NVIDIA GeForce GTX 1660 SUPER vs NVIDIA GeForce GTX 965M Comparison

Head-to-Head Benchmarks

The benchmark data reveals a stark performance gap between these two NVIDIA parts. Across the two shared tests, the GeForce GTX 1660 SUPER records a clean sweep, winning both Geekbench submissions by margins that are difficult to overlook. The comparisons are not subtle; they are decisive shifts in compute capability.

In the Geekbench OpenCL test, the GTX 965M scores 14,509 points. The GTX 1660 SUPER, by contrast, produces 52,490 points. That is a delta of 72.4% in favor of the newer card. To put that in perspective, the GTX 1660 SUPER is not merely faster; it is operating in a different performance class entirely. The 965M’s score places it near the AMD Radeon RX Vega 11, which averages 14,385 points, a difference of only 0.1%. The 1660 SUPER, meanwhile, sits alongside the NVIDIA GeForce RTX 3050 Ti Mobile (12,940 average, 0.4% delta) and the AMD Radeon RX 580 (12,928 average, 0.4% delta). The gap between the two benchmarked cards is roughly equivalent to the entire performance range of the 1660 SUPER’s nearest rivals.

The Vulkan results tell a similar story, though the margin is even wider. The GTX 965M scores 14,299 points in Geekbench Vulkan. The GTX 1660 SUPER scores 57,102 points. That represents a delta of 75%. In this API, the 1660 SUPER is exactly three times faster in raw score terms, 57,102 versus 14,299. The 965M’s Vulkan performance is nearly identical to its OpenCL result, suggesting the Maxwell architecture is not gaining any particular advantage from the newer API. The 1660 SUPER, however, shows a modest improvement in Vulkan over OpenCL, 57,102 versus 52,490, indicating the Turing architecture scales slightly better with this workload.

The wins are not distributed in any way that favors the older card. The GTX 965M records zero benchmark wins in the head-to-head comparison, while the GTX 1660 SUPER takes both. The average benchmark score for the 965M is 14,404, derived from its two Geekbench results. The 1660 SUPER’s average is 12,986, but this figure includes a broader set of tests, including Passmark DirectX 9, 10, 11, and 12, G2D, G3D, and compute workloads, plus 3DMark Steel Nomad DX12. That 3DMark score of 1,278 and the Passmark G3D score of 12,699 reflect a card that is designed for sustained gaming workloads, not just compute tasks. The 965M has no comparable DirectX or 3DMark entries in the database, so its average is computed from a narrower sample.

What is curious is the percentile ranking. The GTX 965M holds a 56th percentile position among all GPUs, while the GTX 1660 SUPER sits at the 53rd percentile. This is counterintuitive given the massive performance delta in the shared benchmarks. The explanation lies in the composition of the database: the 965M’s percentile is based on its two strong Geekbench results, while the 1660 SUPER’s percentile incorporates a wider range of tests, including DirectX 9 and 10 legacy workloads where its scores are relatively low (189 for DX9, 65 for DX10). The 1660 SUPER is clearly the faster card in modern workloads, but its percentile is dragged down by older API tests that do not reflect its strengths.

FAQ

Q: Which card wins the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce GTX 1660 SUPER wins with a score of 52,490, compared to the GTX 965M’s 14,509. The delta is 72.4% in favor of the 1660 SUPER.

Q: How large is the Vulkan performance gap?

A: The GTX 1660 SUPER scores 57,102 in Geekbench Vulkan, while the GTX 965M scores 14,299. That is a 75% delta, making the newer card exactly three times faster in this test.

Q: Does the GTX 965M win any benchmark in the head-to-head comparison?

A: No. The GTX 965M records zero wins across the two shared tests. The GTX 1660 SUPER wins both.

Q: How do the two cards compare to their nearest rivals?

A: The GTX 965M’s average score of 14,404 puts it 0.1% ahead of the AMD Radeon RX Vega 1 (14,385) and 0.2% ahead of the NVIDIA GeForce GTX TITAN (14,373). The GTX 1660 SUPER’s average score of 12,986 places it 0.4% behind the RTX 1660 SUPER’s nearest rival, the RTX 3050 Ti Mobile (12,940), and 0.4% behind the AMD Radeon RX 580 (12,928).

Q: What is the average benchmark score for each card?

A: The GTX 965M has an average benchmark score of 14,404. The GTX 1660 SUPER has an average of 12,986, which includes a broader range of tests such as Passmark DirectX 9 through 12, G2D, G3D, compute, and 3DMark Steel Nomad.

Q: Why does the GTX 965M have a higher percentile rank despite losing both head-to-head tests?

A: The GTX 965M sits at the 56th percentile of all GPUs, while the GTX 1660 SUPER is at the 53rd. The 965M’s percentile is based solely on two strong Geekbench scores, whereas the 1660 SUPER’s percentile incorporates legacy DirectX 9 and 10 tests where it scores lower (189 and 65 respectively), pulling its overall standing down.

The Verdict

The data points to a single conclusion: the GeForce GTX 1660 SUPER is the superior GPU in every measured comparison. The 72.4% OpenCL delta and 75% Vulkan delta are not marginal improvements; they represent a generational leap. In the shared benchmarks, the 1660 SUPER does not just beat the 965M, it more than triples its Vulkan score. Anyone choosing between these two cards for modern compute or gaming workloads should select the 1660 SUPER without hesitation.

However, the percentile rankings complicate the picture. The 965M’s 56th percentile versus the 1660 SUPER’s 53rd suggests that, when measured against the entire database of GPUs, the older card holds its own in the specific tests where data exists. This is a narrow result, though. The 965M has only two benchmark entries, both Geekbench. The 1660 SUPER has ten entries spanning multiple APIs and workload types. The broader test coverage for the 1660 SUPER includes Passmark DirectX 9 (189), DirectX 10 (65), DirectX 11 (104), DirectX 12 (50), G2D (805), G3D (12,699), and compute (5,076). These results show a card that is well-rounded, even if its legacy API scores are low.

For users running modern applications, the 1660 SUPER is the only rational choice. For those constrained to legacy workloads or specific Geekbench testing scenarios, the 965M’s percentile rank indicates it is not obsolete, but the raw score differential is too large to ignore. The database records the 1660 SUPER as the clear winner in both shared tests, and the verdict follows the data.

Specification Differences

The two cards differ across nearly every major specification category. The GTX 965M uses the GM204 chip with Maxwell 2.0 architecture, built on a 28 nm process at TSMC. The GTX 1660 SUPER uses the TU116 chip with Turing architecture, built on a 12 nm process, also at TSMC. The transistor counts reflect this: the 965M has 5,200 million transistors on a 398 mm² die, while the 1660 SUPER has 6,600 million transistors on a smaller 284 mm² die. This yields a transistor density of 13.1 million per mm² for the 965M and 23.2 million per mm² for the 1660 SUPER.

Clock speeds show a significant jump. The 965M runs at a base clock of 924 MHz and a boost clock of 950 MHz. The 1660 SUPER runs at 1,530 MHz base and 1,785 MHz boost. Memory clocks also differ: the 965M uses 1,253 MHz (5 Gbps effective), while the 1660 SUPER uses 1,750 MHz (14 Gbps effective).

Memory specifications are substantially different. The 965M has 2 GB of GDDR5 on a 128-bit bus, providing 80.19 GB/s of bandwidth. The 1660 SUPER has 6 GB of GDDR6 on a 192-bit bus, providing 336.0 GB/s of bandwidth. That is more than four times the memory bandwidth.

The compute units also differ. The 965M has 1,024 shading units, 64 TMUs, and 32 ROPs. The 1660 SUPER has 1,408 shading units, 88 TMUs, and 48 ROPs. Pixel rate jumps from 30.40 GPixel/s to 85.68 GPixel/s, and texture rate from 60.80 GTexel/s to 157.1 GTexel/s. FP32 performance rises from 1.946 TFLOPS to 5.027 TFLOPS. The 1660 SUPER also lists FP16 performance of 10.05 TFLOPS (2:1), a figure absent for the 965M.

Power and physical specifications diverge as well. The 965M is an MXM Module with no power connectors, while the 1660 SUPER is a Dual-slot card with a 1x 8-pin connector and a TDP of 125 W. The 1660 SUPER lists a suggested PSU of 300 W. The 965M uses an MXM-B (3.0) bus interface, while the 1660 SUPER uses PCIe 3.0 x16. Display outputs differ: the 965M is portable device dependent, while the 1660 SUPER offers 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a. The 1660 SUPER has recorded dimensions of 229 mm length, 111 mm height, and 35 mm width. The 965M has no recorded dimensions.

Architecture Differences

The architectural gap between these two GPUs spans two NVIDIA generations. The GTX 965M is built on Maxwell 2.0, released in the GeForce 900M series. The GTX 1660 SUPER is built on Turing, part of the GeForce 16 series. The process node shrank from 28 nm to 12 nm, allowing the 1660 SUPER to pack more transistors (6,600 million versus 5,200 million) into a smaller die (284 mm² versus 398 mm²). The transistor density nearly doubled, from 13.1 million per mm² to 23.2 million per mm².

The memory architecture changed fundamentally. The 965M uses GDDR5 with a 128-bit bus, while the 1660 SUPER uses GDDR6 with a 192-bit bus. This is not just a capacity increase from 2 GB to 6 GB; the bandwidth improvement from 80.19 GB/s to 336.0 GB/s suggests a different memory subsystem design. The 1660 SUPER’s memory clock of 1,750 MHz (14 Gbps effective) is substantially higher than the 965M’s 1,253 MHz (5 Gbps effective).

The compute architecture also evolved. The 1660 SUPER has 1,408 shading units versus 1,024, 88 TMUs versus 64, and 48 ROPs versus 32. The FP32 throughput of 5.027 TFLOPS is more than 2.5 times the 965M’s 1.946 TFLOPS. Notably, the 1660 SUPER lists FP16 performance of 10.05 TFLOPS (2:1), implying a rate that the Maxwell-based 965M does not document. Neither card includes ray tracing cores or tensor cores, so those features are absent from both.

Both cards share the same API support: DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The production status for both is end-of-life. The 965M was released in January 2015, while the 1660 SUPER was released in October 2019. The 965M’s predecessor is the GeForce 800M and its successor is GeForce 10 Mobile. The 1660 SUPER’s predecessor is GeForce 10 and its successor is GeForce 20.

Where Each One Wins

The GTX 1660 SUPER wins decisively in every benchmark category where both cards have data. In Geekbench OpenCL, it leads by 72.4%. In Geekbench Vulkan, it leads by 75%. These are the only two direct comparisons available, and the 1660 SUPER dominates both. The card’s broader benchmark suite, including Passmark G3D (12,699), Passmark compute (5,076), and 3DMark Steel Nomad DX12 (1,278), indicates it is capable across modern gaming and compute workloads. Its 6 GB of GDDR6 memory and 336.0 GB/s bandwidth make it suitable for higher-resolution textures and data-heavy tasks.

The GTX 965M’s wins are more contextual. It holds a higher percentile rank (56th versus 53rd) among all GPUs, based on its limited benchmark sample. Its average score of 14,404 is within 0.1% of the AMD Radeon RX Vega 11 and 0.2% of the NVIDIA GeForce GTX TITAN, suggesting it remains competitive with those specific parts in Geekbench workloads. The 965M is an MXM module with no power connectors, making it a candidate for portable or compact systems where the dual-slot, 125 W, 1x 8-pin 1660 SUPER would not fit. The 965M’s launch MSRP was 229 USD, a figure the database records once for the 1660 SUPER, though the 965M has no corresponding launch price.

For users prioritizing raw performance in shared benchmarks, the 1660 SUPER is the clear pick. For users constrained by form factor or power delivery, the 965M’s MXM design and lack of power connectors may hold appeal, but the performance cost is severe: a 72.4% to 75% deficit in the only direct comparisons. The data does not support choosing the 965M on performance grounds. It only becomes viable in scenarios where the 1660 SUPER’s physical requirements cannot be met, and even then, the performance gap will be felt in every workload.