NVIDIA GeForce GT 745M vs NVIDIA GeForce RTX 3080 12 GB Comparison

Head-to-Head Benchmarks

The NVIDIA GeForce GT 745M and the NVIDIA GeForce RTX 3080 12 GB represent two entirely different eras of GPU design, and the recorded data reflects that gulf. The GT 745M, a Kepler-based mobile part from the GeForce 700M generation, posts an average benchmark score of 3953 across its tested workloads. The RTX 3080 12 GB, an Ampere flagship from the GeForce 30-series, records an average benchmark score of 4791 in its sole tested workload, 3DMark Steel Nomad DX12. That is a raw delta of 838 points, or roughly 21% higher average score for the newer card, though the comparison is complicated by the fact that the two GPUs were not subjected to the same suite of tests.

Looking at the individual benchmark entries, the GT 745M has three recorded scores: Geekbench Metal at 2777, Geekbench OpenCL at 3580, and Geekbench Vulkan at 5502. Its strongest showing comes in Vulkan, where it reaches 5502, while its Metal score lags significantly at 2777. The RTX 3080 12 GB has only one recorded benchmark, 3DMark Steel Nomad DX12, at 4791. Because the benchmark suites do not overlap, a direct apples-to-apples comparison on identical workloads is impossible from the database. However, the percentile rankings place both cards in a similar low-to-mid tier relative to all GPUs: the GT 745M sits at the 23rd percentile, while the RTX 3080 12 GB sits at the 28th percentile. This is a striking result, given the massive architectural and generational differences, and it suggests that the RTX 3080 12 GB's single recorded score may not reflect its true standing among the broader GPU population.

The nearest rival data for each card further illustrates how the database situates them. The GT 745M's closest competitors by average score are the AMD Radeon R5 M420 (3956, a 0.1% deficit), the NVIDIA GeForce 830M (3957, a 0.1% deficit), the NVIDIA Quadro K2000 (3964, a 0.3% deficit), and the NVIDIA Quadro 2000D (3930, a 0.6% advantage for the GT 745M). These are all modest deltas, indicating that the GT 745M sits in a tight cluster of low-end and older discrete GPUs. The RTX 3080 12 GB's nearest rivals, by contrast, are the AMD Radeon R5 M255 (4788, a 0.1% advantage for the RTX card), the AMD Radeon R5 M335 (4752, a 0.8% advantage), the NVIDIA GeForce 940MX (4844, a 1.1% deficit), and the NVIDIA GeForce GTX 560M (4855, a 1.3% deficit). All of these rivals are themselves older or lower-tier parts, which reinforces the notion that the RTX 3080 12 GB's single 3DMark score of 4791 places it in an unusual spot: it is barely ahead of or slightly behind mobile GPUs from earlier generations, despite being a desktop-class part with vastly higher specifications.

The biggest win for the GT 745M comes in the Geekbench Vulkan test, where it scores 5502. That figure is higher than the RTX 3080 12 GB's 4791 in Steel Nomad, though again these are different workloads with different API demands. The biggest win for the RTX 3080 12 GB is its average benchmark score of 4791, which exceeds the GT 745M's average of 3953 by 838 points. In percentage terms, the RTX 3080 12 GB's average is about 21% higher. Yet neither card dominates its nearest rivals by more than a couple of percentage points in either direction, which suggests that the database's scoring methodology produces a narrow band for these particular products.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce RTX 3080 12 GB has an average benchmark score of 4791, compared to the NVIDIA GeForce GT 745M's average of 3953, a difference of 838 points.

Q: How do the two cards rank among all GPUs in the database?

A: The GT 745M sits at the 23rd percentile of all GPUs, while the RTX 3080 12 GB sits at the 28th percentile. Despite the RTX card's much higher specifications, both land in the same general tier by percentile.

Q: What is the RTX 3080 12 GB's closest rival according to the database?

A: The AMD Radeon R5 M255 is the closest rival, with an average score of 4788, a 0.1% difference from the RTX 3080 12 GB's 4791. The NVIDIA GeForce 940MX (4844) and NVIDIA GeForce GTX 560M (4855) are slightly ahead of it.

Q: What is the GT 745M's closest rival according to the database?

A: The AMD Radeon R5 M420 is the closest, with an average score of 3956, a 0.1% difference from the GT 745M's 3953. The NVIDIA GeForce 830M (3957) is also within 0.1%.

Q: Does the GT 745M outperform the RTX 3080 12 GB in any recorded benchmark?

A: In the Geekbench Vulkan test, the GT 745M scores 5502, which is higher than the RTX 3080 12 GB's single recorded score of 4791 in 3DMark Steel Nomad DX12. However, these are different workloads, so the comparison is not direct.

Q: What is the RTX 3080 12 GB's launch MSRP?

A: The launch MSRP is 799 USD.

Where Each One Wins

The GT 745M wins in scenarios where its specific benchmark strengths matter. Its Geekbench Vulkan score of 5502 is the highest single result recorded for either card in this comparison, and its Geekbench OpenCL score of 3580 and Metal score of 2777 provide a broader profile across multiple API environments. For workloads that rely on Vulkan compute or graphics, the GT 745M's recorded data suggests it can punch above its low percentile ranking. Its 23rd percentile placement, combined with its tight cluster of rivals (all within 0.6% of its average score), indicates that it is a predictable, mid-pack performer among the low-end GPUs in the database. The GT 745M also wins on power efficiency in a practical sense: its TDP is 45 W, which is far lower than the RTX 3080 12 GB's 350 W, making it suitable for systems with limited thermal and power headroom, such as the integrated-class form factor (IGP slot width) it occupies.

The RTX 3080 12 GB wins on raw average score, with 4791 versus 3953, and it also wins on specification-driven capabilities that are not directly captured in the benchmark data. Its 12 GB of GDDR6X memory, 384-bit bus, and 912.4 GB/s bandwidth are far beyond the GT 745M's 2 GB GDDR5, 128-bit bus, and 64.00 GB/s. The RTX card's 70 RT cores and 280 tensor cores enable hardware-accelerated ray tracing and AI workloads, which the GT 745M cannot perform at all. Its 30.64 TFLOPS of FP32 compute and matching 30.64 TFLOPS of FP16 (1:1) dwarf the GT 745M's 421.6 GFLOPS of FP32. The RTX 3080 12 GB also supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, whereas the GT 745M is limited to DirectX 12 (11_0) and Vulkan 1.2.175. For any modern, demanding workload, the RTX 3080 12 GB is the clear choice, despite its modest percentile ranking. The database's nearest rival data for the RTX card shows it trading blows with the Radeon R5 M255, R5 M335, GeForce 940MX, and GTX 560M, but those are all older or lower-end parts; the RTX 3080 12 GB's actual hardware capabilities are in a different class.

Specification Differences

The two GPUs differ in nearly every measurable specification. The GT 745M uses a 28 nm process from TSMC, while the RTX 3080 12 GB uses an 8 nm process from Samsung. Transistor counts are 1,270 million for the GT 745M versus 28,300 million for the RTX 3080 12 GB, and die sizes are 118 mm² versus 628 mm² respectively. Transistor density is 10.8M per mm² for the GT 745M and 45.1M per mm² for the RTX 3080 12 GB. Memory configurations are starkly different: the GT 745M has 2 GB GDDR5 on a 128-bit bus with 64.00 GB/s bandwidth, while the RTX 3080 12 GB has 12 GB GDDR6X on a 384-bit bus with 912.4 GB/s bandwidth. Memory clocks are 1000 MHz (4 Gbps effective) for the GT 745M and 1188 MHz (19 Gbps effective) for the RTX 3080 12 GB.

Compute resources differ by an order of magnitude or more. The GT 745M has 384 shading units, 32 TMUs, and 16 ROPs, whereas the RTX 3080 12 GB has 8960 shading units, 280 TMUs, and 96 ROPs. The RTX card also has 70 RT cores and 280 tensor cores; the GT 745M has none. Pixel rates are 4.392 GPixel/s versus 164.2 GPixel/s, and texture rates are 17.57 GTexel/s versus 478.8 GTexel/s. FP32 compute is 421.6 GFLOPS versus 30.64 TFLOPS. The RTX 3080 12 GB also records FP16 at 30.64 TFLOPS (1:1), while the GT 745M has no listed FP16 figure. TDP is 45 W for the GT 745M and 350 W for the RTX 3080 12 GB. The GT 745M has no power connector listed, while the RTX 3080 12 GB uses a 1x 12-pin connector and a suggested PSU of 750 W. The GT 745M is IGP slot width, while the RTX 3080 12 GB is dual-slot. The GT 745M's dimensions are listed as "Portable Device Dependent" for display outputs, while the RTX 3080 12 GB has 1x HDMI 2.1 and 3x DisplayPort 1.4a. The RTX 3080 12 GB has physical dimensions of 285 mm (11.2 inches) in length, 112 mm (4.4 inches) in height, and 40 mm (1.6 inches) in width; the GT 745M has no listed dimensions. Bus interfaces are PCIe 3.0 x16 for the GT 745M and PCIe 4.0 x16 for the RTX 3080 12 GB.

Architecture Differences

The GT 745M is built on the Kepler architecture, using the GK107 chip, and belongs to the GeForce 700M generation. The RTX 3080 12 GB is built on the Ampere architecture, using the GA102 chip, and belongs to the GeForce 30-series. Kepler predates Ampere by multiple generations, and the architectural advances are reflected in the API support and feature sets. The GT 745M supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The RTX 3080 12 GB supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The RTX 3080 12 GB's inclusion of RT cores and tensor cores marks a fundamental architectural shift toward hybrid rendering and AI acceleration, capabilities that Kepler lacks entirely. The GT 745M's predecessor is the GeForce 600M and its successor is the GeForce 800M, placing it in a product line that has since been superseded. The RTX 3080 12 GB's predecessor is the GeForce 20 series and its successor is the GeForce 40 series. Both cards are marked as end-of-life in the database, with the GT 745M released in March 2013 and the RTX 3080 12 GB released in January 2022. The foundry change from TSMC to Samsung, the node shrink from 28 nm to 8 nm, and the massive increase in transistor density (from 10.8M to 45.1M per mm²) all point to the same conclusion: these are products from different technological eras, and the architectural gap is far wider than the benchmark scores alone suggest.