NVIDIA GeForce GTX 1080 vs NVIDIA GeForce GTX 950 Comparison

Head-to-Head Benchmarks

The benchmark data is decisive: the NVIDIA GeForce GTX 1080 wins all three recorded head-to-head tests, with margins ranging from substantial to overwhelming. The GTX 950 does not claim a single victory in any measured workload.

In the Geekbench Metal test, the GTX 1080 scores 23,824 against the GTX 950's 7,172. That is a delta of -69.9% for the GTX 950, meaning the GTX 1080 delivers roughly 3.3 times the Metal performance. This is the largest relative gap between the two cards in any shared benchmark.

The Geekbench OpenCL result follows a similar pattern. The GTX 1080 posts 51,204, while the GTX 950 manages 15,662. The delta of -69.4% confirms that the GTX 1080 is approximately 3.3 times faster in OpenCL compute workloads. The near-identical percentage to the Metal test suggests the architectural advantage scales consistently across different compute APIs.

The closest contest appears in Geekbench Vulkan, where the GTX 950 scores 16,734 and the GTX 1080 reaches 30,398. The delta here is -45%, so the GTX 1080 still leads by a wide margin, but the relative gap narrows compared to Metal and OpenCL. This indicates that Vulkan overhead or driver efficiency partially closes the performance chasm, though the GTX 1080 remains clearly ahead.

When placed against the broader database, both cards sit in similar percentile territory, which is notable given the raw performance disparity. The GTX 950 holds a 53rd percentile rank across all GPUs, while the GTX 1080 sits at the 51st percentile. The average benchmark score tells a similar story: the GTX 950 averages 13,189, while the GTX 1080 averages 11,960. This inverted relationship between head-to-head results and aggregate database standing stems from the fact that the GTX 1080's benchmark suite includes additional tests like Passmark DirectX 9, 10, 11, 12, G2D, G3D, and GPU Compute, where its scores vary widely from 55 in DirectX 12 to 211 in DirectX 9 and 15,586 in G3D. The GTX 950, by contrast, has only three recorded benchmarks, all Geekbench variants, which skew its average upward relative to its actual compute abilities.

The nearest rival data for each card reinforces this divergence. The GTX 950's closest competitors, the GTX 480, Tesla M2090, Radeon Pro 555X, and FirePro M6100, all cluster within 1.2% of its average score. The GTX 1080's nearest rivals, the GTX 960A, RX 6500 XT, GTX 1660, and RX 7800 XT, also sit within 2.9% of its average, but those rivals are far less capable in absolute terms than the GTX 1080's head-to-head opponent. The data shows that the GTX 1080, despite its lower aggregate average, utterly dominates the GTX 950 in every directly comparable workload.

The Verdict

The verdict is unambiguous: the GTX 1080 is the superior card in every measured benchmark. Anyone choosing between these two for compute-intensive tasks, modern game rendering, or API-heavy workloads should select the GTX 1080 without hesitation. The data shows leads of 69.9%, 69.4%, and 45% across Metal, OpenCL, and Vulkan respectively. No workload in the recorded tests favors the GTX 950.

The GTX 950 is only relevant if the use case is strictly limited to the lowest-end legacy scenarios, and even then, the GTX 1080's raw power would still win in any shared test. The GTX 950's percentile rank of 53 versus the GTX 1080's 51 does not indicate parity; it reflects the different benchmark sets recorded for each card. The GTX 950's average score of 13,189 comes from three Geekbench tests where it scores between 7,172 and 16,734, while the GTX 1080's average of 11,960 is dragged down by low Passmark DirectX scores (55 to 211) that are not directly comparable to the Geekbench suite.

For a buyer choosing between these two cards today, the GTX 1080 is the only rational pick based on performance data. The GTX 950 offers no advantage in any recorded metric. The GTX 1080's nearest rivals in the database, such as the GTX 1660 (2.4% behind) and RX 7800 XT (2.9% behind), are all modern cards with similar aggregate scores, but the GTX 1080 still outperforms them in direct comparison, just as it outperforms the GTX 950 by massive margins.

Architecture Differences

The two cards belong to different NVIDIA architectures and process nodes. The GTX 950 uses the GM206 chip built on Maxwell 2.0 architecture, fabricated on a 28 nm process at TSMC. The GTX 1080 uses the GP104 chip built on Pascal architecture, fabricated on a 16 nm process, also at TSMC. The node shrink from 28 nm to 16 nm is a major architectural leap, enabling higher transistor density and clock speeds.

Transistor counts differ substantially. The GTX 950 packs 2,940 million transistors into a 228 mm² die, yielding a density of 12.9 million transistors per square millimeter. The GTX 1080 contains 7,200 million transistors on a 314 mm² die, achieving a density of 22.9 million per square millimeter. That is nearly double the density per area, a direct consequence of the smaller process node.

The memory subsystems are also architecturally distinct. The GTX 950 uses 2 GB of GDDR5 memory on a 128-bit bus, delivering 105.8 GB/s of bandwidth. The GTX 1080 uses 8 GB of GDDR5X memory on a 256-bit bus, delivering 320.3 GB/s. The GDDR5X standard, combined with the wider bus, gives the GTX 1080 roughly three times the memory bandwidth.

Shading resources scale accordingly. The GTX 950 has 768 shading units, 48 texture mapping units, and 32 ROPs. The GTX 1080 has 2,560 shading units, 160 texture mapping units, and 64 ROPs. These are not mere clock differences; the GTX 1080 has over three times the shader count, over three times the TMUs, and exactly double the ROPs.

Pixel and texture rates reflect these differences. The GTX 950 achieves 38.02 GPixel/s and 57.02 GTexel/s. The GTX 1080 reaches 110.9 GPixel/s and 277.3 GTexel/s. Floating point performance tells the same story: the GTX 950 delivers 1.825 TFLOPS of FP32 compute, while the GTX 1080 delivers 8.873 TFLOPS, a nearly fivefold increase. The GTX 1080 also has a recorded FP16 figure of 138.6 GFLOPS at a 1:64 ratio, which is not present for the GTX 950.

Both cards support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, so API compatibility is identical. Display outputs differ slightly: both have 1x DVI, 1x HDMI 2.0, and 3x DisplayPort, but the GTX 1080 uses DisplayPort 1.4a while the GTX 950 uses DisplayPort 1.2.

Specification Differences

The following specifications differ between the two cards, based solely on recorded data:

• Process node: 28 nm (GTX 950) versus 16 nm (GTX 1080)
• Transistors: 2,940 million versus 7,200 million
• Die size: 228 mm² versus 314 mm²
• Transistor density: 12.9M / mm² versus 22.9M / mm²
• Base clock: 1024 MHz versus 1607 MHz
• Boost clock: 1188 MHz versus 1733 MHz
• Memory clock: 1653 MHz (6.6 Gbps effective) versus 1251 MHz (10 Gbps effective)
• Memory size: 2 GB versus 8 GB
• Memory type: GDDR5 versus GDDR5X
• Memory bus width: 128 bit versus 256 bit
• Memory bandwidth: 105.8 GB/s versus 320.3 GB/s
• Shading units: 768 versus 2,560
• TMUs: 48 versus 160
• ROPs: 32 versus 64
• Pixel rate: 38.02 GPixel/s versus 110.9 GPixel/s
• Texture rate: 57.02 GTexel/s versus 277.3 GTexel/s
• FP32: 1.825 TFLOPS versus 8.873 TFLOPS
• FP16: not recorded versus 138.6 GFLOPS (1:64)
• TDP: 90 W versus 180 W
• Power connectors: 1x 6-pin versus 1x 8-pin
• Suggested PSU: 250 W versus 450 W
• DisplayPort version: 1.2 versus 1.4a
• Dimensions: 202 mm (8 inches) long versus 267 mm (10.5 inches) long, 112 mm (4.4 inches) tall, 40 mm (1.6 inches) wide
• Release date: 2015-08-19 versus 2016-05-26
• Predecessor: GeForce 700 versus GeForce 900
• Successor: GeForce 10 versus GeForce 20
• Launch MSRP: 159 USD versus 599 USD

Both cards are dual-slot, use PCIe 3.0 x16, and share the same DirectX, OpenGL, and Vulkan versions. Both are end-of-life products.

FAQ

Q: Which card wins in Geekbench Metal?

A: The GTX 1080 wins decisively with a score of 23,824 against the GTX 950's 7,172, a delta of -69.9% for the GTX 950.

Q: How large is the OpenCL performance gap?

A: The GTX 1080 scores 51,204 in Geekbench OpenCL, while the GTX 950 scores 15,662. The GTX 950 trails by 69.4%.

Q: Is the Vulkan gap smaller than the other benchmarks?

A: Yes. In Geekbench Vulkan, the GTX 1080 scores 30,398 versus the GTX 950's 16,734, a delta of -45%. This is the smallest relative margin among the three head-to-head tests.

Q: Do the cards have the same API support?

A: Yes. Both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4.

Q: What are the memory differences?

A: The GTX 950 has 2 GB of GDDR5 on a 128-bit bus with 105.8 GB/s bandwidth. The GTX 1080 has 8 GB of GDDR5X on a 256-bit bus with 320.3 GB/s bandwidth.

Q: Which card has more shading units?

A: The GTX 1080 has 2,560 shading units, while the GTX 950 has 768. The GTX 1080 also has 160 TMUs versus 48, and 64 ROPs versus 32.

Q: What is the FP32 compute difference?

A: The GTX 950 delivers 1.825 TFLOPS, while the GTX 1080 delivers 8.873 TFLOPS. The GTX 1080 is approximately 4.9 times faster in FP32.

Where Each One Wins

The GTX 1080 wins every recorded benchmark category. In the three head-to-head tests, it leads by margins of 69.9% (Metal), 69.4% (OpenCL), and 45% (Vulkan). There is no workload in the database where the GTX 950 comes out ahead.

For use cases involving modern game rendering at high resolutions, compute-heavy tasks such as machine learning inference or video processing, or any application leveraging Vulkan or Metal APIs, the GTX 1080 is the clear choice. Its 8 GB of GDDR5X memory and 320.3 GB/s bandwidth make it suitable for large textures and high-resolution assets, while its 8.873 TFLOPS FP32 compute handles demanding shader workloads.

The GTX 950, with its 2 GB GDDR5 memory and 1.825 TFLOPS, is only viable for legacy 1080p gaming at low to medium settings or basic compute tasks that do not stress memory capacity. Its 53rd percentile rank suggests it sits near the middle of the database, but that rank is based on only three Geekbench tests, all of which it loses to the GTX 1080 by wide margins.

If the choice is strictly between these two cards, the GTX 1080 wins for every purpose: gaming, compute, productivity, and API-specific workloads. The GTX 950 offers no advantage in any measured metric, and its only potential appeal would be lower power draw (90 W versus 180 W) or smaller physical footprint (202 mm versus 267 mm length), but those are not performance wins. The data is unambiguous: the GTX 1080 is the superior card in every recorded benchmark.