AMD Radeon RX 5500M vs NVIDIA GeForce GTX 680 Comparison

The benchmark data is unambiguous: the AMD Radeon RX 5500M is the outright performance winner in every single head-to-head comparison against the NVIDIA GeForce GTX 680. However, the GTX 680 holds a remarkable position in the overall rankings given its age, demonstrating that the two cards occupy different performance tiers despite their similar average benchmark scores.

Head-to-Head Benchmarks

The RX 5500M dominates every shared benchmark test, often by a substantial margin. In the Geekbench Metal test, the RX 5500M scores 50,359 points compared to the GTX 680's 7,897 points. This represents a delta of -84.3% from the perspective of the GTX 680, meaning the AMD card is more than six times faster in this specific API workload. This is the single largest performance gap between the two cards in any test.

The Geekbench OpenCL benchmark shows a narrower but still decisive victory for the RX 5500M. The AMD card posts 38,725 points against the GTX 680's 16,906 points, a delta of -56.3%. In percentage terms, the RX 5500M delivers roughly 2.3 times the OpenCL performance of the older NVIDIA card. This suggests that while the RX 5500M excels across different compute APIs, its advantage is most pronounced in Metal-based workloads.

The closest contest appears in the Geekbench Vulkan test, where the RX 5500M scores 35,693 points versus the GTX 680's 17,646 points. The delta here is -50.6%, indicating that the AMD card is still about twice as fast, but the gap is slightly smaller than in the other two tests. This pattern suggests that the RX 5500M's architecture handles modern graphics APIs more efficiently, while the GTX 680's Kepler design shows its age in newer compute frameworks.

The final tally is decisive: the RX 5500M wins all three head-to-head benchmarks, while the GTX 680 records zero wins. Despite this sweep, the average benchmark scores tell a more nuanced story. The GTX 680 has an average benchmark score of 14,150 across all its tested workloads, while the RX 5500M averages 13,356. This discrepancy occurs because the RX 5500M has a much broader benchmark suite, including several Passmark tests where its scores are relatively low, such as 39 in DirectX 10 and 35 in DirectX 11.

Architecture Differences

The two GPUs represent fundamentally different eras of graphics design. The GTX 680 uses the GK104 chip built on NVIDIA's Kepler architecture, manufactured on a 28 nm process at TSMC. It packs 3,540 million transistors into a die size of 294 mm², resulting in a transistor density of 12.0 million transistors per square millimeter. The RX 5500M, by contrast, uses the Navi 14 chip based on AMD's RDNA 1.0 architecture, also fabbed by TSMC but on a much more advanced 7 nm process. This newer node allows AMD to fit 6,400 million transistors into a smaller 158 mm² die, achieving a density of 40.5 million transistors per square millimeter — more than three times the density of the GTX 680.

Clock speeds show a similar generational leap. The GTX 680 runs at a base clock of 1006 MHz with a boost clock of 1058 MHz. The RX 5500M operates at a significantly higher 1375 MHz base clock and 1645 MHz boost clock, with a game clock of 1448 MHz. These higher clocks, combined with the architectural efficiency of RDNA 1.0, contribute to the RX 5500M's superior compute throughput. The AMD card delivers 4.632 TFLOPS of FP32 performance, while the GTX 680 manages 3.250 TFLOPS. The RX 5500M also supports FP16 at 9.265 TFLOPS (2:1 ratio), a feature the GTX 680 lacks entirely.

Memory configurations differ substantially. The GTX 680 uses 2 GB of GDDR5 memory on a 256-bit bus, providing 192.3 GB/s of bandwidth. The RX 5500M has 4 GB of GDDR6 memory on a narrower 128-bit bus, yet achieves higher bandwidth at 224.0 GB/s thanks to faster effective memory speeds of 14 Gbps versus the GTX 680's 6 Gbps. The RX 5500M also supports PCIe 4.0 x8, while the GTX 680 uses PCIe 3.0 x16.

Shader and texture configurations show both similarities and differences. The GTX 680 has 1536 shading units, 128 texture mapping units, and 32 ROPs. The RX 5500M has 1408 shading units, fewer at 88 TMUs, but the same 32 ROPs. Despite having fewer shading units and TMUs, the RX 5500M achieves higher pixel rate (52.64 GPixel/s versus 33.86 GPixel/s) and texture rate (144.8 GTexel/s versus 135.4 GTexel/s) due to its higher clock speeds.

Where Each One Wins

The RX 5500M is the clear winner for anyone prioritizing modern compute performance. Its dominance in all three Geekbench tests — Metal, OpenCL, and Vulkan — makes it the superior choice for GPU-accelerated workloads such as video encoding, machine learning inference, and content creation applications that leverage these APIs. The AMD card's higher FP32 throughput, FP16 support, and greater memory bandwidth position it as the more capable card for contemporary software that has been optimized for recent GPU architectures.

The GTX 680, despite losing every head-to-head benchmark, still holds its own in the broader performance landscape. Its average benchmark score of 14,150 places it in the 55th percentile of all GPUs, slightly above the RX 5500M's 54th percentile and 13,356 average. This suggests that in legacy DirectX workloads and older games — particularly those that were optimized for the Kepler architecture — the GTX 680 remains a competent performer. The card's 2 GB of VRAM and 256-bit memory bus were considered generous for its era, and its 33.86 GPixel/s pixel rate still handles traditional rasterization tasks respectably.

For gaming specifically, the RX 5500M's DirectX 12 support (version 12_1) versus the GTX 680's more limited DirectX 12 (11_0) support gives it a clear advantage in modern titles that use advanced DirectX 12 features. The RX 5500M also supports Vulkan 1.4, a newer version than the GTX 680's Vulkan 1.2.175, making it better prepared for current and upcoming game engines. The GTX 680's OpenGL 4.6 support matches the RX 5500M's, so legacy OpenGL titles should perform similarly on both cards.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 680 has a higher average benchmark score of 14,150, compared to the AMD Radeon RX 5500M's 13,356. However, this includes different benchmark suites, as the RX 5500M has additional Passmark tests that lower its average.

Q: How much faster is the RX 5500M in Geekbench Metal?

A: The RX 5500M scores 50,359 points in Geekbench Metal, while the GTX 680 scores 7,897 points. This represents a delta of -84.3% from the GTX 680's perspective, meaning the RX 5500M is approximately 6.4 times faster in this test.

Q: What are the memory specifications of each card?

A: The GTX 680 has 2 GB of GDDR5 memory on a 256-bit bus with 192.3 GB/s bandwidth. The RX 5500M has 4 GB of GDDR6 memory on a 128-bit bus with 224.0 GB/s bandwidth.

Q: Which card has a higher transistor density?

A: The RX 5500M has a transistor density of 40.5 million transistors per square millimeter, while the GTX 680 has 12.0 million per square millimeter. The RX 5500M achieves this on a 7 nm process, versus the GTX 680's 28 nm process.

Q: Do both cards support the same DirectX version?

A: No. The RX 5500M supports DirectX 12 (12_1), while the GTX 680 supports DirectX 12 (11_0). Both cards support OpenGL 4.6, but the RX 5500M supports Vulkan 1.4 versus the GTX 680's Vulkan 1.2.175.

Q: What is the power consumption difference?

A: The GTX 680 has a TDP of 195 W and requires 2x 6-pin power connectors with a suggested PSU of 450 W. The RX 5500M has a TDP of 85 W and requires no power connectors, making it suitable for portable devices.

Specification Differences

The two cards differ across nearly every major specification category. The GTX 680 uses a 28 nm process, while the RX 5500M uses 7 nm. Transistor counts differ significantly: 3,540 million for the GTX 680 versus 6,400 million for the RX 5500M, with corresponding die sizes of 294 mm² and 158 mm². Clock speeds show the RX 5500M running higher at 1375 MHz base and 1645 MHz boost, compared to the GTX 680's 1006 MHz base and 1058 MHz boost.

Memory configurations differ in capacity (2 GB versus 4 GB), type (GDDR5 versus GDDR6), bus width (256-bit versus 128-bit), and bandwidth (192.3 GB/s versus 224.0 GB/s). The RX 5500M has fewer shading units (1408 versus 1536) and TMUs (88 versus 128), but identical ROPs at 32. The RX 5500M delivers higher pixel rate (52.64 GPixel/s versus 33.86 GPixel/s) and texture rate (144.8 GTexel/s versus 135.4 GTexel/s).

FP32 performance favors the RX 5500M at 4.632 TFLOPS versus 3.250 TFLOPS, and the RX 5500M adds FP16 capability at 9.265 TFLOPS. Power requirements differ dramatically: 195 W TDP with 2x 6-pin connectors for the GTX 680, versus 85 W TDP with no connectors for the RX 5500M. The GTX 680 is a dual-slot card measuring 256 mm in length, while the RX 5500M is a mobile part with portable-device-dependent display outputs. The GTX 680 uses PCIe 3.0 x16, while the RX 5500M uses PCIe 4.0 x8.

The Verdict

The AMD Radeon RX 5500M is the definitive choice for anyone requiring modern compute performance, particularly in Metal, OpenCL, and Vulkan workloads. Its complete sweep of the head-to-head benchmarks — with margins ranging from 50.6% to 84.3% — demonstrates a clear generational advantage in architectural efficiency and raw compute capability. The RX 5500M's higher clock speeds, newer 7 nm process, superior memory bandwidth, and support for DirectX 12 (12_1) and Vulkan 1.4 make it the better option for contemporary applications and games.

The NVIDIA GeForce GTX 680, however, should not be dismissed entirely. Its higher average benchmark score and 55th percentile ranking indicate that it remains competitive in legacy workloads, particularly those optimized for older DirectX versions and Kepler-era feature sets. The GTX 680's 256-bit memory bus and 2 GB of GDDR5 memory were impressive for its time, and its 3.250 TFLOPS of FP32 performance is still respectable for basic compute tasks.

The decision ultimately depends on the intended use case. For gaming on modern titles, content creation with current software, or any workload that leverages Vulkan or DirectX 12_1 features, the RX 5500M is the clear winner. For users running legacy applications, older games, or software specifically tuned for NVIDIA's Kepler architecture, the GTX 680 remains a viable option. The RX 5500M's lower power requirement of 85 W versus 195 W also makes it far more practical for portable devices, though the GTX 680's dual-slot design and established driver support may appeal to desktop users with older systems.