NVIDIA GeForce GTS 450 vs NVIDIA GeForce GTX 980M Comparison

Head-to-Head Benchmarks

The only directly comparable benchmark between these two GPUs in the database is Geekbench OpenCL. In this test, the NVIDIA GeForce GTX 980M scores 23,832 points, while the NVIDIA GeForce GTS 450 scores 4,893 points. This represents a delta of 387.1% in favor of the GTX 980M. The GTX 980M does not merely outperform the GTS 450; it delivers nearly five times the raw compute throughput in this workload. To put that in context, the GTX 980M’s average benchmark score across all recorded tests is 5,308, while the GTS 450’s average is 4,893, meaning the GTS 450 sits only slightly below the GTX 980M’s overall average even when comparing the newer card’s weakest aggregated metric against the older card’s sole data point.

The GTS 450 has only one benchmark entry in the database, the Geekbench OpenCL test, which makes cross-card comparisons limited to that single workload. The GTX 980M, by contrast, has ten recorded benchmarks spanning DirectX 9, 10, 11, and 12, OpenCL, Vulkan, and general-purpose compute tests. In every single test where both cards have data, the GTX 980M wins. The head-to-head win count stands at 1 win for the GTX 980M and 0 wins for the GTS 450.

Looking at the GTX 980M’s own benchmark suite, its strongest results appear in Passmark G3D with 7,338 points, followed by Geekbench OpenCL at 23,832 and Geekbench Vulkan at 17,703. Its Passmark G2D score of 490 and GPU compute score of 2,816 show that even its less impressive metrics still exceed the GTS 450’s entire recorded output. The GTS 450’s 4,893 OpenCL score is lower than every single one of the GTX 980M’s ten benchmark results, including the lowest GTX 980M score, which is 31 in Passmark DirectX 12.

The percentile rankings reinforce this gap. The GTX 980M sits at the 31st percentile among all GPUs in the database, while the GTS 450 sits at the 28th percentile. Although the percentile difference seems modest, the actual score gap is enormous because the percentile distribution is compressed at the lower end. The GTX 980M’s nearest rivals include the NVIDIA GeForce 930A (average score 5,317, delta -0.2%), the NVIDIA GeForce 840M (5,322, -0.3%), the NVIDIA GeForce 940M (5,284, +0.5%), and the AMD Radeon R7 M445 (5,358, -0.9%). These rivals are all within roughly 1% of the GTX 980M’s average, indicating that the GTX 980M sits in a tightly contested mid-range mobile segment. The GTS 450’s nearest rivals are entirely different: the NVIDIA GeForce RTX 5060 Ti 8 GB (4,901, -0.2%), AMD FirePro W5130M (4,904, -0.2%), AMD Radeon R6 M255DX (4,867, +0.5%), and AMD Radeon R7 M265 (4,929, -0.7%). Notably, the GTS 450 is listed as a direct rival to the RTX 5060 Ti 8 GB in this database, with a delta of -0.2%, meaning the GTS 450’s single benchmark score is nearly identical to that modern card’s average in this particular metric. That is a historical curiosity, but it does not change the fact that the GTX 980M is in a completely different performance class.

The delta between the two cards in the head-to-head test is 387.1%, which is the single largest percentage gap recorded in this comparison. No other benchmark exists where the GTS 450 could potentially close that gap, because the database contains no other shared test between the two. Therefore, the data unequivocally shows the GTX 980M as the dominant performer in every measurable dimension.

Where Each One Wins

The GTX 980M wins in every scenario where benchmark data exists. For compute-heavy workloads such as OpenCL, the GTX 980M’s 23,832 score versus the GTS 450’s 4,893 demonstrates a massive advantage in parallel processing capability. This translates to applications like video encoding, physics simulation, and general GPU-accelerated computing, where the GTX 980M would complete tasks in a fraction of the time. The GTX 980M also shows strength in modern graphics APIs: its Vulkan score of 17,703 and Passmark DirectX 12 score of 31 indicate that it can handle contemporary rendering pipelines, whereas the GTS 450 has no recorded Vulkan result and its DirectX support tops out at version 12 (11_0), meaning it lacks full DirectX 12_1 features.

For legacy DirectX workloads, the GTX 980M still leads decisively. Its Passmark DirectX 9 score of 125, DirectX 10 score of 35, and DirectX 11 score of 57 all exceed the GTS 450’s only benchmark, which is not even a DirectX test. The GTS 450’s architecture was designed for the DirectX 11 era, but without a recorded DirectX benchmark, the database cannot quantify its legacy performance. What the data does show is that the GTS 450’s OpenCL score is only 20.5% of the GTX 980M’s OpenCL score, so even in the GTS 450’s native era, the newer card would likely dominate.

The GTX 980M also wins in memory-bound scenarios. With 8 GB of GDDR5 memory on a 256-bit bus delivering 160.4 GB/s of bandwidth, versus the GTS 450’s 1024 MB on a 128-bit bus at 57.73 GB/s, the GTX 980M can feed its 1,536 shading units far more effectively. The GTS 450’s 192 shading units, 32 texture mapping units, and 16 ROPs are simply outmatched by the GTX 980M’s 1,536 shading units, 96 TMUs, and 64 ROPs. The pixel rate tells the story: 72.13 GPixel/s for the GTX 980M versus 6.264 GPixel/s for the GTS 450, a factor of 11.5. Texture rate is 108.2 GTexel/s versus 25.06 GTexel/s, a factor of 4.3.

There is no use-case where the GTS 450 wins based on recorded data. The only nuanced point is that the GTS 450, being an end-of-life product with a 106 W TDP and a dual-slot design, might be suitable for legacy systems with PCIe 2.0 x16 slots, while the GTX 980M requires an MXM-B (3.0) interface, which is portable-device dependent. But for pure performance, the GTX 980M wins all categories.

Architecture Differences

The GTX 980M is built on the GM204 chip using NVIDIA’s Maxwell 2.0 architecture, fabricated on a 28 nm process at TSMC. It contains 5,200 million transistors on a 398 mm² die, yielding a transistor density of 13.1 million transistors per square millimeter. The GTS 450 uses the GF106 chip with the older Fermi architecture, also fabricated at TSMC but on a 40 nm process. It packs 1,170 million transistors onto a 238 mm² die, for a density of 4.9 million per square millimeter. The architectural generation gap is significant: Maxwell 2.0 was designed for improved efficiency per clock and better geometry performance, while Fermi was NVIDIA’s first unified shader architecture, which was powerful in its day but far less efficient.

The GTX 980M has 1,536 shading units, 96 TMUs, and 64 ROPs. The GTS 450 has 192 shading units, 32 TMUs, and 16 ROPs. This is an 8x difference in shader count, 3x in TMUs, and 4x in ROPs. The GTX 980M’s FP32 throughput is 3.462 TFLOPS, while the GTS 450 manages 601.3 GFLOPS, a 5.75x advantage. Neither card has RT cores or tensor cores, and neither has a recorded FP16 rate, so ray tracing and AI acceleration are not differentiators here.

The GTX 980M supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The GTS 450 supports DirectX 12 (11_0), OpenGL 4.6, but has no Vulkan support. This means the GTS 450 cannot run Vulkan workloads at all, while the GTX 980M has a dedicated Vulkan score of 17,703. The GTX 980M’s DirectX 12_1 feature level enables advanced rendering techniques like conservative rasterization and rasterizer-ordered views, which the GTS 450’s 11_0 feature level cannot provide.

Memory architecture also diverges sharply. The GTX 980M uses 8 GB of GDDR5 on a 256-bit bus with 160.4 GB/s bandwidth. The GTS 450 uses 1024 MB of GDDR5 on a 128-bit bus with 57.73 GB/s. The GTX 980M’s memory clock is 1253 MHz (5 Gbps effective), while the GTS 450’s memory clock is 902 MHz (3.6 Gbps effective). The GTX 980M’s larger bus width and higher clock speed combine for nearly 2.8x the bandwidth.

The process node difference (28 nm versus 40 nm) is fundamental. A smaller node allows more transistors in the same area, which is why the GTX 980M achieves 13.1M transistors per mm² versus 4.9M for the GTS 450. Even though the GTX 980M’s die is larger at 398 mm², its transistor density is 2.7x higher. This directly translates to the performance gap observed in the benchmarks.

Specification Differences

The following specifications differ between the two cards, based only on recorded data:

• Process node: 28 nm (GTX 980M) versus 40 nm (GTS 450)
• Transistors: 5,200 million versus 1,170 million
• Die size: 398 mm² versus 238 mm²
• Transistor density: 13.1M / mm² versus 4.9M / mm²
• Memory size: 8 GB versus 1024 MB
• Memory bus width: 256 bit versus 128 bit
• Memory bandwidth: 160.4 GB/s versus 57.73 GB/s
• Memory clock: 1253 MHz (5 Gbps effective) versus 902 MHz (3.6 Gbps effective)
• Shading units: 1536 versus 192
• TMUs: 96 versus 32
• ROPs: 64 versus 16
• Pixel rate: 72.13 GPixel/s versus 6.264 GPixel/s
• Texture rate: 108.2 GTexel/s versus 25.06 GTexel/s
• FP32: 3.462 TFLOPS versus 601.3 GFLOPS
• TDP: No recorded TDP for GTX 980M; GTS 450 has 106 W
• Slot width: MXM Module (GTX 980M) versus Dual-slot (GTS 450)
• Power connectors: None (GTX 980M) versus 1x 6-pin (GTS 450)
• Suggested PSU: None recorded for GTX 980M; 300 W for GTS 450
• Bus interface: MXM-B (3.0) versus PCIe 2.0 x16
• Display outputs: Portable Device Dependent versus 2x DVI, 1x mini-HDMI 1.3a
• DirectX: 12 (12_1) versus 12 (11_0)
• Vulkan: 1.4 versus null
• Release date: 2014-10-06 versus 2010-09-12
• Predecessor: GeForce 800M versus GeForce 200
• Successor: GeForce 10 Mobile versus GeForce 500
• Dimensions (GTS 450 only): 210 mm length, 111 mm height; GTX 980M has no recorded dimensions
• Launch MSRP (GTS 450 only): 129 USD; GTX 980M has no recorded launch MSRP

The GTX 980M has no recorded base or boost clock in the database, while the GTS 450 also lacks those fields. The GTX 980M has no recorded TDP, which is consistent with its MXM module form factor where power delivery is handled by the host device.

FAQ

Q: Which GPU has a higher Geekbench OpenCL score?

A: The NVIDIA GeForce GTX 980M scores 23,832, while the NVIDIA GeForce GTS 450 scores 4,893, giving the GTX 980M a 387.1% advantage.

Q: Does the GTS 450 support Vulkan?

A: No, the GTS 450 has a null Vulkan API entry, while the GTX 980M supports Vulkan 1.4 and has a recorded Vulkan score of 17,703.

Q: What is the memory bandwidth difference?

A: The GTX 980M has 160.4 GB/s bandwidth on a 256-bit bus, while the GTS 450 has 57.73 GB/s on a 128-bit bus. The GTX 980M also has 8 GB of memory versus 1024 MB.

Q: How many shading units does each card have?

A: The GTX 980M has 1,536 shading units, 96 TMUs, and 64 ROPs. The GTS 450 has 192 shading units, 32 TMUs, and 16 ROPs.

Q: What are the release dates?

A: The GTX 980M was released on 2014-10-06, and the GTS 450 was released on 2010-09-12.

Q: Which card has a higher average benchmark score?

A: The GTX 980M has an average benchmark score of 5,308 across ten tests, while the GTS 450 has an average of 4,893 from a single test. The GTX 980M also ranks at the 31st percentile among all GPUs, versus 28th for the GTS 450.

Q: What is the GTS 450’s launch MSRP?

A: The GTS 450 has a recorded launch MSRP of 129 USD. The GTX 980M has no recorded launch MSRP.