NVIDIA GeForce GTX 1660 Ti vs NVIDIA GeForce GTX 580M Comparison

Head-to-Head Benchmarks

The only shared benchmark between these two GPUs is Geekbench OpenCL, and the result is not close. The GTX 1660 Ti scores 51,029, while the GTX 580M scores 6,389. That is a delta of 698.7%, meaning the newer card is roughly eight times faster in raw compute throughput. In practical terms, any OpenCL workload that runs on the GTX 580M will see a dramatic improvement on the GTX 1660 Ti, often turning slideshows into fluid interactions.

This single data point is the entire head-to-head comparison in the database. The GTX 166 Ti wins the only shared test, giving it one win and the GTX 580M zero. However, that does not tell the whole story. The GTX 166 Ti also has a much richer benchmark history overall, with ten recorded tests spanning DirectX 9 through 12, Vulkan, and compute workloads. The GTX 580M has just one recorded test, which limits direct comparisons but does not change the outcome where they overlap.

Looking at the broader database context, the GTX 166 Ti sits at the 41st percentile among all GPUs, with an average benchmark score of 7,723. Its nearest rivals in the database are the AMD Radeon 540 (7,673, 0.7% slower), the AMD Radeon Pro WX 3100 (7,580, 1.9% slower), the AMD Radeon R7 250 (7,557, 2.2% slower), and the Intel Arc A310 (7,550, 2.3% slower). These are tight margins, indicating that the GTX 166 Ti is not a standout among its direct peers, but it is solidly mid-pack.

The GTX 580M, by contrast, sits at the 37th percentile, with an average benchmark score of 6,389. Its nearest rivals include the NVIDIA GeForce GTX 460 SE (6,389, 0% difference), the NVIDIA RTX PRO 5000 72 GB Blackwell (6,407, 0.3% faster), the AMD Radeon Pro WX 4100 (6,330, 0.9% slower), and the AMD Radeon R7 M350 (6,327, 1% slower). The GTX 580M is effectively tied with a decade-old desktop card, and it is even slightly behind a modern workstation GPU in the database, though by less than half a percent.

FAQ

Q: Which card has the higher average benchmark score?

A: The GTX 166 Ti averages 7,723 across its recorded tests, while the GTX 580M averages 6,389. That is a 20.9% gap in the database.

Q: Is the GTX 580M competitive in any modern workload?

A: Based on the data, the GTX 580M only has one recorded benchmark (Geekbench OpenCL at 6,389), and it loses that test to the GTX 166 Ti by 698.7%. There is no test in the database where the GTX 580M comes out ahead.

Q: How does the GTX 166 Ti compare to its nearest rivals?

A: The GTX 166 Ti is 0.7% faster than the AMD Radeon 540, 1.9% faster than the AMD Radeon Pro WX 3100, 2.2% faster than the AMD Radeon R7 250, and 2.3% faster than the Intel Arc A310. These are all within a few percent, so the card is not dominant in its tier.

Q: What is the GTX 580M's closest competitor?

A: The NVIDIA GeForce GTX 460 SE has an identical average score of 6,389, with a 0% delta. The NVIDIA RTX PRO 5000 72 GB Blackwell is 0.3% faster, and the AMD Radeon Pro WX 4100 is 0.9% slower.

Q: Which card supports newer DirectX features?

A: The GTX 166 Ti supports DirectX 12 (12_1), while the GTX 580M supports DirectX 12 (11_0). The GTX 166 Ti also supports Vulkan 1.4, whereas the GTX 580M has no recorded Vulkan support in the database.

Q: Does the GTX 580M have any advantage in memory bandwidth per watt?

A: The database does not record a power efficiency metric for either card. The GTX 580M has a lower TDP (100 W vs 120 W), but it also has far lower memory bandwidth (96.00 GB/s vs 288.0 GB/s) and a much lower pixel rate (9.920 GPixel/s vs 84.96 GPixel/s), so any efficiency edge is moot given the massive performance gap.

Architecture Differences

The GTX 166 Ti is built on the TU116 chip using the Turing architecture, fabricated on a 12 nm process at TSMC. It packs 6,600 million transistors into a 284 mm² die, giving a transistor density of 23.2 million per square millimeter. The GTX 580M uses the GF114 chip with the Fermi 2.0 architecture, also from TSMC but on a 40 nm process. It contains 1,950 million transistors on a larger 332 mm² die, with a density of only 5.9 million per square millimeter.

This process gap is fundamental. The 12 nm node allows the GTX 166 Ti to nearly triple the transistor count on a smaller die, which translates directly into higher shader counts and faster clocks. The GTX 166 Ti has 1,536 shading units, 96 texture mapping units, and 48 raster output units. The GTX 580M has 384 shading units, 64 TMUs, and 32 ROPs. That is a 4x difference in shader count, 1.5x in TMUs, and 1.5x in ROPs.

Clock speeds also show the generational leap. The GTX 166 Ti has a base clock of 1500 MHz and a boost clock of 1770 MHz. The GTX 580M has no base or boost clock recorded in the database, but its memory runs at 750 MHz (3 Gbps effective), while the GTX 166 Ti memory runs at 1500 MHz (12 Gbps effective). Even without core clock data for the older card, the memory clock alone is a 4x difference.

Neither card has ray tracing cores or tensor cores. The GTX 166 Ti is a raster-only Turing part, and the GTX 580M predates those features entirely. The GTX 166 Ti supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The GTX 580M supports DirectX 12 (11_0) and OpenGL 4.6, but has no Vulkan entry in the database.

Specification Differences

The memory subsystem is a clear differentiator. The GTX 166 Ti has 6 GB of GDDR6 on a 192-bit bus, delivering 288.0 GB/s of bandwidth. The GTX 580M has 2 GB of GDDR5 on a 256-bit bus, delivering 96.00 GB/s. Despite the narrower bus, the newer memory technology gives the GTX 166 Ti a 3x bandwidth advantage.

Pixel and texture throughput follow the same pattern. The GTX 166 Ti achieves 84.96 GPixel/s and 169.9 GTexel/s. The GTX 580M achieves 9.920 GPixel/s and 39.68 GTexel/s. The FP32 compute is 5.437 TFLOPS versus 952.3 GFLOPS, a 5.7x gap. The GTX 166 Ti also supports FP16 at 10.87 TFLOPS (2:1 rate), while the GTX 580M has no FP16 data recorded.

Power and physical design differ significantly. The GTX 166 Ti has a TDP of 120 W, uses a dual-slot cooler, requires a single 8-pin power connector, and has a suggested PSU of 300 W. Its dimensions are 229 mm in length, 111 mm in height, and 35 mm in width. The GTX 580M has a TDP of 100 W, is an MXM module, has no power connectors, and has no suggested PSU. It has no recorded dimensions. The bus interface also differs: PCIe 3.0 x16 for the GTX 166 Ti versus MXM-B (3.0) for the GTX 580M.

Display outputs separate them as well. The GTX 166 Ti offers 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a. The GTX 580M is listed as "Portable Device Dependent," meaning it relies on the laptop or mobile workstation's own outputs. The GTX 166 Ti is a desktop card, while the GTX 580M is a mobile module.

Release timing and market positioning also differ. The GTX 166 Ti launched on 2019-02-21 with a launch MSRP of 279 USD. The GTX 580M launched on 2011-06-27 with no MSRP recorded. The GTX 166 Ti belongs to the GeForce 16 generation, succeeds the GeForce 10 series, and is succeeded by GeForce 20. The GTX 580M belongs to the GeForce 500M generation, succeeds GeForce 400M, and is succeeded by GeForce 600M.

Where Each One Wins

The GTX 166 Ti wins every recorded benchmark it participates in. Its ten tests show strong results across the board: PassMark G3D at 12,871, Geekbench OpenCL at 51,029, Geekbench Vulkan at 5,827, PassMark GPU Compute at 5,226, and older DirectX tests like PassMark DirectX 9 at 190, DirectX 11 at 103, and DirectX 10 at 66. The low DirectX 12 score of 51 and Steel Nomad DX12 score of 1,061 indicate it is not a top-tier card for the latest APIs, but it handles legacy and compute workloads competently.

The GTX 580M has no wins in the database. Its single Geekbench OpenCL score of 6,389 is its only data point, and it loses to the GTX 166 Ti by 698.7%. The card is at the 37th percentile, just below the GTX 166 Ti's 41st percentile. In terms of actual use, the GTX 580M might still run older DirectX 9 or OpenGL titles that are not demanding, but the database offers no evidence of any modern workload where it is preferable.

The GTX 166 Ti also wins on architectural features that matter for longevity. It supports Vulkan 1.4, which opens the door to modern cross-platform titles, while the GTX 580M has no Vulkan support. It has 4x the memory capacity, 3x the bandwidth, and a 5.7x FP32 compute advantage. For any task that is GPU-bound, the GTX 166 Ti is the clear choice.

The Verdict

The data is unambiguous. The GTX 166 Ti is faster in the only shared benchmark, has a higher average score (7,723 vs 6,389), sits at a higher percentile (41 vs 37), and has a far more complete feature set. The GTX 580M is a legacy mobile chip from 2011, built on a 40 nm process, with a single recorded benchmark and no wins.

For a desktop user, the GTX 166 Ti is the only sensible pick. It offers modern API support, 6 GB of GDDR6, and enough compute to handle DirectX 11 and Vulkan titles at playable settings. It is not a high-end card by any means, as its 41st percentile and tight margins against rivals like the AMD Radeon 540 show, but it is a capable 1080p option.

For a laptop or mobile workstation, the GTX 580M is a historical artifact. It could be found in old gaming laptops, but its 2 GB of GDDR5 and 96.00 GB/s bandwidth limit it to very old or very light games. There is no scenario in the recorded data where the GTX 580M outperforms the GTX 166 Ti.

The GTX 166 Ti's launch MSRP was 279 USD, but that is not relevant to its current standing. The card is end-of-life, and its performance tier is now occupied by newer parts. Still, within the database, it is a solid mid-range GPU with a clear edge over its Fermi-era predecessor. The GTX 580M should be considered only for retro computing or compatibility testing, not for any modern workload.