NVIDIA GeForce GTX 1650 SUPER vs NVIDIA GeForce GTX 1660 SUPER Comparison

The Verdict

The data paints a clear picture: the NVIDIA GeForce GTX 1660 SUPER wins every single benchmark in this comparison, taking all 10 recorded tests. The GTX 1650 SUPER does not secure a single victory in any workload measured. For anyone choosing between these two, the GTX 1660 SUPER is the straightforward pick if performance is the priority. It holds a 17.5% higher average benchmark score (12986 versus 11047), and its smallest win margin is 7.5% in the 2D graphics test, while its largest is a 263.1% blowout in 3DMark Steel Nomad DX12.

However, the GTX 1650 SUPER is not without a role. It runs at a 100 W TDP compared to the 125 W TDP of the GTX 1660 SUPER, it uses a single 6-pin power connector instead of an 8-pin, and it carries a launch MSRP of 159 USD versus 229 USD for the larger card. Those differences matter in compact builds or when the power supply is modest. The GTX 1660 SUPER is the choice for gamers and compute users who want the full performance envelope. The GTX 1650 SUPER is the fallback for builds where power draw, connector compatibility, or the lower launch price are the deciding constraints, accepting a meaningful performance drop in exchange.

Architecture Differences

Both cards share the same fundamental architecture. They use the TU116 chip, built on Turing architecture, manufactured by TSMC on a 12 nm process. Both have 6,600 million transistors on a 284 mm² die, giving them the same transistor density of 23.2M per mm². Neither card includes ray tracing cores or tensor cores, which is consistent with the GeForce 16 generation positioning.

The differences begin at the execution units. The GTX 1660 SUPER has 1408 shading units, 88 texture mapping units, and 48 render output units. The GTX 1650 SUPER steps down to 1280 shading units, 80 TMUs, and 32 ROPs. That is a reduction of 128 shaders, 8 TMUs, and 16 ROPs in the smaller card. These cutbacks directly affect pixel throughput, texture throughput, and raw compute.

Clock speeds also diverge. Both cards share a 1530 MHz base clock, but the boost clock differs: 1785 MHz on the GTX 1660 SUPER versus 1725 MHz on the GTX 1650 SUPER. Memory clocks differ as well, with the GTX 1660 SUPER running at 1750 MHz (14 Gbps effective) and the GTX 1650 SUPER at 1500 MHz (12 Gbps effective). The memory subsystem is where the gap widens further. The GTX 1660 SUPER carries 6 GB of GDDR6 on a 192-bit bus, delivering 336.0 GB/s of bandwidth. The GTX 1650 SUPER has 4 GB of GDDR6 on a 128-bit bus, delivering 192.0 GB/s. That is a 75% bandwidth advantage for the GTX 1660 SUPER, a figure that heavily influences the benchmark results.

Power delivery differs as well. The GTX 1660 SUPER has a 125 W TDP and requires a single 8-pin connector. The GTX 1650 SUPER has a 100 W TDP and uses a single 6-pin connector. Both recommend a 300 W power supply, and both are dual-slot cards with identical dimensions: 229 mm long, 111 mm high, and 35 mm wide. Display outputs are identical too, with one DVI, one HDMI 2.0, and one DisplayPort 1.4a. Both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The production status for both is end-of-life.

Where Each One Wins

The GTX 1660 SUPER wins in every measured category, but the magnitude of the wins varies by workload. Its strongest performance is in 3DMark Steel Nomad DX12, where it scores 1278 against 352, a 263.1% advantage. This indicates a massive lead in modern DX12 gaming load scenarios. In PassMark DirectX 11, it scores 104 versus 73, a 42.5% lead, showing strong performance in the most common gaming API. The PassMark DirectX 9 test shows a 27.7% lead (189 versus 148), and the DirectX 10 test shows a 30% lead (65 versus 50). The DirectX 12 PassMark test shows a narrower 11.1% gap (50 versus 45), suggesting the 1660 SUPER's advantage compresses in that specific legacy DX12 path.

Compute workloads also favor the GTX 1660 SUPER. In Geekbench OpenCL it scores 52490 versus 43875, a 19.6% lead. In Geekbench Vulkan it scores 57102 versus 50519, a 13% lead. PassMark GPU Compute shows 5076 versus 4477, a 13.4% lead. The 3D gaming aggregate, PassMark G3D, shows 12699 versus 10179, a 24.8% lead. Even the 2D test, PassMark G2D, favors the GTX 1660 SUPER at 805 versus 749, a 7.5% lead.

The GTX 1650 SUPER has no benchmark wins to claim. Its role is defined by its physical and electrical characteristics: lower TDP, smaller power connector, and the 159 USD launch MSRP. For a system with a weaker power supply or a case where the 8-pin connector is unavailable, the GTX 1650 SUPER becomes the practical option. It is also the card to choose if the 4 GB memory capacity is sufficient for the target games, since the 6 GB frame buffer of the GTX 1660 SUPER is only useful if the workload actually uses it.

FAQ

Q: Which card is faster in gaming benchmarks?

A: The GTX 1660 SUPER wins all gaming-related tests. In PassMark G3D it scores 12699 versus 10179, a 24.8% lead. In 3DMark Steel Nomad DX12 it scores 1278 versus 352, a 263.1% lead. In PassMark DirectX 11 it leads 104 to 73, a 42.5% gap.

Q: Why does the GTX 1660 SUPER have such a large lead in 3DMark Steel Nomad DX12?

A: The 3DMark Steel Nomad DX12 result is 1278 for the GTX 1660 SUPER versus 352 for the GTX 1650 SUPER, a 263.1% delta. The larger card has 48 ROPs versus 32, 88 TMUs versus 80, and 336.0 GB/s memory bandwidth versus 192.0 GB/s, which are the likely contributors to this extreme gap.

Q: Is the GTX 1650 SUPER ever faster than the GTX 1660 SUPER?

A: No. Across all 10 head-to-head benchmarks, the GTX 1660 SUPER wins every test. The GTX 1650 SUPER records zero wins.

Q: Do both cards use the same power supply recommendation?

A: Yes, both list a suggested PSU of 300 W. However, the GTX 1660 SUPER has a 125 W TDP and needs a single 8-pin connector, while the GTX 1650 SUPER has a 100 W TDP and uses a single 6-pin connector.

Q: What are the memory differences between the two cards?

A: The GTX 1660 SUPER has 6 GB of GDDR6 on a 192-bit bus with 336.0 GB/s bandwidth. The GTX 1650 SUPER has 4 GB of GDDR6 on a 128-bit bus with 192.0 GB/s bandwidth. The GTX 1660 SUPER also runs its memory at 14 Gbps effective versus 12 Gbps effective.

Q: Are these cards still in production?

A: Both are listed as end-of-life. The GTX 1660 SUPER was released on 2019-10-28, and the GTX 1650 SUPER on 2019-11-21. Both are successors in the GeForce 20 series lineage and predecessors to the GeForce 10 series.

Head-to-Head Benchmarks

The head-to-head results are one-sided, but the margins tell a nuanced story about where the GTX 1660 SUPER's extra hardware matters most.

The single most dramatic result is 3DMark Steel Nomad DX12. The GTX 1660 SUPER scores 1278, the GTX 1650 SUPER scores 352, a 263.1% delta. This is the largest gap in the entire comparison and signals that modern DX12 workloads heavily favor the larger card's memory bandwidth and ROP count.

Next is PassMark DirectX 11, where the GTX 1660 SUPER scores 104 versus 73, a 42.5% lead. DirectX 11 remains the baseline for many games, so this gap translates directly to real-world gaming performance in older titles. PassMark DirectX 10 shows a 30% lead (65 versus 50), and PassMark DirectX 9 shows a 27.7% lead (189 versus 148). These older API tests follow the same pattern, with the GTX 1660 SUPER maintaining a consistent advantage.

The 3D gaming aggregate, PassMark G3D, puts the GTX 1660 SUPER at 12699 versus 10179, a 24.8% lead. This is the broadest gaming metric and confirms that the advantage is not limited to a single API or test methodology. Geekbench OpenCL shows a 19.6% lead (52490 versus 43875), Geekbench Vulkan shows a 13% lead (57102 versus 50519), and PassMark GPU Compute shows a 13.4% lead (5076 versus 4477). These compute results show that the GTX 1660 SUPER is not just a gaming card; it also holds a solid edge in general-purpose GPU workloads.

The closest result is PassMark DirectX 12, where the GTX 1660 SUPER scores 50 and the GTX 1650 SUPER scores 45, an 11.1% delta. Even at its narrowest, the GTX 1660 SUPER still wins. The PassMark G2D test, which measures 2D graphics performance, shows a 7.5% lead (805 versus 749). This is the smallest gap overall and suggests that 2D workloads are less sensitive to the hardware differences between the two cards.

The average benchmark scores reflect the same hierarchy. The GTX 1660 SUPER averages 12986 across all tests, while the GTX 1650 SUPER averages 11047. In the database's percentile rankings, the GTX 1660 SUPER sits at the 53rd percentile versus the 50th percentile for the GTX 1650 SUPER. The nearest rivals for the GTX 1660 SUPER include the NVIDIA GeForce RTX 3050 Ti Mobile at 12940 (0.4% behind), the AMD Radeon RX 580 at 12928 (0.4% behind), the NVIDIA Tesla M2090 at 13075 (0.7% ahead), and the AMD Radeon 740M at 12870 (0.9% behind). The GTX 1650 SUPER sits near the AMD Radeon RX 550 at 11075 (0.2% ahead), the NVIDIA RTX PRO 6000D Blackwell Max-Q at 11088 (0.4% ahead), the NVIDIA RTX PRO 6000 Blackwell Max-Q at 11088 (0.4% ahead), and the NVIDIA GeForce MX350 at 10883 (1.5% behind). These groupings show that the GTX 1660 SUPER competes at a higher performance tier than the GTX 1650 SUPER, whose rivals are all lower-performing cards.

Specification Differences

The two cards share the same chip, architecture, process node, foundry, die size, transistor count, and transistor density. They also share the same base clock, slot width, bus interface, display outputs, API support, dimensions, and suggested PSU. The differences are concentrated in the execution units, clocks, memory, and power delivery.

The GTX 1660 SUPER has 1408 shading units, 88 TMUs, and 48 ROPs. The GTX 1650 SUPER has 1280 shading units, 80 TMUs, and 32 ROPs. The boost clock is 1785 MHz on the GTX 1660 SUPER and 1725 MHz on the GTX 1650 SUPER. The memory clock is 1750 MHz (14 Gbps effective) on the GTX 1660 SUPER and 1500 MHz (12 Gbps effective) on the GTX 1650 SUPER.

Memory capacity is 6 GB versus 4 GB, bus width is 192-bit versus 128-bit, and bandwidth is 336.0 GB/s versus 192.0 GB/s. Pixel rate is 85.68 GPixel/s versus 55.20 GPixel/s. Texture rate is 157.1 GTexel/s versus 138.0 GTexel/s. FP32 compute is 5.027 TFLOPS versus 4.416 TFLOPS. FP16 compute is 10.05 TFLOPS (2:1) versus 8.832 TFLOPS (2:1). TDP is 125 W versus 100 W, and the power connector is 1x 8-pin versus 1x 6-pin.

The launch MSRP is 229 USD for the GTX 1660 SUPER and 159 USD for the GTX 1650 SUPER. Release dates are 2019-10-28 for the GTX 1660 SUPER and 2019-11-21 for the GTX 1650 SUPER. Both are end-of-life, both use the TU116 chip, and both belong to the GeForce 16 generation with Turing architecture.