NVIDIA GeForce GTX 580M vs NVIDIA GeForce GTX 680M Comparison

The NVIDIA GeForce GTX 680M is the clear performance leader, outperforming the GTX 580M by a substantial 44.5% in the only shared benchmark. This generational leap comes from a complete architecture overhaul, shifting from Fermi 2.0 on 40nm to Kepler on 28nm, resulting in a significantly higher transistor density and a much larger shader count. The data indicates that the GTX 680M is not just an incremental upgrade, but a fundamental leap in mobile graphics capability, making it the superior choice for demanding tasks.

FAQ

Q: What is the performance difference between the GTX 680M and GTX 580M?

A: In the Geekbench OpenCL benchmark, the GTX 680M scores 9230, while the GTX 580M scores 6389. This represents a 44.5% advantage for the GTX 680M.

Q: How do their average benchmark scores compare to all other GPUs?

A: The GTX 680M has an average benchmark score of 7023, placing it in the 39th percentile of all GPUs. The GTX 580M has an average score of 6389, placing it in the 37th percentile.

Q: What are the key differences in their memory subsystems?

A: The GTX 680M comes with 4 GB of GDDR5 memory on a 256-bit bus, providing 115.2 GB/s of bandwidth. The GTX 580M has 2 GB of GDDR5 memory on the same 256-bit bus, but its bandwidth is lower at 96.00 GB/s.

Q: What is the transistor count and process node for each GPU?

A: The GTX 680M uses a 28 nm process and contains 3,540 million transistors. The GTX 580M uses a larger 40 nm process and contains 1,950 million transistors.

Q: Do both GPUs support the same modern APIs?

A: Both GPUs support DirectX 12 (11_0) and OpenGL 4.6. However, the GTX 680M also supports Vulkan 1.2.175, while the GTX 580M does not have Vulkan support listed.

Q: What is the power consumption of these mobile graphics modules?

A: Both the GTX 680M and the GTX 580M have a TDP of 100 W and use an MXM Module slot width with no power connectors.

Architecture Differences

The fundamental difference lies in their architecture and manufacturing process. The GTX 680M is built on the Kepler architecture using the GK104 chip, fabricated on a 28 nm process at TSMC. In contrast, the GTX 580M utilizes the Fermi 2.0 architecture with the GF114 chip, produced on a larger 40 nm process. This process shrink is a major factor in the performance gap, as it allows for a significantly higher transistor density of 12.0M / mm² on the GTX 680M compared to 5.9M / mm² on the GTX 580M.

The transistor count reflects this architectural advancement. The GTX 680M packs 3,540 million transistors into a 294 mm² die, while the GTX 580M contains 1,950 million transistors on a larger 332 mm² die. This means the newer GPU crams nearly twice the transistors into a smaller physical space. The core configuration also differs dramatically, with the GTX 680M featuring 1344 shading units and 112 texture mapping units (TMUs), compared to 384 shading units and 64 TMUs on the GTX 580M. Both GPUs share the same 32 render output units (ROPs) and a 256-bit memory bus.

The memory technology on both is GDDR5, but the GTX 680M runs at a higher effective speed. Its memory is clocked at 3.6 Gbps effective, providing a bandwidth of 115.2 GB/s, while the GTX 580M's memory runs at 3 Gbps effective for a bandwidth of 96.00 GB/s. The GTX 680M also features a base clock of 719 MHz with a boost clock of 758 MHz, whereas the GTX 580M does not have base or boost clock data listed in the provided information.

Head-to-Head Benchmarks

The only direct benchmark comparison available is the Geekbench OpenCL test, and the results are decisive. The GTX 680M scores 9230, while the GTX 580M scores 6389. This gives the GTX 680M a commanding 44.5% performance lead. This substantial delta is not a marginal improvement but a generational leap, highlighting the efficiency and raw compute power advantages of the Kepler architecture over Fermi 2.0.

To contextualize these scores, the GTX 680M's average benchmark score of 7023 places it in the 39th percentile. It sits slightly behind the NVIDIA T600, which scores 7035, and just ahead of the AMD Radeon R5 M240, which scores 6975. The GTX 580M's average score of 6389 places it in the 37th percentile, matching the NVIDIA GeForce GTX 460 SE exactly and trailing the NVIDIA RTX PRO 5000 72 GB Blackwell by a mere 0.3%. While the percentile difference between the two mobile GPUs is only two points, the raw score difference in the head-to-head test is significant, showing a clear performance hierarchy where the GTX 680M is the superior part.

Specification Differences

The specifications that differ between these two GPUs are extensive, reflecting their different generations and architectures.

• Architecture: The GTX 680M uses Kepler, while the GTX 580M uses Fermi 2.0.
• Process Node: The GTX 680M is built on 28 nm, while the GTX 580M uses 40 nm.
• Transistors: The GTX 680M has 3,540 million transistors, compared to the GTX 580M's 1,950 million.
• Die Size: The GTX 680M has a die size of 294 mm², while the GTX 580M's is larger at 332 mm².
• Transistor Density: The GTX 680M has a density of 12.0M / mm², versus 5.9M / mm² for the GTX 580M.
• Base Clock: The GTX 680M has a base clock of 719 MHz; the GTX 580M has no listed base clock.
• Boost Clock: The GTX 680M has a boost clock of 758 MHz; the GTX 580M has no listed boost clock.
• Memory Clock: The GTX 680M's memory runs at 3.6 Gbps effective, while the GTX 580M's runs at 3 Gbps effective.
• Memory Size: The GTX 680M has 4 GB of memory, double the GTX 580M's 2 GB.
• Memory Bandwidth: The GTX 680M provides 115.2 GB/s of bandwidth, compared to 96.00 GB/s on the GTX 580M.
• Shading Units: The GTX 680M has 1344 shading units, a massive increase over the GTX 580M's 384.
• TMUs: The GTX 680M has 112 TMUs, compared to 64 on the GTX 580M.
• Pixel Rate: The GTX 680M has a pixel rate of 21.22 GPixel/s, over double the GTX 580M's 9.920 GPixel/s.
• Texture Rate: The GTX 680M has a texture rate of 84.90 GTexel/s, more than double the GTX 580M's 39.68 GTexel/s.
• FP32 Performance: The GTX 680M delivers 2.038 TFLOPS, while the GTX 580M delivers 952.3 GFLOPS.
• Vulkan Support: The GTX 680M supports Vulkan 1.2.175, but the GTX 580M has no Vulkan support listed.

Where Each One Wins

The GTX 680M wins in every measurable category where data is available. Its advantages are most pronounced in raw compute performance. In the Geekbench OpenCL test, it is 44.5% faster, and its FP32 performance of 2.038 TFLOPS is more than double the GTX 580M's 952.3 GFLOPS. This makes the GTX 680M the clear choice for compute-heavy applications like OpenCL-accelerated workloads, video editing, and 3D rendering. Its higher texture rate (84.90 GTexel/s vs 39.68 GTexel/s) and pixel rate (21.22 GPixel/s vs 9.920 GPixel/s) also indicate a significant advantage in traditional 3D gaming, where filling pixels and applying textures are critical.

The GTX 580M does not have a single benchmark win in the provided data. While it holds its own in the 37th percentile, its architectural age is a clear handicap. Its lower memory bandwidth of 96.00 GB/s and smaller 2 GB frame buffer could be a limiting factor in modern games with high-resolution textures. However, its 100 W TDP matches the newer card, meaning it doesn't offer an efficiency benefit. The GTX 580M's only potential advantage is its slightly larger die size, but that is a physical characteristic and not a performance benefit.

The Verdict

The data is unambiguous: the NVIDIA GeForce GTX 680M is the superior GPU. Its 44.5% lead in the head-to-head OpenCL benchmark, combined with its more than double FP32 performance and higher memory bandwidth, makes it the definitive choice for users seeking maximum performance in a mobile workstation or gaming laptop. The GTX 680M's 4 GB of memory also provides a future-proofing advantage over the GTX 580M's 2 GB.

The GTX 580M, while a capable part in its own right (matching the GTX 460 SE in average score), is a product of an older generation. Its Fermi 2.0 architecture and 40 nm process node put it at a significant disadvantage in both performance and efficiency metrics. For anyone choosing between these two, the GTX 680M is the only logical pick based on the benchmark results. The GTX 580M should only be considered if the system requires a specific legacy compatibility that the newer card does not offer, though no such data is present here. The verdict is simple: the GTX 680M wins on all fronts.