NVIDIA GeForce GTX 950M vs NVIDIA GeForce GTX 980 Comparison

The NVIDIA GeForce GTX 980 and the NVIDIA GeForce GTX 950M represent two vastly different segments of the Maxwell generation, despite both carrying the GeForce GTX badge. The GTX 980 is a desktop flagship built for maximum throughput, while the GTX 950M is a mobile part designed for efficiency within a laptop's thermal envelope. Benchmark data reveals a near-tie in average scores—8167 for the 980 versus 8135 for the 950M—but this aggregate figure masks an enormous performance gap in raw compute workloads. The most striking fact is that the GTX 980 leads the GTX 950M by 255.8% in Geekbench OpenCL and by 245.5% in Geekbench Vulkan, suggesting the average score comparison is misleading due to the limited shared benchmark suite.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA GeForce GTX 980 has an average benchmark score of 8167, while the NVIDIA GeForce GTX 950M scores 8135. The delta between them is just 0.4%, placing them as direct rivals in the database's ranking.

Q: How much faster is the GTX 980 in OpenCL compute?

A: In the Geekbench OpenCL test, the GTX 980 scores 34676, compared to the GTX 950M's 9745. This translates to a 255.8% advantage for the desktop card, a massive lead in raw parallel processing.

Q: What is the memory bandwidth difference between the two cards?

A: The GTX 980 offers 224.4 GB/s of bandwidth using GDDR5 memory on a 256-bit bus. In contrast, the GTX 950M provides just 28.80 GB/s via DDR3 on a 128-bit bus—a difference that heavily impacts memory-intensive workloads.

Q: What are the transistor counts for each chip?

A: The GTX 980 uses the GM204 chip with 5,200 million transistors, while the GTX 950M uses the GM107 chip with 1,870 million transistors. The GTX 980's die size is 398 mm² versus the 950M's 148 mm².

Q: Do both GPUs support the same DirectX feature level?

A: No. The GTX 980 supports DirectX 12 (12_1), while the GTX 950M is limited to DirectX 12 (11_0). Both support OpenGL 4.6 and Vulkan 1.4.

Q: Which GPU has a higher pixel fill rate?

A: The GTX 980 achieves a pixel rate of 77.82 GPixel/s, driven by 64 ROPs. The GTX 950M manages only 17.98 GPixel/s with 16 ROPs, making the 980 roughly four times faster in this metric.

Where Each One Wins

The GTX 980 wins decisively in every shared benchmark, but the nature of those wins reveals distinct use cases. In Geekbench OpenCL, the 980's 34676 score versus the 950M's 9745 indicates that the desktop card excels at general-purpose compute tasks, such as rendering, physics simulation, or any workload that leverages massive parallel execution. The 255.8% delta suggests the 980 is not just faster but in a different performance class entirely for GPU-accelerated applications.

The Geekbench Vulkan result tells a similar story for graphics APIs: the 980 scores 22543 against the 950M's 6525, a 245.5% lead. This indicates that for modern games or applications using Vulkan, the GTX 980 provides a far smoother experience at higher resolutions or detail settings. The GTX 950M, by contrast, is not competitive in these absolute terms, but its design purpose is different—it fits into the IGP slot width with no power connectors and a 75 W TDP, making it suitable for thin-and-light laptops where the GTX 980's 165 W TDP and dual-slot cooler are physically impossible.

Where the GTX 950M "wins" is in the context of its form factor. It offers a 4GB DDR3 memory configuration and a 128-bit bus, which is modest but functional for 1080p gaming on lower settings. The data shows that the 950M appears in the same nearestRivals list as the 980 with only a 0.4% average score difference, but this is an artifact of the sparse benchmark overlap—the 950M has only two recorded tests (OpenCL and Vulkan), both of which it loses by a wide margin. In practical terms, the GTX 980 wins for any user needing desktop-class performance, while the 950M is the only option for portable systems.

Architecture Differences

The architectural split between these two GPUs is fundamental. The GTX 980 is built on the GM204 chip with the Maxwell 2.0 architecture, which introduced features like improved color compression and tile-based rasterization for better efficiency. The GTX 950M uses the GM107 chip with the original Maxwell architecture, a smaller and older design that lacks some of the 2.0 optimizations.

The silicon itself tells the story. The GTX 980 packs 5,200 million transistors onto a 398 mm² die, yielding a transistor density of 13.1M / mm². The GTX 950M has just 1,870 million transistors on a 148 mm² die, with a lower density of 12.6M / mm². Both are fabricated on TSMC's 28 nm process, but the 980's larger die allows for a substantially more complex compute pipeline.

This complexity translates directly into execution resources. The GTX 980 has 2048 shading units, 128 texture mapping units (TMUs), and 64 raster output units (ROPs). The GTX 950M is cut down to 640 shading units, 40 TMUs, and 16 ROPs. That is a 3.2x difference in shading units and a 4x difference in ROPs, which explains the massive pixel rate advantage (77.82 vs 17.98 GPixel/s) and texture rate advantage (155.6 vs 44.96 GTexel/s) for the 980.

The memory subsystem also differs architecturally. The GTX 980 uses GDDR5 at 1753 MHz with a 256-bit interface, achieving 224.4 GB/s. The GTX 950M uses DDR3 at 900 MHz on a 128-bit bus, yielding just 28.80 GB/s. This 7.8x bandwidth gap is a critical architectural difference, as it limits the 950M's ability to feed its smaller set of cores in bandwidth-hungry scenarios. Neither GPU has RT cores or tensor cores, as both predate the RTX series.

Specification Differences

The spec sheets diverge sharply across nearly every field. Clock speeds: the GTX 980 runs at a 1127 MHz base and 1216 MHz boost, while the GTX 950M runs at 993 MHz base and 1124 MHz boost—a smaller gap in clocks than in core counts, but still a 13% advantage for the 980 at base.

Power and physical requirements differ by design intent. The GTX 980 has a 165 W TDP, requires a 450 W suggested PSU, uses 2x 6-pin power connectors, and occupies a dual-slot footprint with dimensions of 267 mm length, 111 mm height, and 40 mm width. The GTX 950M has a 75 W TDP, no power connectors, an IGP slot width, and no listed dimensions, as it's soldered into laptops. The bus interface also differs: the 980 uses PCIe 3.0 x16, while the 950M is limited to PCIe 3.0 x8, halving the available bandwidth to the host system.

Display outputs separate them further. The GTX 980 offers 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2, enabling multi-monitor desktop setups. The GTX 950M's outputs are "Portable Device Dependent," meaning they vary by laptop manufacturer. Memory size is identical at 4 GB, but the type, bus width, and bandwidth are all superior on the 980, as noted earlier.

Compute throughput: the GTX 980 delivers 4.981 TFLOPS of FP32 performance, while the GTX 950M manages 1,438.7 GFLOPS (approximately 1.44 TFLOPS). This 3.5x difference in floating-point capability underpins the OpenCL benchmark gap. The 950M has no launch MSRP, while the 980 has a launch MSRP of 549 USD. Release dates also differ: the 980 launched on 2014-09-18, and the 950M followed on 2015-03-12.

Head-to-Head Benchmarks

The shared benchmark suite between these two GPUs is limited to just two tests, but both tell a consistent story of overwhelming GTX 980 dominance.

In Geekbench OpenCL, the GTX 980 scores 34676 against the GTX 950M's 9745. The delta is 255.8%, meaning the 980 delivers more than 3.5 times the performance. This is not a marginal victory but a categorical one, reflecting the 980's 2048 shading units and 224.4 GB/s bandwidth versus the 950M's 640 units and 28.80 GB/s. For any OpenCL workload—whether it's video encoding, scientific computation, or data processing—the 980 is the clear choice, and the data suggests it will complete tasks in less than a third of the time.

The Geekbench Vulkan test shows a similar pattern. The GTX 980 scores 22543, and the GTX 950M scores 6525, a 245.5% delta. Vulkan is a low-overhead graphics API, and the 980's superior ROP count (64 vs 16) and pixel rate (77.82 vs 17.98 GPixel/s) directly contribute to this result. For gaming, this means the 980 can handle higher resolutions, more demanding effects, and higher frame rates, while the 950M will struggle to maintain playable performance in modern titles.

The wins tally is 2 for the GTX 980 and 0 for the GTX 950M, with no benchmark where the mobile part comes out ahead. Even the average benchmark score (8167 vs 8135) is misleadingly close, because the 950M's average is based on only two tests, both of which it loses by over 200%. The GTX 980's average includes additional tests like Passmark G3D (11095) and Geekbench Metal (15163), which the 950M cannot participate in due to its different feature set. In a direct comparison, the GTX 980 is the definitive winner, with the 950M serving as a reminder that mobile GPUs of this era traded massive amounts of performance for portability.