NVIDIA GeForce GTX 680M vs NVIDIA GeForce GTX 980 Comparison

The Verdict

The data presents a decisive outcome: the NVIDIA GeForce GTX 980 outperforms the NVIDIA GeForce GTX 680M in every recorded benchmark. The GTX 980 wins 2 out of 2 head-to-head tests, while the GTX 680M records zero wins. This is not a close contest; the performance gap is substantial and consistent across both compute and graphics workloads.

The GTX 980 is the clear choice for users who prioritize raw performance and are working with desktop platforms that can accommodate its dual-slot cooler and 267 mm length. Its average benchmark score of 8167 places it in the 43rd percentile of all GPUs, while the GTX 680M's average score of 7023 lands in the 39th percentile. The 16.3% difference in average scores translates to a tangible performance advantage in real-world applications.

The GTX 680M, meanwhile, is a mobile-oriented part built as an MXM module. It consumes less power (100 W TDP versus 165 W for the GTX 980) and requires no external power connectors, making it suitable for laptop implementations where space and thermal constraints are critical. However, the benchmark data shows that this efficiency comes at a steep performance cost. Users who need a mobile solution with the best possible performance should still consider the GTX 980 only if their chassis can support a desktop-class card, which is rare; otherwise, the GTX 680M remains the only viable option from this pair for portable systems.

For desktop builders, the GTX 980 is the unambiguous recommendation. For laptop users, the GTX 680M is the only realistic choice, but the data makes clear that it sacrifices over 70% of the GTX 980's compute throughput in OpenCL and Metal workloads.

Where Each One Wins

The GTX 980 wins in every category where both cards have recorded data. The head-to-head results show a 214.9% advantage in Geekbench Metal and a 275.7% advantage in Geekbench OpenCL. These are not marginal leads; the GTX 980 more than triples the GTX 680M's OpenCL score and more than doubles its Metal score.

The GTX 680M's only "wins" are in the absence of competition. It has no recorded scores in DirectX tests, PassMark suites, or 3DMark Steel Nomad, so the GTX 980 also holds unopposed leads in those categories. The GTX 980's PassMark G3D score of 11095 and GPU Compute score of 4753 have no corresponding GTX 680M data points, meaning the GTX 680M cannot claim any performance advantage in any measurable metric within this database.

For users running Metal-based applications (common on macOS platforms), the GTX 980's 15163 score versus 4815 for the GTX 680M represents a massive real-world difference in graphics and compute performance. For OpenCL workloads, which span a wider range of professional and scientific applications, the GTX 980's 34676 versus 9230 is an even larger gap. The GTX 680M's only practical benefit is its lower power draw and compact MXM form factor, but neither of those translates to a benchmark win.

Architecture Differences

The GTX 980 uses the GM204 chip built on the Maxwell 2.0 architecture, while the GTX 680M uses the GK104 chip based on the older Kepler architecture. Both are manufactured on the same 28 nm process node at TSMC, but the GTX 980 packs significantly more hardware into its die.

The GTX 980 contains 5,200 million transistors on a 398 mm² die, yielding a transistor density of 13.1 million per square millimeter. The GTX 680M has 3,540 million transistors on a 294 mm² die, with a density of 12.0 million per square millimeter. The GTX 980's larger die and higher transistor count enable substantial architectural improvements, including 2048 shading units versus 1344, 128 texture mapping units versus 112, and 64 raster output units versus 32.

Clock speeds also favor the GTX 980 significantly. It runs at a base clock of 1127 MHz with a boost of 1216 MHz, while the GTX 680M operates at 719 MHz base and 758 MHz boost. This clock advantage compounds with the higher core counts to produce the GTX 980's 4.981 TFLOPS of FP32 compute versus 2.038 TFLOPS for the GTX 680M.

Memory configurations are similar in capacity (both 4 GB GDDR5 on a 256-bit bus), but the GTX 980's memory runs at 1753 MHz (7 Gbps effective) delivering 224.4 GB/s of bandwidth, while the GTX 680M's memory at 900 MHz (3.6 Gbps effective) provides only 115.2 GB/s. The GTX 980 also supports DirectX 12 (12_1) and Vulkan 1.4, compared to the GTX 680M's DirectX 12 (11_0) and Vulkan 1.2.175. Both support OpenGL 4.6.

The GTX 980 is a dual-slot desktop card requiring 2x 6-pin power connectors and a 450 W power supply, with display outputs including DVI, HDMI 2.0, and three DisplayPort 1.2 connections. The GTX 680M is an MXM module with no external power connectors and display outputs dependent on the portable device. The GTX 980 was released in September 2014, while the GTX 680M launched in June 2012, representing two full generations of architectural advancement.

FAQ

Q: Which GPU has higher raw compute performance in OpenCL?

A: The GTX 980 scores 34676 in Geekbench OpenCL, which is 275.7% higher than the GTX 680M's score of 9230. This means the GTX 980 delivers more than three times the OpenCL compute throughput.

Q: What is the difference in Metal performance between the two?

A: The GTX 980 scores 15163 in Geekbench Metal, while the GTX 680M scores 4815. The GTX 980 leads by 214.9%, more than doubling the GTX 680M's Metal performance.

Q: How do their memory bandwidths compare?

A: The GTX 980 provides 224.4 GB/s of bandwidth using GDDR5 memory at 7 Gbps effective on a 256-bit bus. The GTX 680M provides 115.2 GB/s using GDDR5 at 3.6 Gbps effective on the same 256-bit bus width. The GTX 980 has nearly double the memory bandwidth.

Q: Are both GPUs on the same manufacturing process?

A: Yes, both use TSMC's 28 nm process node. However, the GTX 980 packs 5,200 million transistors on a 398 mm² die, while the GTX 680M has 3,540 million transistors on a 294 mm² die.

Q: Which GPU supports newer graphics APIs?

A: The GTX 980 supports DirectX 12 (12_1) and Vulkan 1.4, while the GTX 680M supports DirectX 12 (11_0) and Vulkan 1.2.175. Both support OpenGL 4.6.

Q: What is the power consumption difference?

A: The GTX 980 has a 165 W TDP and requires 2x 6-pin power connectors with a 450 W suggested power supply. The GTX 680M has a 100 W TDP and requires no external power connectors, reflecting its mobile MXM design.

Head-to-Head Benchmarks

The database contains two direct head-to-head comparisons between these GPUs, and both show overwhelming wins for the GTX 980.

The first test, Geekbench Metal, records a GTX 980 score of 15163 against a GTX 680M score of 4815. This represents a 214.9% delta in favor of the GTX 980. In practical terms, any Metal-based rendering or compute task will complete in less than half the time on the GTX 980, assuming the rest of the system is not a bottleneck. The 10348-point absolute difference is larger than the GTX 680M's entire score, which underscores the generational leap between Kepler and Maxwell 2.0.

The second test, Geekbench OpenCL, shows an even wider margin. The GTX 980 scores 34676, while the GTX 680M scores 9230. The 275.7% delta means the GTX 980 delivers 3.76 times the OpenCL performance of the GTX 680M. This gap is consistent with the architectural differences: the GTX 980 has 2048 shading units versus 1344, a 1.52x core advantage, combined with a 1.57x clock advantage (1127 MHz versus 719 MHz base) and a 1.94x memory bandwidth advantage (224.4 GB/s versus 115.2 GB/s). When these factors compound, the resulting performance ratio aligns closely with the observed benchmark deltas.

The GTX 980 also holds unopposed scores in several other tests. In PassMark G3D, it scores 11095, which places it among mid-range desktop GPUs. Its PassMark GPU Compute score of 4753 indicates strong general-purpose compute capability. The GTX 980 achieves 77.82 GPixel/s pixel fill rate and 155.6 GTexel/s texture fill rate, compared to 21.22 GPixel/s and 84.90 GTexel/s for the GTX 680M. These fill rates directly impact resolution scaling and texture-heavy workloads.

In 3DMark Steel Nomad DX12, the GTX 980 scores 474, a relatively low absolute number that reflects the demanding nature of this test, but no comparable GTX 680M score exists in the database. Similarly, the GTX 980's DirectX 9 score of 164, DirectX 11 score of 83, and DirectX 12 score of 46 have no GTX 680M counterparts, meaning the GTX 680M cannot claim any lead in legacy or modern DirectX workloads.

The average benchmark scores further contextualize the performance gap. The GTX 980 averages 8167 across all recorded benchmarks, while the GTX 680M averages 7023. This 1144-point difference is somewhat compressed because the GTX 680M only has two recorded benchmarks (both compute-oriented), while the GTX 980 has eleven. If the GTX 680M were subjected to the same battery of DirectX and PassMark tests, its average would likely drop further, given that its architecture lacks the raw throughput to compete in those workloads.

The nearest rival data also places the two cards in different competitive brackets. The GTX 980's closest competitors (GTX 950M at 8135, Radeon R9 M360 at 8129, GeForce 945M at 8099, GRID K2 at 8080) are all within 1.1% of its average score, indicating it sits at the top of a dense cluster of mid-range mobile and entry desktop parts. The GTX 680M's nearest rivals (T600 at 7035, Radeon R5 M240 at 6975, GTX 675M at 6946, GTX 970 at 7157) are similarly clustered but at a significantly lower performance level. The GTX 970, notably, is 1.9% ahead of the GTX 680M, while the GTX 950M is 0.4% behind the GTX 980, which shows how far GPU performance has advanced between these two release windows.

In summary, every measurable performance metric in the database favors the GTX 980 by a wide margin. The GTX 680M remains a functional mobile GPU, but it is outclassed by more than 200% in both available head-to-head comparisons, and its architectural limitations (fewer cores, lower clocks, less memory bandwidth) fully explain the observed results.