can outperform a much more powerful card. This makes it an interesting option for scenarios where Vulkan-based applications are the primary use case, and where the lower 55 W TDP and single-slot form factor are advantageous. The R7 350’s lack of a power connector also simplifies installation in systems with limited power delivery.

The GTX 1660 Ti’s launch MSRP was 279 USD. Neither card is in current production, with both marked as end-of-life. The GTX 1660 Ti offers a longer feature list with DirectX 12_1 support and DisplayPort 1.4a, while the R7 350 provides only DisplayPort 1.2 and HDMI 1.4a. For most users, the GTX 1660 Ti’s overwhelming compute advantage and greater memory capacity make it the superior all-rounder, but the R7 350’s Vulkan performance ensures it remains relevant in niche scenarios.

Specification Differences

The following table highlights the key differences between the two cards based solely on the provided data:

• Process Node: The GTX 1660 Ti uses a 12 nm process, while the R7 350 uses a 28 nm process.
• Transistors: The GTX 1660 Ti has 6,600 million transistors, compared to the R7 350’s 1,500 million.
• Die Size: The GTX 1660 Ti’s die is 284 mm², versus 123 mm² for the R7 350.
• Transistor Density: The GTX 1660 Ti achieves 23.2M / mm², while the R7 350 has 12.2M / mm².
• Base Clock: The GTX 1660 Ti has a base clock of 1500 MHz; the R7 350 has no listed base clock.
• Boost Clock: The GTX 1660 Ti has a boost clock of 1770 MHz; the R7 350 has no listed boost clock.
• Memory Clock: The GTX 1660 Ti runs at 1500 MHz (12 Gbps effective), while the R7 350 runs at 1125 MHz (4.5 Gbps effective).
• Memory Size: The GTX 1660 Ti has 6 GB, while the R7 350 has 2 GB.
• Memory Type: The GTX 1660 Ti uses GDDR6, while the R7 350 uses GDDR5.
• Memory Bus Width: The GTX 1660 Ti has a 192-bit bus, while the R7 350 has a 128-bit bus.
• Memory Bandwidth: The GTX 1660 Ti delivers 288.0 GB/s, compared to 72.00 GB/s for the R7 350.
• Shading Units: The GTX 1660 Ti has 1,536, while the R7 350 has 512.
• TMUs: The GTX 1660 Ti has 96, while the R7 350 has 32.
• ROPs: The GTX 1660 Ti has 48, while the R7 350 has 16.
• Pixel Rate: The GTX 1660 Ti achieves 84.96 GPixel/s, versus 12.80 GPixel/s for the R7 350.
• Texture Rate: The GTX 1660 Ti achieves 169.9 GTexel/s, versus 25.60 GTexel/s for the R7 350.
• FP32 Performance: The GTX 1660 Ti has 5.437 TFLOPS, while the R7 350 has 819.2 GFLOPS.
• FP16 Performance: The GTX 1660 Ti has 10.87 TFLOPS (2:1); the R7 350 has no listed FP16 performance.
• TDP: The GTX 1660 Ti has a TDP of 120 W, while the R7 350 has a TDP of 55 W.
• Slot Width: The GTX 1660 Ti is dual-slot, while the R7 350 is single-slot.
• Power Connectors: The GTX 1660 Ti requires 1x 8-pin, while the R7 350 requires none.
• Suggested PSU: The GTX 1660 Ti suggests a 300 W PSU, while the R7 350 suggests a 250 W PSU.
• Display Outputs: The GTX 1660 Ti has 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a; the R7 350 has 1x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2.
• DirectX Support: The GTX 1660 Ti supports 12 (12_1), while the R7 350 supports 12 (11_1).
• Vulkan Support: The GTX 1660 Ti supports 1.4, while the R7 350 supports 1.2.170.
• Dimensions: The GTX 1660 Ti is 229 mm long, 111 mm high, and 35 mm wide; the R7 350 is 168 mm long with no listed height or width.
• Release Date: The GTX 1660 Ti was released on 2019-02-21, while the R7 350 was released on 2016-07-05.