AMD Radeon HD 8690M vs NVIDIA Quadro M500M Comparison

FAQ

Q: Which GPU is faster in OpenCL compute workloads?

A: The AMD Radeon HD 8690M scores 6,137 in Geekbench OpenCL, while the NVIDIA Quadro M500M scores 5,986. The AMD card wins by 2.5%.

Q: How do these two cards compare to their closest rivals?

A: The AMD Radeon HD 8690M sits 0.3% ahead of the Intel Iris Pro Graphics 6200 and 1% ahead of both the NVIDIA RTX A400 and NVIDIA GeForce MX230. The NVIDIA Quadro M500M is 1.2% ahead of the AMD FirePro M4000 and 1.7% ahead of the NVIDIA GeForce MX130.

Q: What is the difference in memory bandwidth between the two?

A: The AMD Radeon HD 8690M delivers 32.00 GB/s using GDDR5 memory, while the NVIDIA Quadro M500M provides 14.40 GB/s with DDR3 memory. That is a substantial gap favoring AMD.

Q: Which card has a higher boost clock speed?

A: The NVIDIA Quadro M500M boosts to 1,124 MHz, while the AMD Radeon HD 8690M boosts to 975 MHz. NVIDIA runs about 149 MHz faster at peak.

Q: Do both cards support the same DirectX version?

A: No. The AMD Radeon HD 8690M supports DirectX 12 (11_1), while the NVIDIA Quadro M500M supports DirectX 12 (11_0). Both support OpenGL 4.6, but NVIDIA lists Vulkan 1.4 compared to AMD's Vulkan 1.2.170.

Q: What is the transistor density difference between the two chips?

A: The NVIDIA Quadro M500M has a transistor density of 13.2M per mm², while the AMD Radeon HD 8690M has 12.3M per mm². NVIDIA packs transistors slightly more densely despite both using a 28 nm TSMC process.

Architecture Differences

The AMD Radeon HD 8690M is built on GCN 1.0 architecture using the Sun chip, while the NVIDIA Quadro M500M employs Maxwell architecture with the GM108S chip. Both are fabricated on a 28 nm process at TSMC, but the underlying designs diverge significantly.

The AMD chip contains 690 million transistors on a 56 mm² die, while the NVIDIA chip packs 1,020 million transistors onto a 77 mm² die. This gives NVIDIA a higher transistor density at 13.2M per mm² versus AMD's 12.3M per mm². The larger transistor budget allows the Quadro M500M to field 384 shading units compared to AMD's 320, a 64-unit advantage.

Texture and rasterization resources tell a different story. The AMD Radeon HD 8690M has 20 texture mapping units (TMUs) and 8 raster operation units (ROPs). The NVIDIA Quadro M500M has 16 TMUs and 8 ROPs. AMD's extra 4 TMUs help its texture throughput, while ROP counts are identical.

Clock behavior also differs. The AMD card runs at a base of 925 MHz with a boost of 975 MHz. The NVIDIA card starts at 1,029 MHz and boosts to 1,124 MHz, giving it a meaningful frequency advantage in both states.

Memory architecture shows a stark split. AMD uses 2 GB of GDDR5 on a 64-bit bus, reaching 32.00 GB/s of bandwidth. NVIDIA also has 2 GB but uses DDR3 on the same 64-bit bus, capping at 14.40 GB/s. That is less than half of AMD's bandwidth, which has direct implications for memory-bound workloads.

The API feature sets differ slightly. AMD supports DirectX 12 (11_1), while NVIDIA supports DirectX 12 (11_0). Both expose OpenGL 4.6, but NVIDIA's Vulkan support is listed at 1.4 versus AMD's 1.2.170.

Interface and power characteristics also diverge. The AMD card uses PCIe 3.0 x8, while the NVIDIA card is an MXM-A (3.0) module. NVIDIA lists a TDP of 30 W with no power connectors, indicating an integrated mobile solution. AMD does not list TDP in the database.

Head-to-Head Benchmarks

The only direct benchmark comparison in the database is Geekbench OpenCL. Here, the AMD Radeon HD 8690M scores 6,137 against the NVIDIA Quadro M500M's 5,986. The delta is 2.5% in AMD's favor, making it the sole head-to-head win and giving AMD a 1-0 record in wins.

This result is notable because the NVIDIA card has more shading units and higher clocks. The data suggests that AMD's memory bandwidth advantage of 32.00 GB/s versus 14.40 GB/s, combined with its wider texture throughput, offsets NVIDIA's compute unit count and frequency lead.

Looking at broader context, the AMD card's 6,137 OpenCL score places it near the NVIDIA RTX A400 (6,078, 1% slower) and the NVIDIA GeForce MX230 (6,077, 1% slower). It also edges past the Intel Iris Pro Graphics 6200 (6,117, 0.3% slower). The AMD Radeon HD 6950 sits slightly ahead at 6,210, which is 1.2% higher than the HD 8690M.

The NVIDIA Quadro M500M's OpenCL score of 5,986 sits close to its rivals. The AMD FirePro M4000 trails by 1.2% at 5,537, while the NVIDIA GeForce MX130 is 1.7% behind at 5,508, and the NVIDIA GeForce GTX 765M is 1.9% behind at 5,501. The AMD Radeon HD 8790M leads the M500M by 1.5% with a score of 5,691.

The Quadro M500M also has a Vulkan benchmark score of 5,222 in the database, though no comparable Vulkan result exists for the AMD card. This limits direct API-level comparison to OpenCL only.

Specification Differences

The two cards differ across several specification fields. The process node is identical at 28 nm, and both use TSMC as the foundry. Transistor counts differ: AMD has 690 million, NVIDIA has 1,020 million. Die sizes are 56 mm² for AMD and 77 mm² for NVIDIA, resulting in transistor densities of 12.3M per mm² and 13.2M per mm² respectively.

Base clocks are 925 MHz for AMD and 1,029 MHz for NVIDIA. Boost clocks are 975 MHz for AMD and 1,124 MHz for NVIDIA. Memory clock differs as well: AMD runs at 1000 MHz with 4 Gbps effective, while NVIDIA runs at 900 MHz with 1800 Mbps effective.

Memory type is a key differentiator: GDDR5 for AMD, DDR3 for NVIDIA. Both have 2 GB capacity and a 64-bit bus width. Bandwidth is 32.00 GB/s for AMD and 14.40 GB/s for NVIDIA.

Shading units number 320 for AMD and 384 for NVIDIA. TMUs are 20 for AMD and 16 for NVIDIA. ROPs are equal at 8. Pixel rate is 7.800 GPixel/s for AMD and 8.992 GPixel/s for NVIDIA. Texture rate is 19.50 GTexel/s for AMD and 17.98 GTexel/s for NVIDIA. FP32 performance is 624.0 GFLOPS for AMD and 863.2 GFLOPS for NVIDIA.

Power and interface specs differ. NVIDIA lists a TDP of 30 W, while AMD has no listed TDP. NVIDIA's slot width is MXM Module, with no power connectors. AMD uses PCIe 3.0 x8; NVIDIA uses MXM-A (3.0). Display outputs are portable device dependent for NVIDIA, while AMD does not list them.

API support differs in DirectX and Vulkan versions. DirectX is 12 (11_1) for AMD and 12 (11_0) for NVIDIA. Vulkan is 1.2.170 for AMD and 1.4 for NVIDIA. OpenGL is 4.6 for both.

Where Each One Wins

The AMD Radeon HD 8690M wins in OpenCL compute performance, taking the only direct head-to-head benchmark by 2.5%. Its memory bandwidth of 32.00 GB/s is more than double the NVIDIA card's 14.40 GB/s, making it better suited for bandwidth-sensitive tasks such as large data transfers, texture-heavy workloads, and OpenCL kernels that frequently access memory. Its higher texture rate of 19.50 GTexel/s versus 17.98 GTexel/s also favors AMD in scenarios involving rapid texture sampling.

The NVIDIA Quadro M500M wins in raw compute throughput. Its FP32 output of 863.2 GFLOPS exceeds AMD's 624.0 GFLOPS by about 38%. Its pixel rate of 8.992 GPixel/s is also higher than AMD's 7.800 GPixel/s. The higher base and boost clocks, along with 384 shading units versus 320, give NVIDIA an edge in compute-bound operations that are not limited by memory bandwidth. The Vulkan 1.4 support also indicates more modern API exposure, though no Vulkan comparison exists in the database.

For general mobile use, the NVIDIA card's 30 W TDP and MXM module form factor suggest a specific integration path, whereas AMD's PCIe 3.0 x8 interface offers a different connectivity approach. The AMD card's GDDR5 memory is a clear advantage for sustained bandwidth, while NVIDIA's higher clock speeds help latency-sensitive workloads.

The Verdict

The data points to a split decision. The AMD Radeon HD 8690M is the better choice for OpenCL workloads and any task that depends on memory bandwidth. Its 2.5% head-to-head win in Geekbench OpenCL, coupled with 32.00 GB/s of bandwidth, makes it the stronger candidate for compute applications that move large data sets. The 1-0 win record in the database confirms this directly.

The NVIDIA Quadro M500M is the better choice for compute-bound tasks that rely on raw floating-point throughput and pixel fill. Its FP32 performance of 863.2 GFLOPS is approximately 38% higher than the AMD card, and its pixel rate of 8.992 GPixel/s exceeds AMD's 7.800 GPixel/s. For workloads where shading unit count and clock speed dominate, NVIDIA has the structural advantage.

The AMD card also holds a slight percentile edge, sitting at the 36th percentile of all GPUs compared to NVIDIA's 32nd. Its average benchmark score of 6,137 also exceeds NVIDIA's average of 5,604, though this includes NVIDIA's Vulkan result in the averaging.

Overall, the AMD Radeon HD 8690M wins the direct benchmark comparison, but the NVIDIA Quadro M500M offers superior compute specifications that could win in workloads not captured by the single OpenCL test. The choice depends on whether memory bandwidth or raw FP32 throughput matters more for the intended application.