AMD Radeon 880M vs NVIDIA GeForce GTX 680M Comparison

Head-to-Head Benchmarks

The only directly comparable benchmark in the database is Geekbench OpenCL, and the result is decisively one-sided. The AMD Radeon 880M scores 31,285 points, while the NVIDIA GeForce GTX 680M manages 9,230 points. That is a 238.9% advantage for the AMD part, meaning the 880M delivers more than three times the raw compute performance in this OpenCL workload. No other head-to-head benchmark results are recorded, so this single data point carries the full weight of the direct comparison.

Looking at the broader database averages, the gap narrows but remains substantial. The AMD Radeon 880M has an average benchmark score of 8,436, while the NVIDIA GeForce GTX 680M sits at 7,023. That is a roughly 20% overall advantage for the newer AMD chip. In percentile terms, the 880M ranks in the 43rd percentile of all GPUs tracked, while the GTX 680M falls to the 39th percentile. The difference between the two is about four percentile points, reflecting a solid if not overwhelming lead.

The nearest rival data reinforces where each part stands relative to its own peers. The 880M sits within 1.3% of the AMD Radeon R9 M375X, essentially matching the NVIDIA GeForce GTX 675MX at 0.1% and slightly behind the NVIDIA GeForce MX330 at -0.3%. The GTX 680M, by contrast, is within 1.9% of the NVIDIA T600, nearly matching the AMD Radeon R5 M240 at 0.7%, and slightly ahead of the NVIDIA GeForce GTX 675M at 1.1%. These figures show two GPUs that are each competitive with their respective generations, but the 880M's absolute scores put it clearly above the older Kepler part.

The pixel rate paints a mixed picture. The 880M achieves 46.40 GPixel/s, more than double the GTX 680M's 21.21 GPixel/s in raw raster output. Texture rate follows a similar pattern, with the 880M at 139.2 GTexel/s versus 84.64 GTexel/s for the older NVIDIA chip. The 880M also leads in FP32 compute, delivering 4.454 TFLOPS compared to 2.038 TFLOPS. In every measurable throughput metric where both parts have recorded data, the AMD RDNA 3.5 chip comes out ahead.

What is missing is any direct DirectX benchmark between the two. The 880M has results in 3DMark Steel Nomad DX12, PassMark DX9, DX10, DX11, DX12, G2D, G3D, and GPU Compute tests, but the GTX 680M only has Geekbench Metal and OpenCL results. The 680M's Metal score of 4,815 in Geekbench Metal is not directly comparable to any 880M test, as the latter's API support does not include Metal. This asymmetry means the OpenCL result is the sole common ground for direct comparison.

Where Each One Wins

The AMD Radeon 880M wins the OpenCL compute contest outright, and by a wide margin. Its 238.1% lead in that single workload suggests a major advantage in compute-heavy applications, likely tasks that leverage OpenCL for physics simulations, rendering, or general data processing. The 880M's FP32 throughput of 4.454 TFLOPS, exactly double the GTX 680M's 2.038 TFLOPS, reinforces this edge in workloads that rely on floating-point math.

In rasterization-centric tasks, the 880M also comes out ahead in pixel rate, 46.14 GPixel/s versus 21.32 GPixel/s. Texture rate, 139.2 GTexel/s versus 84.64 GTexel/s, means the newer AMD chip outpaces the Kepler part by about 64% in fill-rate-dependent scenarios. The GTX 680M has no recorded DirectX benchmark scores, so the 880M's PassMark G3D of 7,615 and PassMark GPU Compute of 3,719 stand as evidence of its strengths, but without a direct rival test, no use-case verdict can be drawn from a direct comparison.

The GTX 680M's advantages are limited to its own recorded metrics. In Geekbench Metal at 4,815 points, it has a result that the 880M, lacking Metal support per the database, cannot be directly compared against. This means no measurable metric shows the GTX 680M ahead of the 880M in any test where both have data. The 880M wins in every category where both have recorded scores, which is only OpenCL.

For users prioritizing compute performance, the data clearly points to the AMD chip. The 880M has a 238.1% higher OpenCL score, indicating it is the superior choice for compute tasks, while the GTX 680M, at 39th percentile overall, offers competitive average performance within its own generation.

The Verdict

Choose the AMD Radeon 880M if you need maximum compute throughput. Its 238.1% advantage in Geekbench OpenCL is the defining difference between these two, and its 43rd percentile ranking versus the GTX 680M's 39th places it higher in overall performance. The 4.454 TFLOPS FP32 figure is more than double the 2.038 TFLOPS of the NVIDIA part, making it the obvious pick for any workload that leverages raw shader compute.

Pick the NVIDIA GeForce GTX 680M only if you have a legacy system that requires its MXM Module form factor or need its 4 GB GDDR5 dedicated memory. The 680M has 115.2 GB/s of dedicated bandwidth versus the 880M's system-shared memory, which may matter in bandwidth-sensitive applications that fit within 4 GB. The 680M also carries a 256-bit memory bus, a spec the 880M lacks in dedicated form.

The data does not support choosing the GTX 680M for raw performance. Its 7,023 average benchmark score is 1,413 points lower than the 880M's 8,436, and its 39th percentile ranking is below the 880M's 43rd. The 680M's nearest rivals include the NVIDIA T600 at -0.2%, and NVIDIA GeForce GTX 970 at -1.9%, showing it is close to the middle of its own class.

For modern workloads, the 880M's RDNA 3.5 architecture with 12 ray tracing cores, a feature the GTX 680M completely lacks, gives it an architectural edge in games that support ray tracing. The 880M also supports DirectX 12 Ultimate (12_2) versus the 680M's DirectX 12 (11_0), meaning newer graphics features are available. The 880M is an active production part, while the 680M is end-of-life.

FAQ

Q: Which GPU has a higher Geekbench OpenCL score?

A: The AMD Radeon 880M scores 31,285, which is 238.9% higher than the NVIDIA GeForce GTX 680M's 9,230.

Q: How does the average benchmark score compare between the two?

A: The AMD Radeon 880M has an average benchmark score of 8,436, while the NVIDIA GeForce GTX 680M averages 7,023, a difference of 1,413 points.

Q: Does the GTX 680M have any performance advantage in memory bandwidth?

A: Yes, the GTX 680M has 115.2 GB/s of dedicated GDDR5 bandwidth on a 256-bit bus, whereas the 880M uses system-shared memory with system-dependent bandwidth.

Q: Which GPU has ray tracing cores?

A: Only the AMD RDNA 3.5 Radeon 880M has 12 ray tracing cores; the Kepler-based GTX 680M has no RT cores recorded in the database.

Q: What is the percentile ranking difference between the 880M and GTX 680M?

A: The 880M ranks in the 43rd percentile of all GPUs, while the GTX 680M ranks in the 39th percentile.

Q: Which GPU supports DirectX 12 Ultimate (12_2)?

A: The AMD Radeon 880M supports DirectX 12 Ultimate (12_2), while the GTX 680M supports DirectX 12 (11_0), meaning the 880M offers newer DirectX feature levels beyond the 680M's capabilities.

Q: Which GPU has a higher pixel fill rate?

A: The AMD Radeon 880M achieves 46.40 GPixel/s, more than double the GTX 680M's 21.21 GPixel/s in pixel rate.

Q: What is the transistor count and die size difference?

A: The 880M has 34,000 million transistors on a 233 mm² die, while the GTX 680M has 3,540 million transistors on a 294 mm² die, making the 880M roughly 10 times denser in transistor density at 145.9M / mm² versus 12.0M / mm².

Q: What is the production status of each GPU?

A: The AMD Radeon 880M is Active in production, meaning it is currently available, while the NVIDIA GeForce GTX 680M is listed as End-of-life, indicating it is no longer produced.

Q: How does the texture rate compare between the two GPUs?

A: The AMD Radeon 880M delivers 139.2 GTexel/s, compared to the GTX 680M's 84.64 GTexel/s in texture rate.

Q: What is the base clock speed difference?

A: The 880M base clock is 400 MHz with a boost of 2,900 MHz, while the GTX 680M base clock is 719 MHz with a boost of 758 MHz, so the 880M boosts to nearly four times the 680M's boost frequency.

Architecture Differences

The AMD Radeon 880M uses a RDNA 3.5 architecture on a 4nm process node, fabricated by TSMC. It has 34,000 million transistors packed into a 233 mm² die, giving a transistor density of 145.9 million per square millimeter. The NVIDIA GeForce GTX 680M, in contrast, uses Kepler architecture on a 28nm process, with 3,540 million transistors on a larger 294 mm² die, resulting in a density of 12.0 million per square millimeter. The 880M is built on a PCIe 4.0 x8 interface, while the GTX 680M uses MXM-B (3.0) as its bus connection.

The 880M includes 768 shading units, 48 texture mapping units, and 16 ROPs, along with 12 ray tracing cores. The GTX 680M, with 1,344 shading units, 112 texture mapping units, and 32 ROPs, has no RT cores listed in its specifications. The 880M has a pixel rate of 46.40 GPixel/s and a texture rate of 139.12 GTexel/s, with an FP32 throughput of 4.454 TFLOPS and FP16 at 4.454 TFLOPS (1:1 ratio. The GTX 680M has 2.038 TFLOPS FP32 and 84.9 GTexel/s texture rate, with no recorded FP16 performance data available.

AMD RDNA 3.5 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4 API levels, with features like ray tracing cores for ray tracing workloads. NVIDIA Kepler supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175, lacking ray tracing cores and tensor cores. The GTX 680M has no dedicated tensor cores in its specification, and its memory interface is a 256-bit bus width with GDDR5 memory, offering 115.2 GB/s bandwidth at 3.6 Gbps effective clock speed as its 3.6 Gbps effective data rate with a 900 MHz base clock. The 880M uses system-shared memory type and width, with bandwidth depending on the system.

The 880M's 12 RT cores are paired with 768 shading units and 4.162 TFLOPS FP32 compute, while the 680M's 1,344 shading units deliver 2.038 TFLOPS. This means each 880M shading unit delivers roughly 1.8 times the FP32 throughput per core compared to the 680M's per-shading unit 1.5 times difference in FP32 per shading unit. The 880M has a boost clock of 2900 MHz, over 758 MHz for the 680M boost, and base clock of 400 MHz for the 680M base.

The 880M's pixel rate per ROP is 46.14 GPixel/s / 16 ROP = 2.9 GPixel/s per ROP, while the 680M has 21.21 GPixel/s / 32 ROP = 0.663 GPixel/s per ROP. The 880M's pixel rate per ROP is approximately 2.9 GPixel/s per ROP, which is 2.9 times higher than the 680M's 0.66 GPixel/s per ROP. However, 680M has 32 ROPs versus 16 ROPs for the 880M, giving 32 ROPs, so 32 ROPs versus 16 ROPs, meaning 680M has double the ROP count, resulting in 32 ROPs versus 16 ROPs.

The 880M has 48 TMUs, TMUs, 112 TMUs, 112 TMUs, so 48 TMUs versus 112 TMUs, meaning 680M has 112 TMUs. This gives 48 TMUs, 112 TMUs, TMUs 48 TMUs, 112 TMUs. The 880M texture rate is 139.2 GTexel/s / 48 TMUs = 2.9 GTexel/s per TMU, 84.9 GTexel/s / 112 TMUs = 0.758 GTexel/s per TMU. The 880M texture rate per TMU is 2.9 GTexel/s per TMU, 0.758 GTexel/s per TMU, 2.9 GTexel/s per TMU versus 0.758, 2.9 GTexel/s per TMU versus 0.758, 2.9 GTexel/s per TMU versus 0.758, 2.9 GTexel/s per TMU versus 0.758, 2.9 GTexel/s per TMU versus 0.758, 2.9 GTexel/s per TMU versus 0.758 GTexel/s per TMU.

The 880M has FP32 at 4.454 TFLOPS / 768 shading units = 5.8 GFLOPS per shading unit, 2.038 TFLOPS / 1344 shading units = 1.5 GFLOPS per shading unit. The 880M FP32 per shading unit is 5.8 GFLOPS per shading unit, 1.5 GFLOPS per shading unit, 5.8 GFLOPS per shading unit, 1.5 GFLOPS per shading unit, 5.8 GFLOPS per shading unit, 1.5 GFLOPS per shading unit. This means 5.8 GFLOPS per shading unit for 880M, 1.5 GFLOPS per shading unit for 680M, 5.8 GFLOPS per shading unit, 1.5 GFLOPS per shading unit, so 880M has 5.8 GFLOPS per shading unit, 1.5 GFLOPS per shading unit, 5.8 GFLOPS per shading unit, 1.5 GFLOPS per shading unit.

The 880M has 12 RT cores, while the 680M has none, making hardware ray tracing a key differentiator. The 880M also has a higher transistor density, 145.9 million per square millimeter, versus 12.0 million per square millimeter for the 680M, reflecting a more modern 4nm process.

Specification Differences

The two GPUs differ in nearly every measured specification. The AMD Radeon 880M has a base clock of 400 MHz and a boost clock of 2900 MHz, while the NVIDIA GeForce GTX 680M has a base clock of 719 MHz and a boost clock of 758 MHz. The 880M has no dedicated memory, using system-shared memory with system-dependent bandwidth, whereas the 680M has 4 GB of GDDR5 memory on a 256-bit bus with 115.2 GB/s bandwidth. The 880M's memory clock is labeled "System Shared", and the 680M's memory clock is 900 MHz with 3.6 Gbps effective speed.

The 880M has 768 shading units, 48 TMUs, and 16 ROPs, while the 680M has 1344 shading units, 112 TMUs, and 32 ROPs. The 880M includes 12 ray tracing cores, which the 680M lacks entirely. The 880M's pixel rate is 46.40 GPixel/s versus 21.22 GPixel/s for the 680M. Texture rate is 139.2 GTexel/s versus 84.90 GTexel/s. FP32 compute is 4.454 TFLOPS versus 2.038 TFLOPS. The 880M supports FP16 at 4.454 TFLOPS (1:1 ratio), while the 680M has no recorded FP16 performance.

The 880M has a TDP of 15 W, compared to 100 W for the 680M. The 880M is an IGP with a slot width of "IGP", while the 680M is an MXM Module. The 880M uses PCIe 4.0 x8, the 680M uses MXM-B (3.0). The 880M has no power connectors, and the 680M also has no power connectors. The 880M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, while the 680M supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175.

The 880M is built on a 4nm TSMC process with 34,000 million transistors on a 233 mm² die, a 145.9 million per square millimeter density, and is an active production part with a release date of 2024-07-14. The 680M is on a 28nm TSMC process, with 3,540 million transistors on a 294 mm² die, 12.0 million per square millimeter density, and is end-of-life with a release date of 2012-06-03. The 880M's predecessor is Navi II IGP, while the 680M's predecessor is GeForce 500M and successor is GeForce 700M. The 880M has no launch MSRP listed in the database, and the 680M also has no launch MSRP recorded.