NVIDIA GeForce GTX 1660 Ti vs NVIDIA GeForce GTX 960 Comparison

FAQ

Q: How do the two cards compare in overall average benchmark score?

A: The GTX 960 records an average benchmark score of 9273, while the GTX 1660 Ti records 7723. Despite this, the GTX 960 sits at the 45th percentile among all GPUs, whereas the GTX 1660 Ti sits at the 41st percentile.

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

A: The GTX 1660 Ti wins decisively with a score of 1061 versus the GTX 960's 162, a delta of -84.7% from the perspective of the GTX 960. This represents the largest single-test margin in the head-to-head data.

Q: Is there any benchmark where the GTX 960 comes out ahead?

A: Yes, in Geekbench Vulkan the GTX 960 scores 9231 against the GTX 1660 Ti's 5827, a 58.4% advantage. This is the only head-to-head test the older card wins.

Q: What are the memory specifications for each card?

A: The GTX 960 has 2 GB of GDDR5 on a 128-bit bus with 112.2 GB/s bandwidth. The GTX 1660 Ti has 6 GB of GDDR6 on a 192-bit bus with 288.0 GB/s bandwidth.

Q: What is the launch MSRP for each card?

A: The GTX 960 has a launch MSRP of 199 USD, and the GTX 1660 Ti has a launch MSRP of 279 USD.

Q: Which card has a higher transistor density?

A: The GTX 1660 Ti has a transistor density of 23.2M per mm², significantly higher than the GTX 960's 12.9M per mm², reflecting the newer 12 nm process versus 28 nm.

Where Each One Wins

The head-to-head data splits cleanly across three tests. The GTX 1660 Ti dominates in two of them: 3DMark Steel Nomad DX12 and Geekbench OpenCL. In Steel Nomad, the newer card achieves 1061 points versus 162, a 6.5x raw score advantage. In OpenCL, the GTX 1660 Ti posts 51029 against 18925, roughly 2.7x higher. These are compute-heavy and modern API workloads where the Turing architecture's additional shading units, higher clocks, and faster memory produce overwhelming leads.

The GTX 960's sole victory comes in Geekbench Vulkan, where it scores 9231 against 5827. This 58.4% margin is substantial, but it is confined to a single API test. The pattern suggests the GTX 960 retains an edge in certain Vulkan workloads, possibly due to driver optimizations or specific workload characteristics, but this does not translate to broader compute or DX12 performance.

Looking at the wins tally, the GTX 1660 Ti takes 2 wins out of 3 head-to-head tests. The GTX 960 takes 1. For users prioritizing modern DX12 gaming or OpenCL compute, the GTX 1660 Ti is the clear choice. For Vulkan-specific applications where the GTX 960's score holds up, the older card remains relevant, but that relevance narrows as the rest of the benchmark suite favors the successor.

Architecture Differences

The GTX 960 is built on the GM206 chip using Maxwell 2.0 architecture, fabricated on a 28 nm process at TSMC. It packs 2,940 million transistors into a 228 mm² die, yielding a density of 12.9M per mm². The GTX 1660 Ti uses the TU116 chip with Turing architecture, manufactured on a 12 nm process, also at TSMC. It contains 6,600 million transistors across a 284 mm² die, for a density of 23.2M per mm². The transistor count more than doubles, and the density nearly doubles, reflecting the process shrink and architectural improvements.

The GTX 1660 Ti includes FP16 support at 10.87 TFLOPS (2:1 ratio), while the GTX 960 lists no FP16 capability. The GTX 1660 Ti also features a boost clock of 1770 MHz versus 1178 MHz for the GTX 960, and a base clock of 1500 MHz versus 1127 MHz. These clock increases, combined with 1536 shading units against 1024, drive the substantial compute advantage.

Neither card includes ray tracing or tensor cores, as both lack RT and tensor core fields. The GTX 1660 Ti carries 96 texture mapping units and 48 ROPs, while the GTX 960 has 64 TMUs and 32 ROPs. The newer card's texture rate is 169.9 GTexel/s versus 75.39 GTexel/s, and pixel rate is 84.96 GPixel/s versus 37.70 GPixel/s. Both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, so API compatibility is identical. The GTX 1660 Ti's display outputs include DisplayPort 1.4a, while the GTX 960 uses DisplayPort 1.2, a minor but real difference for modern monitors.

Specification Differences

The two cards diverge on nearly every core specification. The GTX 960 has 1024 shading units, 64 TMUs, and 32 ROPs, while the GTX 1660 Ti has 1536 shading units, 96 TMUs, and 48 ROPs. Clock speeds differ: the GTX 960 runs at 1127 MHz base and 1178 MHz boost, the GTX 1660 Ti at 1500 MHz base and 1770 MHz boost. Memory capacity is 2 GB GDDR5 versus 6 GB GDDR6, with bus widths of 128 bit versus 192 bit and bandwidth of 112.2 GB/s versus 288.0 GB/s. Effective memory speed is 7 Gbps for the GTX 960 and 12 Gbps for the GTX 1660 Ti.

FP32 compute is 2.413 TFLOPS for the GTX 960 and 5.437 TFLOPS for the GTX 1660 Ti, more than doubling. The GTX 1660 Ti also adds FP16 at 10.87 TFLOPS, which the GTX 960 lacks. Pixel rate is 37.70 GPixel/s versus 84.96 GPixel/s, and texture rate is 75.39 GTexel/s versus 169.9 GTexel/s. The GTX 1660 Ti uses a 1x 8-pin power connector, while the GTX 960 uses 1x 6-pin. Both have a TDP of 120 W and a suggested PSU of 300 W. Physical dimensions differ: the GTX 960 is 241 mm long and dual-slot, the GTX 1660 Ti is 229 mm long, 111 mm high, 35 mm wide, also dual-slot. The GTX 1660 Ti was released on 2019-02-21, the GTX 960 on 2015-01-21. Both are end-of-life products.

Head-to-Head Benchmarks

The 3DMark Steel Nomad DX12 test shows the most lopsided result. The GTX 1660 Ti scores 1061, the GTX 960 scores 162, a delta of -84.7% for the older card. This is not a marginal improvement; it is a 6.5x multiplier. The test likely stresses DX12 features and raw throughput, where the GTX 1660 Ti's higher shading unit count, doubled FP32 rate, and 2.5x memory bandwidth combine for an overwhelming win.

In Geekbench OpenCL, the GTX 1660 Ti scores 51029 against 18925, a delta of -62.9%. This reflects compute-heavy workloads that scale with shader count and clock speed. The GTX 1660 Ti's 1536 shading units at 1770 MHz boost produce far more parallelism than the GTX 960's 1024 units at 1178 MHz. The memory bandwidth advantage also helps feed the compute units.

The GTX 960's win in Geekbench Vulkan is notable: 9231 versus 5827, a 58.4% advantage. This is the inverse of the other two tests. The data does not explain why, but the result is consistent within the measurement. It suggests Vulkan workloads on this particular test favor the Maxwell architecture, or the GTX 960's driver path handles the specific draw calls more efficiently. This single win does not offset the magnitude of the losses, but it prevents a clean sweep.

Across all three tests, the GTX 1660 Ti's average score in the head-to-head is pulled down by the Vulkan result, yet it still wins 2 of 3. The GTX 960's average benchmark score of 9273 is higher than the GTX 1660 Ti's 7723, driven largely by its strong Vulkan and OpenCL numbers relative to the newer card's lower Vulkan score. The percentile rankings reflect this: the GTX 960 at 45th versus the GTX 1660 Ti at 41st.

The Verdict

The data points to the GTX 1660 Ti as the stronger card for modern workloads, but not without qualification. It wins the two most demanding tests, DX12 and OpenCL, with margins of 84.7% and 62.9% respectively. For any user running DX12 games or OpenCL compute, the GTX 1660 Ti is the only rational choice. Its 6 GB memory versus 2 GB also matters for texture-heavy scenes, though the benchmark data does not directly test that.

The GTX 960 retains a clear advantage in Vulkan, winning by 58.4%. For users locked into Vulkan-based applications, the older card still holds its own. However, the GTX 960's lower FP32 rate, smaller memory pool, and older architecture limit its ceiling. The GTX 1660 Ti's nearest rivals in the database are the AMD Radeon 540 (0.7% ahead), AMD Radeon Pro WX 3100 (1.9% ahead), AMD Radeon R7 250 (2.2% ahead), and Intel Arc A310 (2.3% ahead), indicating it sits in a crowded mid-range cluster. The GTX 960's nearest rivals include the GTX 465 (-0.2%), GTX 850M (-0.3%), Radeon R7 M380 (-0.4%), and Radeon Vega 8 (0.6%), a similar tier.

The verdict from the recorded data: choose the GTX 1660 Ti for DX12 and general compute, choose the GTX 960 only if Vulkan performance is the primary concern. The GTX 1660 Ti's 2-1 win record, combined with the sheer size of its victories, makes it the better overall card. The GTX 960's single Vulkan win is real but narrow in scope. Both cards are end-of-life, so the decision is about legacy use cases, not future upgrades. The GTX 1660 Ti's higher transistor count, doubled FP32, and tripled memory capacity provide a structural advantage that the GTX 960 cannot overcome outside its one niche.