AMD Radeon HD 6950 vs NVIDIA GeForce GTX 765M Comparison

Head-to-Head Benchmarks

The single recorded head-to-head comparison in the database is the Geekbench OpenCL test, and it produces a clear, if not overwhelming, winner. The NVIDIA GeForce GTX 765M scores 7,176 points, while the AMD Radeon HD 6950 scores 6,210 points. This gives NVIDIA a 13.5% advantage in raw compute throughput as measured by this workload.

That margin is substantial enough to be meaningful, but it is not a blowout. The AMD card is an older, desktop-class part with a higher power envelope, yet it trails the mobile NVIDIA chip by more than a tenth in this synthetic compute test. The GTX 765M's Kepler architecture appears to handle the OpenCL workload more efficiently, which is notable given the disparity in their physical specifications.

Looking at the broader database context, the AMD Radeon HD 6950's average score of 6,210 places it at the 36th percentile of all GPUs tracked. Its nearest rivals are tightly clustered: the AMD FirePro W600 scores 6,223 (0.2% higher), the NVIDIA GeForce RTX 4070 Ti SUPER AD102 scores 6,270 (0.9% higher), and the NVIDIA Quadro K620 scores 6,282 (1.1% higher). This means the HD 6950 is essentially trading blows with much newer, higher-end parts in this specific test, which speaks to the strength of its raw compute resources.

The NVIDIA GeForce GTX 765M, by contrast, has an average benchmark score of 5,501 across all its recorded tests, which places it at the 32nd percentile. Its nearest rivals include the NVIDIA GeForce MX130 (5,508, 0.1% higher), the AMD Radeon R7 M440 (5,483, 0.3% lower), and the NVIDIA Quadro M4000 (5,467, 0.6% lower). The GTX 765M's OpenCL result of 7,176 is significantly higher than its own average, indicating that this particular workload plays to its strengths relative to its overall performance profile.

It is worth remembering the GTX 765M has three benchmark entries in the database, while the HD 6950 has only one. The NVIDIA part also records a Geekbench Metal score of 2,612 and a Geekbench Vulkan score of 6,714, which are not directly comparable to the AMD card since no matching tests were run for it. The AMD part's sole data point is the OpenCL result, meaning the head-to-head comparison is singular but decisive.

Where Each One Wins

The data structure shows a single win for the NVIDIA GeForce GTX 765M and zero wins for the AMD Radeon HD 6950 in their direct head-to-head matchup. That categorical result comes from the Geekbench OpenCL test, where NVIDIA leads by 13.5%.

However, the use-case split is not as simple as a single benchmark suggests. The AMD card, with its 1,408 shading units, 88 texture mapping units, and 32 raster operation pipelines, is built for raw throughput. Its pixel rate of 25.60 GPixel/s and texture rate of 70.40 GTexel/s are impressive numbers that indicate strong fill-rate capabilities, which would matter in traditional rasterization workloads. The NVIDIA part counters with 768 shading units, 64 TMUs, and 16 ROPs, yielding a pixel rate of 13.81 GPixel/s and a texture rate of 55.23 GTexel/s. On paper, the AMD card has the architectural muscle for pixel-pushing tasks.

The GTX 765M's wins come from its software stack and API support. It supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175, whereas the HD 6950 is limited to DirectX 11.2 (11_0) and OpenGL 4.4, with no Vulkan support at all. For modern workloads that leverage newer APIs, the NVIDIA part is the only one with a path forward. Its Vulkan score of 6,714 in the database shows that it can perform well in that environment.

For compute-heavy tasks like OpenCL, the GTX 765M demonstrates superiority in the recorded test. For legacy DirectX 11 and OpenGL applications, the AMD card's higher fill rates and larger memory bus (256-bit versus 128-bit) could theoretically provide advantages, though no direct benchmark confirms this. The database records only the OpenCL result, so claims about other workloads remain speculative based on architectural analysis.

Architecture Differences

The two GPUs come from different architectural eras and design philosophies. The AMD Radeon HD 6950 uses the Cayman chip, built on TeraScale 3 architecture, and belongs to the Northern Islands generation. It is fabricated on a 40 nm process at TSMC, with 2,640 million transistors on a die size of 389 mm². The transistor density is 6.8 million per square millimeter.

The NVIDIA GeForce GTX 765M uses the GK106 chip, built on the Kepler architecture, and belongs to the GeForce 700M generation. It is fabricated on a more advanced 28 nm process at TSMC, packing 2,540 million transistors into a much smaller die of 221 mm². This yields a transistor density of 11.5 million per square millimeter, nearly 70% higher than the AMD part.

The memory subsystems are starkly different. The HD 6950 has a 256-bit bus with 2 GB of GDDR5 running at 1250 MHz (5 Gbps effective), producing 160.0 GB/s of bandwidth. The GTX 765M has a 128-bit bus with 2 GB of GDDR5 at 1002 MHz (4 Gbps effective), yielding just 64.13 GB/s of bandwidth. The AMD card has more than double the memory bandwidth, which is critical for high-resolution textures and bandwidth-hungry workloads.

Clock speeds tell another story. The GTX 765M runs at a base clock of 797 MHz with a boost clock of 863 MHz. The HD 6950 has no recorded base or boost clock, but its memory clock is higher than the NVIDIA part's. The AMD card's compute throughput is rated at 2.253 TFLOPS for FP32, while the NVIDIA part is rated at 1,325.6 GFLOPS (approximately 1.33 TFLOPS). Despite the NVIDIA win in OpenCL, the AMD part has a higher theoretical FP32 ceiling.

Power consumption is a major differentiator. The HD 6950 has a TDP of 200 W, requires dual-slot cooling, and needs two 6-pin power connectors with a suggested 550 W power supply. The GTX 765M is a mobile MXM module with a 75 W TDP, no power connectors, and no suggested PSU. This is a desktop versus mobile comparison, and the power envelope reflects that. The AMD card is 286 mm long, while the NVIDIA module has no recorded dimensions.

The API support gap is significant. The GTX 765M supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The HD 6950 supports DirectX 11.2 (11_0), OpenGL 4.4, and no Vulkan. This makes the NVIDIA part substantially more future-proof for software that adopts modern graphics APIs.

The Verdict

The data points to a clear, though nuanced, conclusion. In the only directly comparable benchmark, the NVIDIA GeForce GTX 765M outperforms the AMD Radeon HD 6950 by 13.5% in Geekbench OpenCL. This is a meaningful margin that indicates NVIDIA's Kepler architecture handles this compute workload more efficiently, despite the AMD card's higher theoretical FP32 throughput and significantly larger memory bus.

The GTX 765M also carries the advantage of modern API support, including Vulkan and DirectX 12, which the HD 6950 lacks entirely. For any user considering these parts for modern software, the NVIDIA card is the only viable option. Its Vulkan score of 6,714 in the database suggests it performs well in that environment.

However, the AMD HD 6950 has strengths that should not be dismissed. Its 160.0 GB/s of memory bandwidth is more than double the GTX 765M's 64.13 GB/s, which could matter in bandwidth-sensitive applications. Its higher pixel and texture rates (25.60 GPixel/s and 70.40 GTexel/s versus 13.81 GPixel/s and 55.23 GTexel/s) suggest it could be competitive or superior in traditional rasterization workloads, even if the database does not record such a test.

The percentile rankings are close: the HD 6950 sits at the 36th percentile of all GPUs, while the GTX 765M sits at the 32nd percentile. This indicates that, on average, the AMD card is slightly better positioned relative to the entire GPU landscape, even though it loses the direct head-to-head OpenCL test.

For users focused on compute tasks that resemble the Geekbench OpenCL workload, the GTX 765M is the better choice. For users with older software that relies on DirectX 11 and OpenGL 4.4, the AMD card's higher bandwidth and fill rates may offer advantages. For anyone needing Vulkan or DirectX 12 support, the NVIDIA card is the only option. The GTX 765M wins the recorded matchup, but the HD 6950 remains a capable part for its era and power class.

FAQ

Q: Which GPU has the higher Geekbench OpenCL score?

A: The NVIDIA GeForce GTX 765M scores 7,176 points, while the AMD Radeon HD 6950 scores 6,210 points. The NVIDIA part leads by 13.5% in this test.

Q: What are the FP32 compute ratings for these two GPUs?

A: The AMD Radeon HD 6950 is rated at 2.253 TFLOPS, while the NVIDIA GeForce GTX 765M is rated at 1,325.6 GFLOPS for FP32 compute.

Q: Which GPU has more memory bandwidth?

A: The AMD Radeon HD 6950 has 160.0 GB/s of bandwidth from its 256-bit memory bus, while the NVIDIA GeForce GTX 765M has 64.13 GB/s from its 128-bit bus.

Q: Does the AMD Radeon HD 6950 support Vulkan?

A: No. The AMD Radeon HD 6950 supports DirectX 11.2 (11_0) and OpenGL 4.4, but has no Vulkan support. The NVIDIA GeForce GTX 765M supports Vulkan 1.2.175, DirectX 12 (11_0), and OpenGL 4.6.

Q: How do the power requirements compare?

A: The AMD Radeon HD 6950 has a 200 W TDP and requires a 550 W power supply with two 6-pin connectors. The NVIDIA GeForce GTX 765M has a 75 W TDP and requires no external power connectors.

Q: What are the process nodes for each GPU?

A: The AMD Radeon HD 6950 is built on a 40 nm process, while the NVIDIA GeForce GTX 765M uses a 28 nm process. Both are fabricated by TSMC.