NVIDIA GeForce GTX 765M vs NVIDIA GeForce GTX 980M Comparison

The NVIDIA GeForce GTX 765M and GTX 980M represent two distinct generations of NVIDIA's mobile graphics, separated by a significant architectural leap. While both are end-of-life MXM modules aimed at laptops, the benchmark data reveals a stark performance gulf, with the newer GTX 980M dominating the shared tests. The GTX 765M, based on the Kepler architecture, is a capable entry-level part from 2013, while the GTX 980M, built on Maxwell 2.0, was a high-end performer from 2014. This analysis examines their head-to-head results, specification sheets, and architectural differences to determine which GPU holds the advantage and for whom.

Head-to-Head Benchmarks

The direct comparison between these two mobile GPUs is brief but decisive, as both were subjected to two identical benchmark tests. In the Geekbench OpenCL test, the GTX 980M delivers a score of 23,832, which is a staggering 69.9% higher than the GTX 765M’s score of 7,176. This is not a marginal victory; it is a generational blowout. The GTX 980M’s result is more than triple that of its older counterpart, indicating that the Maxwell architecture in the 980M is fundamentally more efficient at raw compute tasks than the Kepler design in the 765M.

The second shared test, Geekbench Vulkan, tells a similar story. The GTX 980M scores 17,703, while the GTX 765M trails at 6,714. The delta here is 62.1% in favor of the GTX 980M. While the gap is slightly narrower than in OpenCL, it remains a dominant performance advantage. The Vulkan API is designed to reduce driver overhead and provide more direct hardware control, and the data suggests the GTX 980M’s superior hardware resources—more shaders, more memory bandwidth, and a newer architecture—allow it to exploit this API far more effectively.

It is important to contextualize these scores against the broader competitive landscape. The GTX 765M’s average benchmark score is 5,501, placing it in the 32nd percentile of all GPUs. Its nearest rivals include the NVIDIA GeForce MX130, which scores 5,508 (0.1% higher), and the AMD Radeon R7 M440, which scores 5,483 (0.3% lower). This shows that the GTX 765M, by average score, sits in a crowded mid-pack of low-end mobile parts. Its performance is essentially a dead heat with the MX130 and R7 M440, making its direct benchmark victories over those parts negligible.

Conversely, the GTX 980M has an average benchmark score of 5,308, which places it in the 31st percentile, despite its crushing victories in the head-to-head tests. This apparent contradiction is explained by its nearest rivals: the NVIDIA GeForce 930A (5,317, 0.2% higher), the GeForce 840M (5,322, 0.3% higher), and the GeForce 940M (5,284, 0.5% lower). The GTX 980M’s average score is dragged down by a wider range of benchmark results, including some older DirectX tests where its lead is less pronounced. However, in the modern compute-oriented workloads of OpenCL and Vulkan, its dominance over the GTX 765M is unmistakable. The data clearly shows that the GTX 980M is the superior performer, achieving wins in 2 out of 2 shared tests, while the GTX 765M records zero wins.

The Verdict

The benchmark data is unambiguous: the NVIDIA GeForce GTX 980M is the far more powerful GPU. Its victory margins of 69.9% in OpenCL and 62.1% in Vulkan are not just wins; they are overwhelming generational leaps. Any user prioritizing raw compute performance in modern APIs would unequivocally choose the GTX 980M. The data shows that the GTX 980M is in a different performance class entirely, even though its average benchmark score is slightly lower due to its inclusion of a broader set of tests.

However, the GTX 765M is not without a niche. Its average benchmark score of 5,501 is actually higher than the GTX 980M’s average of 5,308. This is a peculiar statistical artifact, but it suggests that in a mixed workload of older and newer tests, the GTX 765M is more consistent. Its nearest rivals, such as the MX130 and R7 M440, are all within 1% of its average score, meaning it was a solid, balanced performer for its time. For users running legacy applications or DirectX 9/10-era games, the GTX 765M might provide a more stable experience relative to its expected performance tier.

From a pure performance standpoint, the choice is clear. The GTX 980M is the pick for anyone who needs maximum compute throughput, particularly in Vulkan or OpenCL workloads. The data indicates it is more than two and a half times faster in OpenCL and over two and a half times faster in Vulkan. The GTX 765M, on the other hand, should only be considered if the system in question is legacy hardware where the lower average score is less relevant than the specific application compatibility. The verdict is simple: for modern tasks, the GTX 980M wins decisively; for a balanced legacy performer, the GTX 765M holds its own, but cannot compete with the 980M’s raw power.

FAQ

Q: How much faster is the GTX 980M than the GTX 765M in Geekbench OpenCL?

A: The GTX 980M scores 23,832 compared to the GTX 765M’s 7,176, a difference of 69.9% in favor of the GTX 980M.

Q: Which GPU has a higher average benchmark score?

A: The GTX 765M has a higher average score of 5,501, while the GTX 980M averages 5,308. This is despite the 980M winning all head-to-head tests.

Q: What is the percentile ranking for each GPU?

A: The GTX 765M is in the 32nd percentile of all GPUs, while the GTX 980M is in the 31st percentile.

Q: Who are the closest rivals to the GTX 765M based on average score?

A: The nearest rivals are 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).

Q: Does the GTX 980M win every head-to-head benchmark test listed?

A: Yes, the GTX 980M wins both shared tests: Geekbench OpenCL and Geekbench Vulkan. The GTX 765M has zero wins in the head-to-head data.

Q: What is the difference in memory bus width between the two GPUs?

A: The GTX 980M has a 256-bit memory bus, while the GTX 765M has a 128-bit bus. This contributes to the 980M’s higher bandwidth of 160.4 GB/s versus 64.13 GB/s.

Specification Differences

The specification sheets for these two GPUs highlight the substantial generational gap. The most obvious difference is in memory configuration. The GTX 980M comes equipped with 8 GB of GDDR5 memory on a 256-bit bus, delivering a bandwidth of 160.4 GB/s. In contrast, the GTX 765M has only 2 GB of GDDR5 on a 128-bit bus, yielding 64.13 GB/s of bandwidth. This represents a 2.5x increase in bandwidth for the 980M, which is critical for high-resolution textures and compute workloads.

The core configurations also differ dramatically. The GTX 980M has 1,536 shading units, 96 texture mapping units (TMUs), and 64 raster operation units (ROPs). The GTX 765M, by comparison, has 768 shading units, 64 TMUs, and only 16 ROPs. This means the 980M has double the shaders and six times the ROP count, explaining its massive lead in pixel rate (72.13 GPixel/s vs. 13.81 GPixel/s) and texture rate (108.2 GTexel/s vs. 55.23 GTexel/s). The FP32 compute performance is also vastly different: the 980M delivers 3.462 TFLOPS, while the 765M manages 1,325.6 GFLOPS.

Clock speeds are another differentiator. The GTX 980M has a base clock of 1038 MHz and a boost clock of 1127 MHz, while the GTX 765M runs at a base of 797 MHz and a boost of 863 MHz. The 980M also runs its memory at a higher effective speed of 5 Gbps, compared to 4 Gbps on the 765M. The thermal design power (TDP) is listed as 75 W for the 765M, but is not specified for the 980M, making a direct power comparison impossible. Both use an MXM-B (3.0) bus interface and have no power connectors, relying on the motherboard for power delivery. The GTX 980M is larger in terms of die size and transistor count, but both are "Portable Device Dependent" for display outputs.

Architecture Differences

The architectural divide between these two GPUs is profound, representing a full generation shift from Kepler to Maxwell 2.0. The GTX 765M is built on the GK106 chip using the Kepler architecture, while the GTX 980M uses the GM204 chip with the Maxwell 2.0 architecture. Both are manufactured by TSMC on the same 28 nm process node, but the execution is vastly different. The Kepler chip contains 2,540 million transistors on a die size of 221 mm², resulting in a transistor density of 11.5 million per mm². The Maxwell 2.0 chip, however, packs 5,200 million transistors onto a 398 mm² die, achieving a density of 13.1 million per mm².

This increase in transistor count allows Maxwell 2.0 to implement a more efficient scheduling and workload distribution system, which is why the 980M achieves higher clock speeds (1038 MHz base vs. 797 MHz) despite the larger, more complex chip. The fundamental design philosophy shifted from Kepler’s emphasis on raw shader throughput to Maxwell’s focus on per-clock efficiency and improved cache hierarchies. This is reflected in the benchmark results, where the 980M’s superior architecture yields massive gains in compute-heavy APIs like Vulkan and OpenCL.

The memory subsystem also reflects architectural changes. The 980M’s 256-bit bus, combined with a higher effective memory clock, allows for significantly more data to be moved per second, which is essential for feeding the 1,536 shading units. The 765M’s 128-bit bus is a bottleneck for its 768 shaders. In terms of API support, the 980M supports DirectX 12 (12_1) and Vulkan 1.4, whereas the 765M only reaches DirectX 12 (11_0) and Vulkan 1.2.175. This means the 980M is better positioned for modern games and applications that leverage the latest feature sets. The generational leap is clear: the GTX 980M’s Maxwell 2.0 architecture is a complete rethinking of the GPU pipeline compared to the older Kepler design in the GTX 765M, yielding superior performance and modern API compatibility.