AMD Radeon 760M

The AMD Radeon 760M is an integrated graphics processor built on the RDNA 3.0 architecture, manufactured on a 4 nm process at TSMC with 25,390 million transistors on a 178 mm² die. It operates with a base clock of 800 MHz and a boost clock of 2599 MHz, delivering a measured average benchmark score of 22999, which places it in the 66th percentile of all GPUs. The data indicates that this IGP occupies a peculiar performance tier, sitting within a 1% margin of several discrete desktop and mobile graphics cards, a positioning that warrants closer examination of its benchmark deltas.

Benchmark Performance

The benchmark results for the AMD Radeon 760M show a split personality depending on the API used. In Geekbench OpenCL, the GPU scores 20255, while in Geekbench Vulkan, it reaches 25742, representing a 27% improvement in the Vulkan workload over the OpenCL workload. This substantial gap suggests that the RDNA 3.0 architecture responds particularly well to the lower-level Vulkan API, possibly indicating driver optimizations or architectural efficiencies that favor explicit multi-threading and asynchronous compute.

The average benchmark score of 22999 places the 760M in a remarkable position relative to discrete GPUs. The nearest rival, the Intel Arc B580, scores 23021, which is just 0.1% higher than the 760M. That 22-point difference is statistically negligible, yet the B580 is a dedicated graphics card with its own memory subsystem, while the 760M is an integrated processor sharing system memory. The NVIDIA P106-100 scores 22950, a 0.2% deficit relative to the 760M, and the RTX 4060 Mobile scores 22729, placing it 1.2% behind. The RTX 4060 desktop scores 23181, which is 0.8% ahead of the 760M.

These deltas are extraordinarily tight, with the entire rival cluster spanning just 452 points, or roughly 2% from top to bottom. The data suggests that in synthetic compute workloads, the 760M matches or nearly matches several discrete cards that draw significantly more power and have dedicated VRAM. The pixel rate of 41.58 GPixel/s and texture rate of 83.17 GTexel/s, derived from 16 ROPs and 32 TMUs respectively, indicate that the 760M is not bandwidth-starved in the same way typical IGPs are, at least within the constraints of its system-dependent memory architecture.

Power and Cooling

The AMD Radeon 760M carries a thermal design power of just 15 W, which is remarkably low for the performance level the benchmark scores indicate. This TDP figure classifies the chip as an integrated graphics processor with a slot width of IGP, meaning it occupies no expansion slot and requires no dedicated cooling solution beyond what the host system provides. The power connectors list shows "None," so the GPU draws all power through the motherboard socket, and no PSU recommendation is specified in the data.

The 4 nm manufacturing process likely contributes to this efficiency, as the transistor density of 142.6 million transistors per square millimeter allows for high compute density within the 15 W envelope. The data does not provide a suggested PSU wattage, so any system integration guidance must remain qualitative: the 760M is a low-power IGP that will not stress typical power supplies. Cooling requirements are similarly modest, as the IGP form factor implies reliance on the system's existing thermal solution for the CPU package, which houses the GPU. The absence of a dedicated cooler or power connector reinforces that this is a drop-in component for laptops or compact desktops, not an upgrade path for existing desktops.

How It Compares

Intel Arc B580 (avg score 23021, delta -0.1%): The 760M trails the Arc B580 by a razor-thin 0.1%, which is effectively a tie. The B580 is a discrete card with its own memory, yet the 760M matches it in average benchmark score. This suggests that the 760M's system-shared memory, when paired with fast system RAM, can compensate for the lack of dedicated VRAM in synthetic workloads. The delta is so small that run-to-run variance could flip the ordering.

NVIDIA P106-100 (avg score 22950, delta +0.2%): The 760M leads the P106-100 by 0.2%, a 49-point advantage. The P106-100 is a mining-oriented card with no display outputs, so its relevance to gaming is limited, but the comparison shows the 760M outperforming a chip that was designed for compute-heavy workloads. This 0.2% lead is notable because the P106-100 likely runs at a much higher TDP than 15 W.

NVIDIA GeForce RTX 4060 (avg score 23181, delta -0.8%): The desktop RTX 4060 sits 0.8% ahead, a 182-point gap. This is the largest deficit among the rivals, but it is still under 1%, meaning the integrated 760M comes within striking distance of a modern discrete mid-range card. The RTX 4060 has dedicated VRAM and higher power limits, yet the 760M's RDNA 3.0 architecture closes the synthetic benchmark gap to under one percent.

NVIDIA GeForce RTX 4060 Mobile (avg score 22729, delta +1.2%): The 760M leads the RTX 4060 Mobile by 1.2%, a 270-point margin. This is the only rival with a delta exceeding 1%, and it shows the 760M outperforming a mobile discrete card that draws significantly more power. The mobile 4060 likely operates under thermal constraints, but the 760M's 15 W envelope still manages to surpass it in average benchmark scores.

Who Should Consider It

Based on the benchmark data, the AMD Radeon 760M is positioned for users who need modest gaming and compute performance without a discrete GPU. The 66th percentile ranking against all GPUs means it outperforms the majority of graphics processors in the database, but the synthetic scores of around 23000 in Geekbench indicate it is not a high-end part. For 1080p gaming at medium settings, the 760M should handle older titles and esports games comfortably, given its FP32 throughput of 5.323 TFLOPS and 512 shading units.

At 1440p, the picture becomes more nuanced. The system-dependent memory bandwidth means performance will scale with the host system's RAM speed and configuration. Users with fast dual-channel memory will see better results than those with single-channel or slower modules. The RTX 4060 comparison, where the 760M trails by only 0.8%, suggests that for lighter loads, the 760M could substitute for a low-end discrete card, but the lack of dedicated VRAM will likely manifest in texture-heavy scenes. For 4K, the 16 ROPs and system-shared memory present a hard limitation; the pixel rate of 41.58 GPixel/s suggests fill-rate constraints at high resolutions.

Ray Tracing and Feature Set

The AMD Radeon 760M includes 8 ray tracing cores, which is notable for an integrated GPU. The presence of dedicated RT hardware means the chip can accelerate ray-traced effects, but the 15 W TDP and 512 shading units will limit the complexity of RT workloads. The data does not provide RT-specific benchmark scores, so conclusions must be drawn from the core counts alone. The FP16 throughput of 10.65 TFLOPS (2:1 ratio) indicates that the GPU can handle half-precision compute at double the rate of FP32, which is useful for AI workloads and some graphics effects.

API support is comprehensive: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The DirectX 12 Ultimate designation implies support for features like mesh shaders and variable rate shading, though the hardware's low power envelope will govern real-world utilization. The Vulkan 1.4 support aligns with the strong Vulkan benchmark score of 25742, suggesting the architecture is well-optimized for this API. No tensor cores are listed, so AI-accelerated features rely on the standard shader or RT cores.

Memory Subsystem

The memory subsystem is entirely system-shared, with no dedicated VRAM. The size, type, and bus width are all listed as "System Shared," and the bandwidth is "System Dependent." This is the single most important caveat for the 760M's performance. In the benchmark results, the average score of 22999 was achieved under specific system memory conditions, but users with slower RAM will see lower scores, and those with faster RAM could potentially push higher.

The lack of dedicated VRAM means the GPU competes with the CPU for memory bandwidth. For gaming, this translates to texture loading and frame buffer operations relying on system RAM speed and channel configuration. The 178 mm² die size and 25,390 million transistors suggest the memory controllers are optimized for efficiency, but they cannot overcome the fundamental physics of shared bandwidth. At high resolutions, where frame buffers grow and texture data increases, the system-dependent nature of the memory will become a bottleneck. The 16 ROPs and 32 TMUs help mitigate some of this by keeping pixel and texture throughput reasonable, but memory latency and bandwidth will ultimately cap performance.

FAQ

Q: How does the AMD Radeon 760M compare to the Intel Arc B580?

A: The 760M scores 22999 on average, which is 0.1% behind the Arc B580's 23021. The 22-point difference is negligible, meaning the integrated 760M effectively matches the discrete B580 in synthetic benchmarks.

Q: What is the TDP of the AMD Radeon 760M and what power connectors does it need?

A: The TDP is 15 W, and the power connectors are listed as "None." It is an IGP that draws power from the motherboard, requiring no external power connections and no dedicated PSU recommendation.

Q: Does the AMD Radeon 760M support ray tracing?

A: Yes, it includes 8 ray tracing cores, and it supports DirectX 12 Ultimate (12_2), which includes ray tracing features. The FP32 throughput of 5.323 TFLOPS will limit RT complexity.

Q: What memory does the AMD Radeon 760M use?

A: The memory is System Shared, meaning it uses the host system's RAM. The size, type, bus width are all system-shared, and bandwidth is system-dependent, so performance varies with the installed system memory.

Q: How does the AMD Radeon 760M perform in different APIs?

A: In Geekbench OpenCL, it scores 20255, while in Geekbench Vulkan, it scores 25742. The Vulkan score is about 27% higher, indicating strong Vulkan 1.4 optimization.

Q: Is the AMD Radeon 760M faster than the NVIDIA RTX 4060 Mobile?

A: Yes, the 760M has an average score of 22999, which is 1.2% higher than the RTX 4060 Mobile's 22729. The desktop RTX 4060, however, is 0.8% faster than the 760M.