NVIDIA GeForce GTX 1660 vs NVIDIA GeForce GTX 950 Comparison

Head-to-Head Benchmarks

The recorded data shows a decisive victory for the NVIDIA GeForce GTX 1660 across the two shared benchmark tests. The GTX 1660 wins both head-to-head comparisons, and the margins are substantial, indicating a generational leap in compute performance.

In Geekbench OpenCL, the GTX 1660 scores 47,850, while the GTX 950 manages 15,662. This represents a 67.3% advantage for the GTX 1660. To put that number in perspective, the GTX 950's score places it near rivals like the NVIDIA Tesla M2090 (13,075, a 0.9% delta) and the AMD FirePro M6100 (13,354, a 1.2% delta). The GTX 1660's OpenCL result, in contrast, sits close to the AMD Radeon RX 7800 XT (11,627, a 0.5% delta) and the AMD Radeon RX 6500 XT (11,842, a 1.4% delta). The GTX 1660 is not just faster, it is operating in a completely different performance tier.

The Vulkan results tell the same story. The GTX 1660 posts 50,137 in Geekbench Vulkan, against the GTX 950's 16,734. The delta is 66.6% in favor of the newer card. This is a massive gap. The GTX 950's Vulkan score is closer to its own OpenCL result, suggesting that its Maxwell architecture handles the compute workload in a similar fashion, but the Turing-based GTX 1660 shows a marked improvement in this API.

There are no head-to-head results where the GTX 950 wins; the win count is zero for the older card and two for the GTX 1660. The data is unambiguous. The GTX 1660 delivers more than triple the compute performance in both OpenCL and Vulkan workloads. This is not a marginal upgrade; it is a wholesale replacement in terms of raw throughput.

The average benchmark score across all recorded tests reinforces this. The GTX 950 has an average score of 13,189, while the GTX 1660's average is 11,680. This might seem contradictory at first glance, but the averages are drawn from different test suites. The GTX 950's average is based on three Geekbench results (Metal, OpenCL, Vulkan), while the GTX 1660's average includes a wider range of Passmark tests (DirectX 9 through 12, G2D, G3D, GPU Compute) and a 3DMark Steel Nomad result. The Passmark DirectX 9 score of 177 and the DirectX 11 score of 79 are low absolute numbers, which drag down the GTX 1660's average. The shared head-to-head tests are the correct basis for comparison, and they favor the GTX 1660 overwhelmingly.

The Verdict

The database analysis points to one clear conclusion: the NVIDIA GeForce GTX 1660 is the superior product for any workload that depends on compute performance. The GTX 950, while a capable card in its day, is outclassed by a wide margin in the recorded benchmarks. The 67.3% lead in OpenCL and the 66.6% lead in Vulkan are not small differences; they represent a fundamental performance gap that cannot be closed by driver optimizations or overclocking.

Who should pick the GTX 950? Based strictly on the data, very few users. Its only advantage is its smaller physical footprint. The GTX 950 measures 202 mm in length (8 inches), while the GTX 1660 is 229 mm (9 inches). For a compact build with strict length constraints, the GTX 950 might fit where the GTX 1660 cannot. It also requires less power from the system: the GTX 950 has a TDP of 90 W and a suggested PSU of 250 W, whereas the GTX 1660 has a TDP of 120 W and a suggested PSU of 300 W. If the power supply is low-capacity, the GTX 950 is the safer bet.

Who should pick the GTX 1660? Anyone who needs compute performance, which includes most modern gaming and productivity scenarios. The GTX 1660 has more than double the shading units (1,408 vs. 768), more texture mapping units (88 vs. 48), and more render output units (48 vs. 32). It also has three times the memory capacity (6 GB vs. 2 GB) and nearly double the memory bandwidth (192.1 GB/s vs. 105.8 GB/s). In the recorded data, the GTX 1660 is the only reasonable choice for Vulkan-based games or OpenCL compute tasks. The GTX 950's 2 GB memory is a hard limitation for modern workloads, and its compute scores reflect that.

The percentile rankings are close: the GTX 950 sits at the 53rd percentile among all GPUs, and the GTX 1660 sits at the 51st percentile. This is misleading. The percentile is calculated against the entire database of GPUs, which includes far more powerful modern cards that push the GTX 1660 down. The head-to-head data is the relevant metric, and it shows a dominant GTX 1660 victory.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 950 has a higher average benchmark score of 13,189, compared to the GTX 1660's 11,680. However, this average is based on different test sets, and the direct head-to-head comparisons favor the GTX 1660 by a large margin.

Q: How much faster is the GTX 1660 in OpenCL?

A: The GTX 1660 scores 47,850 in Geekbench OpenCL, while the GTX 950 scores 15,662. This is a 67.3% advantage for the GTX 1660.

Q: What is the memory capacity of each card?

A: The GTX 950 has 2 GB of GDDR5 memory on a 128-bit bus, while the GTX 1660 has 6 GB of GDDR5 memory on a 192-bit bus.

Q: Which card has a higher boost clock?

A: The GTX 1660 has a boost clock of 1785 MHz, while the GTX 950 has a boost clock of 1188 MHz.

Q: Do both cards support the same DirectX version?

A: Yes, both cards support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4.

Q: What is the TDP difference between the two cards?

A: The GTX 950 has a TDP of 90 W, while the GTX 1660 has a TDP of 120 W. The suggested PSU is 250 W for the GTX 950 and 300 W for the GTX 1660.

Specification Differences

The two cards differ in nearly every measurable specification. The GTX 950 is built on a 28 nm process, while the GTX 1660 uses a 12 nm process. The GTX 950 has 2,940 million transistors on a 228 mm² die, with a transistor density of 12.9 million per mm². The GTX 1660 has 6,600 million transistors on a 284 mm² die, with a density of 23.2 million per mm².

Clock speeds are significantly higher on the GTX 1660. Its base clock is 1530 MHz and boost is 1785 MHz. The GTX 950's base clock is 1024 MHz with a boost of 1188 MHz. Memory clocks also differ: the GTX 950 runs at 1653 MHz (6.6 Gbps effective), while the GTX 1660 runs at 2001 MHz (8 Gbps effective).

Memory configuration is a major differentiator. The GTX 950 has 2 GB of GDDR5 on a 128-bit bus, delivering 105.8 GB/s of bandwidth. The GTX 1660 has 6 GB of GDDR5 on a 192-bit bus, delivering 192.1 GB/s.

Compute resources are vastly different. The GTX 950 has 768 shading units, 48 TMUs, and 32 ROPs. The GTX 1660 has 1,408 shading units, 88 TMUs, and 48 ROPs. Pixel rate is 38.02 GPixel/s for the GTX 950 versus 85.68 GPixel/s for the GTX 1660. Texture rate is 57.02 GTexel/s versus 157.1 GTexel/s. FP32 throughput is 1.825 TFLOPS for the GTX 950 and 5.027 TFLOPS for the GTX 1660. The GTX 1660 also has an FP16 rating of 10.05 TFLOPS (2:1), while the GTX 950 has no recorded FP16 rating.

Power and physical specs differ as well. The GTX 950 has a TDP of 90 W and uses a single 6-pin power connector. The GTX 1660 has a TDP of 120 W and uses a single 8-pin connector. The GTX 950 is 202 mm long (8 inches). The GTX 1660 is longer at 229 mm (9 inches), and also has recorded dimensions of 111 mm height (4.4 inches) and 35 mm width (1.4 inches). Both are dual-slot cards. Display outputs are similar: the GTX 950 has 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2, while the GTX 1660 has 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a.

Architecture Differences

The GTX 950 is based on the GM206 chip using the Maxwell 2.0 architecture, part of the GeForce 900 generation. The GTX 1660 is based on the TU116 chip using the Turing architecture, part of the GeForce 16 generation. The GTX 1660 is listed as belonging to the GeForce 10-series, despite being a Turing card, which reflects its positioning as a direct successor to the GTX 10-series lineup.

The process node is a major architectural difference. The GTX 950 uses a 28 nm process from TSMC, while the GTX 1660 uses a 12 nm process from the same foundry. This node shrink allows the GTX 1660 to pack more than double the transistors (6,600 million vs. 2,940 million) into a modestly larger die (284 mm² vs. 228 mm²).

The GTX 950's Maxwell architecture has a predecessor of GeForce 700 and a successor of GeForce 10. The GTX 1660's Turing architecture has a predecessor of GeForce 10 and a successor of GeForce 20. Neither card has ray tracing cores or tensor cores, so those features are absent from both.

The API support is identical: both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The GTX 950 was released on 2015-08-19, while the GTX 1660 was released on 2019-03-13. Both are marked as end-of-life in production status.

Where Each One Wins

The GTX 1660 wins in every compute benchmark recorded. It is the clear choice for OpenCL workloads, Vulkan-based applications, and any task that leverages the GPU's shading units and memory bandwidth. The 67.3% lead in OpenCL and 66.6% lead in Vulkan translate directly to faster execution in games that use these APIs, as well as in productivity applications that offload compute tasks to the GPU. The 6 GB memory capacity and 192.1 GB/s bandwidth give it a substantial edge in handling larger textures and datasets without spilling to system memory.

The GTX 1660 also wins on raw throughput metrics. Its FP32 performance of 5.027 TFLOPS is nearly triple the GTX 950's 1.825 TFLOPS. Its texture rate of 157.1 GTexel/s is 2.75 times higher, and its pixel rate of 85.68 GPixel/s is 2.25 times higher. In any scenario that stresses these units, the GTX 1660 will finish the job faster.

The GTX 950 has no benchmark wins in the recorded data. Its advantages are limited to physical and power characteristics. It is shorter at 202 mm, which matters for small form factor cases. It has a lower TDP of 90 W and a lower suggested PSU of 250 W, which matters for systems with modest power supplies. It uses a single 6-pin power connector, while the GTX 1660 requires a single 8-pin connector. If a build has a hard length or power ceiling, the GTX 950 is the only one of the two that fits.

In terms of use-case split, the data suggests the GTX 1660 for any performance-oriented task, and the GTX 950 only for constrained builds where its smaller size and lower power draw are non-negotiable requirements. The GTX 1660's launch MSRP is 219 USD, while the GTX 950's launch MSRP is 159 USD, but the performance gap in the recorded benchmarks justifies the difference for anyone who needs the compute power.