NVIDIA GeForce GTX 1650 vs NVIDIA GeForce GTX 980 Comparison

The GeForce GTX 980 and GeForce GTX 1650 are separated by nearly five years of GPU architecture evolution, yet the benchmark data reveals a fascinating reversal of expectations. The older GTX 980, built on the Maxwell 2.0 architecture with a 28 nm process, consistently outperforms the newer GTX 1650 in most traditional rasterization workloads, while the Turing-based GTX 1650 strikes back decisively in modern API compute tasks. This comparison is not a simple generational upgrade; it is a clash between a former flagship with raw hardware muscle and a modern entry-level part with contemporary feature support.

Head-to-Head Benchmarks

The GTX 980 wins nine of the ten head-to-head benchmark comparisons, and the margins are frequently substantial. In the demanding 3DMark Steel Nomad DX12 test, the GTX 980 scores 474 against the GTX 1650's 305, a commanding 55.4% advantage. This is the largest single-test delta in the entire comparison, and it underscores how the GTX 980's broader 256-bit memory bus and higher shading unit count translate into raw throughput that the GTX 1650 cannot match. The GTX 980's lead extends to compute workloads as well, where it scores 4753 in Passmark GPU Compute versus 3048 for the GTX 1650, a 55.9% difference that shows the older card's 2048 shading units are far more numerous than the GTX 1650's 896.

The pattern continues across DirectX 11 and DirectX 12 legacy benchmarks. In Passmark DirectX 11, the GTX 980 posts 83 points to the GTX 1650's 58, a 43.1% lead. The DirectX 12 test shows a similar story: 46 versus 35, a 31.4% margin. Even in older API tests, the GTX 980 dominates—Passmark DirectX 9 shows 164 versus 124 (32.3% ahead), and DirectX 10 shows 53 versus 39 (35.9% ahead). The GTX 980's advantage in the 2D test is also notable: 792 versus 561, a 41.2% gap that reflects its higher pixel rate of 77.82 GPixel/s against the GTX 1650's 53.28 GPixel/s.

The Geekbench OpenCL results reinforce the GTX 980's lead, with a score of 34676 against 29629, a 17% advantage. However, the Geekbench Vulkan test tells an entirely different story. Here, the GTX 1650 scores 33042, decisively beating the GTX 980's 22543 by 31.8%. This is the sole win for the GTX 1650, and it is a significant one, suggesting that the newer Turing architecture has substantial advantages in Vulkan compute workloads that the older Maxwell design lacks.

Where Each One Wins

The GTX 980 is the clear winner for traditional gaming and general-purpose GPU compute. Its performance in Passmark G3D (11095 versus 7880, a 40.8% lead) and across all DirectX API tests indicates that for DirectX-based games—particularly those using DX11 or DX12—the GTX 980 provides noticeably higher frame rates and smoother experiences. The 17% OpenCL advantage also makes the GTX 980 the better choice for compute-heavy applications that leverage OpenCL, such as certain video encoding or scientific workloads. The GTX 980's 55.4% lead in the 3DMark Steel Nomad DX12 test further confirms its superiority in modern DirectX 12 gaming scenarios.

The GTX 1650's single victory in Geekbench Vulkan is not trivial. Vulkan is increasingly used in modern game engines and emulators, and the 31.8% advantage here suggests that the GTX 1650 will perform better in Vulkan-native titles and applications that prioritize this API. The GTX 1650's higher boost clock of 1665 MHz (versus 1216 MHz on the GTX 980) and its Turing architecture's improved compute efficiency likely contribute to this result. For users who primarily play Vulkan-based games or use Vulkan-powered software, the GTX 1650 is the more capable option despite its overall lower raw specifications.

Architecture Differences

The architectural gap between these two cards is substantial. The GTX 980 uses the GM204 chip based on Maxwell 2.0, built on a 28 nm process at TSMC. This older design packs 5,200 million transistors into a 398 mm² die, yielding a transistor density of 13.1M per mm². In contrast, the GTX 1650 uses the TU117 chip based on Turing, fabricated on a 12 nm process, also at TSMC. The newer process allows 4,700 million transistors in a much smaller 200 mm² die, achieving a significantly higher density of 23.5M per mm². This density improvement is a hallmark of the node transition, but the GTX 980's larger die and higher transistor count give it a raw hardware advantage.

The shading unit count reflects this disparity. The GTX 980 has 2048 shading units, 128 texture mapping units, and 64 ROPs, while the GTX 1650 has only 896 shading units, 56 TMUs, and 32 ROPs. This nearly 2:1 ratio in core resources explains the GTX 980's superior fill rates and compute throughput. The GTX 980's texture rate of 155.6 GTexel/s is 67% higher than the GTX 1650's 93.24 GTexel/s, and its pixel rate of 77.82 GPixel/s is 46% higher. In terms of FP32 performance, the GTX 980 delivers 4.981 TFLOPS against the GTX 1650's 2.984 TFLOPS, a 67% advantage.

The GTX 1650 does have a feature the GTX 980 lacks: FP16 support at 5.967 TFLOPS (2:1), which can accelerate certain workloads that leverage half-precision math. However, the GTX 980 has no listed FP16 capability, so this is a Turing-specific advantage. Both cards support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, so API feature levels are identical, but the GTX 1650's Vulkan performance suggests the architecture itself is better optimized for this API.

Specification Differences

The most glaring difference is memory bandwidth. The GTX 980 uses a 256-bit memory bus with GDDR5 running at 7 Gbps effective, delivering 224.4 GB/s of bandwidth. The GTX 1650 uses a 128-bit bus with GDDR5 at 8 Gbps effective, yielding only 128.1 GB/s. This 75% bandwidth advantage for the GTX 980 is a primary reason for its superior performance in high-resolution textures and memory-intensive workloads. Both cards have 4 GB of VRAM, so capacity is equal, but the GTX 980 can feed its cores much faster.

Power requirements also differ significantly. The GTX 980 has a 165 W TDP and requires two 6-pin power connectors, with a suggested PSU of 450 W. The GTX 1650, by contrast, has a 75 W TDP, requires no power connectors, and works with a 250 W PSU. This makes the GTX 1650 far easier to install in existing systems without PSU upgrades. Physical dimensions also favor the GTX 1650: it is 229 mm (9 inches) long versus 267 mm (10.5 inches) for the GTX 980, and it is thinner at 35 mm versus 40 mm. Both are dual-slot cards with a height of 111 mm.

Display outputs are nearly identical, with both offering 1x DVI and 1x HDMI 2.0, but the GTX 980 adds 3x DisplayPort 1.2 while the GTX 1650 has only 1x DisplayPort 1.4a. The GTX 1650's DisplayPort 1.4a supports higher bandwidth than the GTX 980's DisplayPort 1.2, which may matter for high-refresh-rate or high-resolution monitors. The GTX 980 was released in 2014 with a launch MSRP of 549 USD, while the GTX 1650 launched in 2019 with a launch MSRP of 149 USD. Both are now end-of-life products.

FAQ

Q: Which card is faster in DirectX 11 games?

A: The GTX 980 is significantly faster. In the Passmark DirectX 11 test, it scores 83 versus 58 for the GTX 1650, a 43.1% advantage.

Q: Does the GTX 1650 beat the GTX 980 in any benchmark?

A: Yes, in the Geekbench Vulkan test, the GTX 1650 scores 33042 against the GTX 980's 22543, a 31.8% lead. This is the only head-to-head test the GTX 1650 wins.

Q: How much more memory bandwidth does the GTX 980 have?

A: The GTX 980 has 224.4 GB/s of bandwidth thanks to its 256-bit bus, while the GTX 1650 has 128.1 GB/s on a 128-bit bus. That is a 75% bandwidth advantage for the GTX 980.

Q: Can the GTX 1650 run on a lower wattage power supply?

A: Yes, the GTX 1650 has a 75 W TDP and requires no power connectors, with a suggested PSU of 250 W. The GTX 980 needs 165 W, two 6-pin connectors, and a 450 W PSU.

Q: Do both cards support the same modern APIs?

A: Both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. However, the GTX 1650 also supports FP16 compute at 5.967 TFLOPS, which the GTX 980 does not list.

Q: Which card has higher core count?

A: The GTX 980 has 2048 shading units, 128 TMUs, and 64 ROPs. The GTX 1650 has 896 shading units, 56 TMUs, and 32 ROPs. The GTX 980 has more than double the shading units.

The Verdict

The data is unambiguous: the GTX 980 is the superior performer for the vast majority of workloads. With nine wins out of ten head-to-head tests and leads ranging from 17% to 55.9%, it dominates in DirectX-based gaming, OpenCL compute, and general 3D rendering. Its 40.8% lead in Passmark G3D and 55.4% lead in 3DMark Steel Nomad DX12 make it the clear choice for anyone prioritizing raw gaming performance in traditional titles. The GTX 980's larger memory bus and higher core counts are the reasons for this dominance, and for users with a PSU capable of handling 165 W and two 6-pin connectors, it is the unequivocal pick.

The GTX 1650 is not without merit, but its appeal is narrow and specific. Its 31.8% Vulkan advantage makes it the better option for Vulkan-native applications, and its 75 W TDP with no power connectors means it can drop into almost any system without PSU concerns. Its smaller 229 mm length and thinner 35 mm profile also make it more case-friendly. The GTX 1650's FP16 support adds compute versatility, and its DisplayPort 1.4a output is more modern. However, for users who want the best frame rates in DirectX games or need maximum compute throughput, the GTX 980's superior specifications and benchmark results make it the definitive choice. The GTX 1650 is a pragmatic, low-power alternative, but the GTX 980 is the performance king.