AMD Radeon 780M vs NVIDIA GeForce GTX 690 Comparison

The AMD Radeon 780M and NVIDIA GeForce GTX 690 represent two distinct eras of GPU design, separated by over a decade of architectural evolution. The data shows a clear overall winner in the 780M, which secures a 2-0 victory in head-to-head benchmarks, yet the comparison reveals nuanced strengths for the older card. The 780M posts an average benchmark score of 17,588 against the GTX 690’s 17,037, a difference of roughly 3.2%, placing both within the same performance tier—the 780M sits at the 61st percentile of all GPUs, while the GTX 690 sits at the 60th. This narrow gap in aggregate scores, however, belies dramatic differences in individual workloads and underlying technology.

Head-to-Head Benchmarks

The most striking result comes from the Geekbench Vulkan test, where the AMD Radeon 780M achieves a score of 33,683 compared to the GTX 690’s 16,675. That is a 102% advantage—more than double the performance—and represents the single largest margin in this comparison. The 780M’s Vulkan score is not just a win; it is a demonstration of modern API efficiency, as the GTX 690’s support for Vulkan 1.2.175 lags behind the 780M’s Vulkan 1.4. This delta is so large that it alone accounts for the majority of the average score gap between the two cards.

In the Geekbench OpenCL test, the 780M again takes the lead, scoring 18,602 against the GTX 690’s 17,399. The 6.9% advantage here is more modest but still decisive. This test reflects compute-heavy workloads, where the 780M’s 8.909 TFLOPS of FP32 performance vastly outstrips the GTX 690’s 3.130 TFLOPS. Interestingly, the GTX 690’s raw shading power—1,536 shading units versus the 780M’s 768—does not translate into a win, suggesting that architectural efficiency and clock speeds matter more than sheer unit count. The 780M boosts to 2900 MHz, while the GTX 690 tops out at 1019 MHz, a 2.85x clock advantage that helps explain the performance inversion.

Neither benchmark favors the GTX 690, and the data shows zero wins for the older card. However, the OpenCL result is close enough that in real-world compute tasks with different optimization profiles, the margin could narrow. The Vulkan result, by contrast, is a categorical rout, indicating that the GTX 690 cannot compete in modern graphics APIs regardless of workload.

Architecture Differences

The architectural gap between these two GPUs is generational. The 780M is built on TSMC’s 4 nm process node, packing 25,390 million transistors into a 178 mm² die. This yields a transistor density of 142.6 million per square millimeter. The GTX 690, in contrast, uses a 28 nm process with 3,540 million transistors on a 294 mm² die, for a density of just 12.0 million per square millimeter. The 780M achieves over 11x higher transistor density, enabling far more complex logic in a smaller footprint.

The 780M uses the RDNA 3.0 architecture, part of the Navi III IGP generation, and is built around the Phoenix chip. It features 768 shading units, 48 texture mapping units, and 32 ROPs, along with 12 dedicated ray tracing cores. The GTX 690 uses the Kepler architecture with the GK104 chip, offering 1,536 shading units, 128 TMUs, and 32 ROPs, but no ray tracing hardware. The 780M’s FP32 throughput of 8.909 TFLOPS is nearly triple the GTX 690’s 3.130 TFLOPS, and it also supports FP16 at a 1:1 ratio, which the GTX 690 lacks entirely.

Memory configurations could not be more different. The GTX 690 has 2 GB of dedicated GDDR5 memory on a 256-bit bus, delivering 192.3 GB/s of bandwidth. The 780M uses System Shared memory, with its bandwidth marked as System Dependent. This means the 780M’s memory performance is tied to the host system’s RAM, which can be a bottleneck or an advantage depending on the platform. The GTX 690 also draws 300 W of power with dual 8-pin connectors and requires a 700 W power supply, while the 780M is an IGP with a 15 W TDP and no power connectors. The 780M supports PCIe 4.0 x8, whereas the GTX 690 uses PCIe 3.0 x16.

API support further separates them. The 780M offers DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The GTX 690 is limited to DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. This means the 780M can leverage modern features like ray tracing and variable rate shading, while the GTX 690 is stuck with legacy feature sets. The GTX 690’s production status is End-of-life, having launched in May 2012, while the 780M remains Active, launching in January 2024.

Where Each One Wins

The AMD Radeon 780M wins in every scenario where modern API features or compute throughput matter. Its 102% lead in Vulkan makes it the clear choice for any current-generation game or application that utilizes Vulkan’s advanced features. The 12 ray tracing cores enable hardware-accelerated ray tracing, a capability entirely absent from the GTX 690. For productivity tasks that rely on OpenCL or FP32 compute, the 780M’s 8.909 TFLOPS and 6.9% OpenCL advantage provide a tangible edge, particularly in workloads that scale with raw shader math. The 15 W TDP also makes it viable for thin-and-light laptops, where the GTX 690’s 300 W requirement is physically impossible to accommodate.

The NVIDIA GeForce GTX 690 retains advantages in specific legacy scenarios. Its 2 GB of dedicated GDDR5 memory with 192.3 GB/s bandwidth is fixed and predictable, whereas the 780M’s System Shared memory depends entirely on system RAM speed and capacity. In a system with slow or insufficient RAM, the GTX 690 could outperform the 780M in memory-bound tasks. The GTX 690 also has double the shading units (1,536 vs 768) and over double the TMUs (128 vs 48), which could benefit workloads that are heavily texture-bound and do not leverage modern APIs. Its 279 mm length and dual-slot design are irrelevant for performance but indicate a card built for desktop expansion, not integrated efficiency.

For raw pixel throughput, the 780M wins with 92.80 GPixel/s versus the GTX 690’s 32.61 GPixel/s. Texture rate is close, with the 780M at 139.2 GTexel/s and the GTX 690 at 130.4 GTexel/s. The 780M’s higher clocks and newer architecture drive these wins, even with fewer execution units. The GTX 690’s only clear victory is in memory bandwidth per watt, but since its total power draw is 20x higher, that is a pyrrhic win.

FAQ

Q: Which GPU has higher raw compute performance?

A: The AMD Radeon 780M delivers 8.909 TFLOPS of FP32 performance, nearly triple the GTX 690’s 3.130 TFLOPS. This is reflected in the OpenCL benchmark, where the 780M scores 18,602 versus 17,399.

Q: Can the GTX 690 handle modern graphics APIs?

A: The GTX 690 supports DirectX 12 (11_0) and Vulkan 1.2.175, but not DirectX 12 Ultimate or Vulkan 1.4. The 780M supports both, plus 12 ray tracing cores, giving it a massive advantage in modern workloads.

Q: How does memory configuration affect performance?

A: The GTX 690 has 2 GB of dedicated GDDR5 with 192.3 GB/s bandwidth. The 780M uses System Shared memory with System Dependent bandwidth. The GTX 690 offers predictable, fixed memory performance, while the 780M’s performance varies with the host system’s RAM.

Q: What is the power consumption difference?

A: The 780M has a 15 W TDP and requires no power connectors, while the GTX 690 has a 300 W TDP and needs dual 8-pin connectors plus a 700 W power supply. The 780M is an IGP, the GTX 690 is a dual-slot desktop card.

Q: Which GPU has better Vulkan performance?

A: The 780M wins decisively, scoring 33,683 versus 16,675 in Geekbench Vulkan, a 102% advantage. This is the largest margin in the entire comparison and indicates the GTX 690 cannot compete in Vulkan-based applications.

Q: Are these GPUs in the same performance tier?

A: Yes, their average benchmark scores are close—17,588 for the 780M and 17,037 for the GTX 690—placing them at the 61st and 60th percentiles of all GPUs, respectively. However, the distribution of wins heavily favors the 780M.

The Verdict

The data is unambiguous: the AMD Radeon 780M is the superior GPU for virtually all modern use cases. Its 102% lead in Vulkan performance is the headline statistic, but the 6.9% OpenCL win, 2.85x boost clock advantage, and 8.909 TFLOPS of FP32 throughput cement its position. The 780M’s 4 nm process node with 142.6M transistors per mm² versus the GTX 690’s 12.0M per mm² shows a generational leap in efficiency, allowing it to deliver more performance at 15 W than the GTX 690 does at 300 W. For anyone building a new system or upgrading from an older discrete GPU, the 780M offers modern API support, ray tracing, and compute power that the GTX 690 simply cannot match.

The NVIDIA GeForce GTX 690 is not without merit in this comparison. Its 2 GB of dedicated GDDR5 memory with 192.3 GB/s bandwidth provides a fixed, reliable memory subsystem that the 780M cannot guarantee due to its System Dependent bandwidth. In scenarios where system RAM is slow or shared with other components, the GTX 690’s dedicated memory could offer more consistent performance. Its 1,536 shading units and 128 TMUs also give it an edge in legacy, texture-heavy workloads that do not utilize modern API features.

The verdict depends on the user’s context. For a modern system, the 780M is the obvious choice—it wins both head-to-head benchmarks, offers ray tracing, and consumes 20x less power. For a retro or specialized build where dedicated GDDR5 memory and legacy architecture are required, the GTX 690 remains functional, but the data shows it loses in every benchmark recorded here. The 780M’s active production status and 2024 release date also ensure driver support and compatibility, while the GTX 690 is end-of-life. The numbers point one way: the 780M is the better GPU, and the GTX 690 is a historical artifact that, while competitive in average score, cannot keep pace in modern workloads.