AMD Radeon HD 8690M vs NVIDIA GeForce GTX 460 SE Comparison

NVIDIA GeForce GTX 460 SE and AMD Radeon HD 8690M are two end-of-life graphics cards from different eras, targeting very different segments of the market. The GTX 460 SE is a desktop part from NVIDIA’s Fermi generation, while the HD 8690M is a mobile chip from AMD’s GCN 1.0 architecture. Benchmark data shows the GTX 460 SE winning the single available head-to-head test, but the two cards have distinct strengths that make them suitable for different use cases.

Where Each One Wins

The NVIDIA GeForce GTX 460 SE takes the clear win in raw compute performance. In the Geekbench OpenCL test, it scores 6389 against the HD 8690M’s 6137, a 4.1% advantage. This margin is modest but consistent, and it places the GTX 460 SE in the 37th percentile of all GPUs, one point above the HD 8690M’s 36th percentile. For tasks that rely heavily on general-purpose compute—like OpenCL-accelerated rendering, physics simulation, or data processing—the GTX 460 SE is the stronger choice. Its higher FP32 throughput of 748.8 GFLOPS versus 624.0 GFLOPS reinforces this, giving it roughly 20% more floating-point headroom. The desktop card also has a significantly wider memory bus at 256-bit versus the mobile chip’s 64-bit, delivering 108.8 GB/s of bandwidth compared to just 32.00 GB/s. That bandwidth advantage is critical for memory-bound workloads, where the GTX 460 SE can feed its shaders far more efficiently.

The AMD Radeon HD 8690M, however, wins on architectural efficiency and modern feature support. It is built on a 28 nm process, versus the GTX 460 SE’s 40 nm node, and it packs 320 shading units into a 56 mm² die—remarkably dense at 12.3M transistors per mm². The GTX 460 SE has fewer shading units (288) on a much larger 332 mm² die with a density of only 5.9M / mm². The HD 8690M also supports Vulkan 1.2.170, while the GTX 460 SE has no Vulkan support listed. For software that leverages Vulkan, the AMD card is the only option here. Its PCIe 3.0 x8 interface is also newer than the GTX 460 SE’s PCIe 2.0 x16, which can matter in systems with newer CPUs and motherboards. The HD 8690M also has double the memory capacity at 2 GB versus 1 GB, which helps in texture-heavy applications or when running multiple displays, despite its narrower bus.

Architecture Differences

The architectural divide is stark. The NVIDIA GeForce GTX 460 SE uses the GF104 chip, part of the Fermi architecture, fabricated by TSMC on a 40 nm process. It contains 1,950 million transistors on a 332 mm² die. The Fermi design emphasizes raw throughput, with 288 shading units, 48 texture mapping units (TMUs), and 32 raster operation units (ROPs). Its memory subsystem is robust: 1024 MB of GDDR5 on a 256-bit bus, yielding 108.8 GB/s bandwidth. The card runs at a memory clock of 850 MHz (3.4 Gbps effective) and supports DirectX 12 (11_0), OpenGL 4.6, but no Vulkan. Its pixel rate is 7.800 GPixel/s, and its texture rate is 31.20 GTexel/s.

The AMD Radeon HD 8690M, in contrast, uses the Sun chip, built on GCN 1.0 architecture, also by TSMC but on a 28 nm process. It has just 690 million transistors on a tiny 56 mm² die. Despite the smaller size, it features 320 shading units—more than the GTX 460 SE—but only 20 TMUs and 8 ROPs. Its memory is 2 GB of GDDR5 on a 64-bit bus, which limits bandwidth to 32.00 GB/s. The base clock is 925 MHz with a boost up to 975 MHz, and memory runs at 1000 MHz (4 Gbps effective). The HD 8690M supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. Its pixel rate matches the GTX 460 SE at 7.800 GPixel/s, but its texture rate drops to 19.50 GTexel/s. The transistor density tells the story: 12.3M / mm² for AMD versus 5.9M / mm² for NVIDIA, highlighting how much more modern the GCN design is. The HD 8690M also has a predefined boost clock, which the GTX 460 SE lacks entirely.

Head-to-Head Benchmarks

The only direct benchmark comparison available is the Geekbench OpenCL test, where the NVIDIA GeForce GTX 460 SE scores 6389 against the AMD Radeon HD 8690M’s 6137. That 4.1% delta is the entire head-to-head dataset. It is a narrow win, but it is consistent with the GTX 460 SE’s higher FP32 throughput and vastly superior memory bandwidth. In compute-heavy scenarios, the data suggests the GTX 460 SE will pull ahead by a small but measurable margin.

Looking at the nearest rival scores puts both cards in context. The GTX 460 SE’s 6389 ties the NVIDIA GeForce GTX 580M exactly (deltaPct 0), and it sits just 0.3% behind the NVIDIA RTX PRO 5000 72 GB Blackwell’s 6407. It also beats the AMD Radeon Pro WX 4100 (6330) by 0.9% and the AMD Radeon R7 M350 (6327) by 1%. This places the GTX 460 SE in a competitive zone for its era, despite its age. The HD 8690M’s 6137 is within 0.3% of the Intel Iris Pro Graphics 6200’s 6117, and 1% ahead of both the NVIDIA RTX A400 (6078) and NVIDIA GeForce MX230 (6077). However, it trails the AMD Radeon HD 6950 (6210) by 1.2%. These numbers show the HD 8690M performing slightly below its desktop-class rivals but holding its own against integrated and entry-level discrete options.

The deltaPct values highlight a key nuance: the GTX 460 SE’s nearest rivals are all within 1% of its score, meaning its performance tier is tightly packed. The HD 8690M, meanwhile, has a slightly wider spread, with the HD 6950 being the only rival more than 1% away. In practical terms, the GTX 460 SE edges out the HD 8690M in compute, but neither card is a performance champion by modern standards. Their percentiles (37th and 36th) confirm they sit in the lower-middle range of all GPUs.

FAQ

Q: Which GPU is faster in the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce GTX 460 SE scores 6389, which is 4.1% higher than the AMD Radeon HD 8690M’s 6137.

Q: Does the AMD Radeon HD 8690M have any feature the GTX 460 SE lacks?

A: Yes, the HD 8690M supports Vulkan 1.2.170, while the GTX 460 SE has no Vulkan support listed. The HD 8690M also has a boost clock (975 MHz), whereas the GTX 460 SE has no boost clock defined.

Q: How do their memory configurations compare?

A: The GTX 460 SE has 1024 MB of GDDR5 on a 256-bit bus with 108.8 GB/s bandwidth. The HD 8690M has 2 GB of GDDR5 on a 64-bit bus with only 32.00 GB/s bandwidth.

Q: Which card has more shading units?

A: The AMD Radeon HD 8690M has 320 shading units, while the NVIDIA GeForce GTX 460 SE has 288.

Q: What are their manufacturing process nodes?

A: The GTX 460 SE is built on a 40 nm process, while the HD 8690M uses a more advanced 28 nm process. Both are fabricated by TSMC.

Q: How do their transistor densities compare?

A: The HD 8690M has a transistor density of 12.3M / mm², which is over twice the GTX 460 SE’s 5.9M / mm².

The Verdict

The data points to a clear split: pick the NVIDIA GeForce GTX 460 SE if your priority is raw compute performance and memory bandwidth. Its 4.1% lead in Geekbench OpenCL, combined with 748.8 GFLOPS of FP32 throughput and 108.8 GB/s of bandwidth, makes it the stronger choice for number-crunching or high-resolution texture work. Its 256-bit memory bus is a substantial advantage over the HD 8690M’s 64-bit bus, which is critical for any workload that saturates memory. The GTX 460 SE also holds a slightly higher percentile ranking (37th versus 36th) and ties the GeForce GTX 580M in its nearest rival group, indicating solid performance for its class.

Pick the AMD Radeon HD 8690M if you need modern API support or operate in a power-constrained environment. Its Vulkan 1.2.170 compatibility is a dealbreaker for any Vulkan-based software, and its 28 nm process with 2 GB of memory makes it a more efficient, capacity-friendly option. The HD 8690M’s 320 shading units are also more numerous, which could help in shader-bound scenarios, despite the lower texture rate of 19.50 GTexel/s. It is also a mobile chip, so it is designed for laptops where the GTX 460 SE’s desktop form factor and 150 W TDP are not feasible. The GTX 460 SE requires a 450 W power supply and dual-slot cooling, while the HD 8690M has no listed TDP or power connectors, reflecting its low-power mobile nature.

In short, the GTX 460 SE wins on performance metrics, but the HD 8690M wins on efficiency and feature set. There is no single “better” card—only the one that fits your system and software requirements. For desktop compute tasks, the GTX 460 SE is the data-backed pick. For mobile use or Vulkan-based workloads, the HD 8690M is the only choice.