NVIDIA GeForce GTX 1660 Ti vs NVIDIA GeForce GTX 750 Comparison

The NVIDIA GeForce GTX 1660 Ti and the NVIDIA GeForce GTX 750 are separated by a full product generation and a massive technological gap. The GTX 1660 Ti, built on Turing architecture, is a modern 12 nm part with 6 GB of GDDR6 memory, while the GTX 750 is a Maxwell-era 28 nm card with just 1 GB of GDDR5. The data shows a clear generational divide, but the GTX 750 surprisingly wins one of the two directly comparable head-to-head benchmark tests, making this a nuanced comparison rather than a simple sweep.

FAQ

Q: Which card has the higher average benchmark score?

A: The GTX 1660 Ti holds the higher average benchmark score at 7723, compared to the GTX 750's 7222. This places the 1660 Ti at the 41st percentile of all GPUs, while the 750 sits just below at the 40th percentile.

Q: How do the two cards compare on memory capacity and type?

A: The GTX 1660 Ti uses 6 GB of GDDR6 memory on a 192-bit bus, delivering 288.0 GB/s of bandwidth. The GTX 750 uses 1024 MB (1 GB) of GDDR5 on a 128-bit bus, with a much lower bandwidth of 80.19 GB/s.

Q: Which card wins in the Geekbench Vulkan test?

A: The GTX 750 wins the Geekbench Vulkan test with a score of 8078, beating the GTX 1660 Ti's score of 5827. This represents a 27.9% advantage for the older card in this specific workload.

Q: What are the power requirements for each card?

A: The GTX 1660 Ti has a TDP of 120 W and requires a single 8-pin power connector, with a suggested PSU of 300 W. The GTX 750 has a TDP of just 55 W and requires no power connectors, with a suggested PSU of 250 W.

Q: What is the difference in shading units and texture mapping units?

A: The GTX 1660 Ti has 1536 shading units and 96 TMUs, while the GTX 750 has only 512 shading units and 32 TMUs. This gives the 1660 Ti three times the raw shader count of the 750.

Q: Which card supports a higher DirectX version?

A: The GTX 1660 Ti supports DirectX 12 (12_1), while the GTX 750 only supports DirectX 12 (11_0). Both cards support OpenGL 4.6 and Vulkan 1.4.

Where Each One Wins

The GTX 1660 Ti is the clear winner in raw compute and modern API performance. In the Geekbench OpenCL test, it scores 51029 versus the GTX 750's 9315, a staggering 447.8% advantage. This indicates that for compute-heavy workloads, content creation, or any task leveraging OpenCL, the 1660 Ti is in a completely different league. The 1660 Ti also dominates in the Passmark suite, scoring 12871 in G3D, 190 in DirectX 9, 103 in DirectX 11, and 51 in DirectX 12, while the GTX 750 has no such benchmark scores recorded in the data.

The GTX 750 wins the Geekbench Vulkan test with a score of 8078 against the 1660 Ti's 5827, a 27.9% lead. This is an unusual result, suggesting that in this specific Vulkan workload, the older Maxwell architecture outperforms Turing. The 750 also has a Geekbench Metal score of 4274, though the 1660 Ti has no Metal score listed for comparison. For users running Vulkan-based applications that favor the 750's specific driver path, the older card holds a genuine advantage.

In practical terms, the 1660 Ti wins for gaming at higher resolutions, texture-heavy titles, and any task requiring more than 1 GB of VRAM. The GTX 750 wins only in the narrow scenario of the specific Vulkan workload measured here, plus it offers the benefit of lower power draw and a smaller physical footprint.

Architecture Differences

The architectural divide between these two cards is generational. The GTX 1660 Ti uses the TU116 chip on TSMC's 12 nm process, packing 6,600 million transistors into a 284 mm² die. This results in a transistor density of 23.2M per mm². It is built on the Turing architecture, which is a direct successor to the Pascal line, though it lacks the RT cores and tensor cores found in the higher-end GeForce 20-series.

The GTX 750 uses the GM107 chip on TSMC's 28 nm process, with 1,870 million transistors on a 148 mm² die. Its transistor density is 12.6M per mm², roughly half that of the 1660 Ti. The Maxwell architecture was designed for power efficiency, and the 750's 55 W TDP reflects that. The 1660 Ti's TDP of 120 W is more than double, but it delivers far more performance per watt in most workloads.

The memory subsystems are fundamentally different. The 1660 Ti uses GDDR6 with a 192-bit bus, achieving 288.0 GB/s of bandwidth. The 750 uses GDDR5 with a 128-bit bus, capped at 80.19 GB/s. The 1660 Ti also has a much higher memory clock at 1500 MHz (12 Gbps effective) versus 1253 MHz (5 Gbps effective) on the 750. The 1660 Ti's pixel rate of 84.96 GPixel/s and texture rate of 169.9 GTexel/s dwarf the 750's 17.36 GPixel/s and 34.72 GTexel/s.

Specification Differences

The two cards differ in nearly every specification field. The GTX 1660 Ti has a base clock of 1500 MHz and a boost clock of 1770 MHz, while the GTX 750 runs at 1020 MHz base and 1085 MHz boost. The 1660 Ti has 1536 shading units, 96 TMUs, and 48 ROPs, versus the 750's 512 shading units, 32 TMUs, and 16 ROPs.

Memory is a major differentiator: the 1660 Ti offers 6 GB of GDDR6 with 288.0 GB/s bandwidth, while the 750 has 1 GB of GDDR5 with 80.19 GB/s. The FP32 compute performance is 5.437 TFLOPS for the 1660 Ti versus 1,111.0 GFLOPS for the 750. The 1660 Ti also has FP16 performance of 10.87 TFLOPS (2:1), while the 750 has no FP16 capability listed.

Physical specifications diverge significantly. The 1660 Ti is a dual-slot card measuring 229 mm in length, 111 mm in height, and 35 mm in width, requiring a 1x 8-pin power connector. The 750 is a single-slot card at 145 mm length, with no power connector needed. The 750 has a lower TDP of 55 W versus 120 W, and a lower suggested PSU of 250 W versus 300 W. Display outputs also differ: the 1660 Ti has 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a, while the 750 has 2x DVI and 1x mini-HDMI 1.4a.

Head-to-Head Benchmarks

The only two directly comparable benchmark results between these cards are in Geekbench OpenCL and Geekbench Vulkan. In OpenCL, the GTX 1660 Ti scores 51029 against the GTX 750's 9315. This is a 447.8% difference, meaning the 1660 Ti delivers roughly 5.5 times the OpenCL performance of the 750. This is the largest single-specification gap between the two cards, reflecting the massive difference in shading units, memory bandwidth, and compute capability.

In Vulkan, the result is reversed. The GTX 750 scores 8078, beating the 1660 Ti's 5827 by 27.9%. This is a surprising outcome, as the 1660 Ti is a much more powerful card on paper. The deltaPct of -27.9 indicates the 1660 Ti is behind in this test. Possible explanations include driver optimization differences for the older Maxwell architecture in this specific Vulkan path, or the test being more sensitive to latency than raw throughput.

Looking at the broader benchmark picture, the 1660 Ti has a Passmark G3D score of 12871, while the 750 has no Passmark scores listed. The 1660 Ti also scores 804 in Passmark G2D, 5226 in GPU compute, and various DirectX scores ranging from 51 in DX12 to 190 in DX9. The 750's only other benchmark is Geekbench Metal at 4274, which the 1660 Ti does not have a comparable score for. The wins are split evenly at 1-1 in the head-to-head tests, but the magnitude of the 1660 Ti's OpenCL win far exceeds the 750's Vulkan win.

The Verdict

The data points to a clear choice for most users: the GTX 1660 Ti is the superior card for essentially all modern workloads. Its 447.8% lead in OpenCL and its substantial lead in Passmark G3D (12871 vs. no comparable score) demonstrate that it is in a different performance class. The 1660 Ti's 6 GB of GDDR6 memory versus 1 GB of GDDR5 alone disqualifies the 750 for any modern gaming or content creation task, as 1 GB is insufficient for current titles. The 1660 Ti's higher percentile ranking (41st vs. 40th) and higher average score (7723 vs. 7222) reinforce this.

The GTX 750's sole victory in Vulkan is noteworthy but narrow. It suggests that for users running a specific legacy Vulkan application where the 750's driver path is better optimized, the older card could offer a smoother experience. However, the 750's 55 W TDP and lack of power connectors make it an attractive option for low-power builds or as a simple display adapter. Its single-slot, 145 mm length also fits in cases where the 1660 Ti's 229 mm dual-slot design would not.

The choice depends on the use case. For gaming, content creation, compute, or any future-proofing needs, the GTX 1660 Ti is the only rational pick. Its 5.437 TFLOPS FP32 performance and 288.0 GB/s bandwidth are essential for modern software. For a budget, low-power office machine or a legacy system where Vulkan performance is the sole priority, the GTX 750 might suffice, but its 1 GB memory limit will be a hard ceiling. The data overwhelmingly favors the 1660 Ti for anyone who needs actual graphics performance.