AMD Radeon 760M vs AMD Radeon HD 8790M Comparison

FAQ

Q: How does the AMD Radeon 760M compare to the AMD Radeon HD 8790M in overall average benchmark score?

A: The Radeon 760M has an average benchmark score of 6019, while the Radeon HD 8790M scores 5691. This places the 760M in the 35th percentile of all GPUs, slightly ahead of the HD 8790M's 33rd percentile.

Q: Which GPU wins in the Geekbench OpenCL test, and by how much?

A: The AMD Radeon 760M wins decisively with a score of 20255 versus 5017 for the HD 8790M, a delta of 303.7%. This is the largest performance gap between the two in any shared benchmark.

Q: What is the difference in their manufacturing process nodes?

A: The Radeon 760M is built on TSMC's 4 nm process, while the Radeon HD 8790M uses a 28 nm process from the same foundry. The 760M's node is significantly more advanced, though the HD 8790M was released in 2013.

Q: How do their memory configurations differ?

A: The Radeon 760M uses system-shared memory with bandwidth described as "System Dependent," whereas the HD 8790M has 2 GB of dedicated GDDR5 memory on a 128-bit bus with 64.00 GB/s bandwidth. The 760M's memory performance is tied to the host system.

Q: Which architecture does each GPU use, and what does that mean for feature support?

A: The 760M uses RDNA 3.0 architecture from the Navi III IGP (Phoenix) generation, while the HD 8790M uses GCN 1.0 from the Solar System (HD 8700M) generation. The 760M supports DirectX 12 Ultimate (12_2), whereas the HD 8790M only reaches DirectX 12 (11_1).

Q: What is the production status of each GPU?

A: The AMD Radeon 760M is listed as "Active" production, released on 2024-01-30, while the AMD Radeon HD 8790M is "End-of-life," having been released on 2013-03-31.

The Verdict

The data presents a clear generational divide. The AMD Radeon 760M outperforms the AMD Radeon HD 8790M in both shared benchmark tests, with wins in Geekbench OpenCL (20255 vs 5017) and Geekbench Vulkan (30336 vs 6365). The 760M's average benchmark score of 6019 also exceeds the HD 8790M's 5691.

For users choosing between these two, the decision hinges on context. The Radeon 760M is an integrated GPU (IGP) with a 15 W TDP, designed for modern laptops where it shares system memory. It offers modern API support including DirectX 12 Ultimate and Vulkan 1.4. The HD 8790M is a dedicated MXM module with 2 GB of GDDR5 memory, but its GCN 1.0 architecture is dated, and its production status is end-of-life.

Benchmark results indicate the 760M is the superior performer in raw compute and graphics workloads. However, the HD 8790M's dedicated memory might be preferable in scenarios where system memory bandwidth is a bottleneck for the 760M. The HD 8790M sits in the 33rd percentile of all GPUs, while the 760M sits in the 35th, so both are near the lower-middle range of the GPU spectrum.

Head-to-Head Benchmarks

The head-to-head comparison includes only two shared tests, and the AMD Radeon 760M wins both. The most dramatic result is in Geekbench Vulkan, where the 760M scores 30336 against the HD 8790M's 6365, a delta of 376.6%. This is a substantial lead that indicates the 760M handles modern graphics APIs far more efficiently than the older GCN 1.0 part.

In Geekbench OpenCL, the 760M again dominates with 20255 points versus 5017 points, a 303.7% advantage. This test reflects general-purpose compute performance, and the 760M's RDNA 3.0 architecture with 512 shading units clearly outperforms the HD 8790M's 384 shading units.

Beyond these shared tests, the 760M has additional benchmark data that the HD 8790M lacks. The 760M scores 5310 in Passmark G3D, 2840 in Passmark GPU Compute, and 890 in Passmark G2D. It also shows API-specific results: 65 in DirectX 9, 52 in DirectX 11, 25 in DirectX 12, and 19 in DirectX 10. The HD 8790M has no comparable Passmark entries in the data, so these results are not directly comparable, but they do show the 760M's broader testing profile.

The 760M's average benchmark score of 6019 is within 0.3% of the AMD Radeon RX 6400 (6001) and within 0.5% of the NVIDIA Quadro P2000 (6049). The HD 8790M's average of 5691 is 0.4% behind the Intel Iris Pro Graphics P6300 (5712) and 0.5% behind the NVIDIA GeForce GTX 670MX (5721). These nearest-rival comparisons place the 760M in slightly better company, with its closest competitors being more modern parts.

Specification Differences

The two GPUs differ across nearly every specification category. The Radeon 760M uses a Phoenix chip with a 4 nm TSMC process, while the HD 8790M uses a Mars chip with a 28 nm TSMC process. Transistor counts reflect this: the 760M has 25,390 million transistors on a 178 mm² die, whereas the HD 8790M has 950 million transistors on a 77 mm² die. Transistor density is dramatically different at 142.6M per mm² for the 760M versus 12.3M per mm² for the HD 8790M.

Clock speeds show the 760M has a base of 800 MHz and a boost of 2599 MHz, while the HD 8790M has a base of 850 MHz and a boost of 900 MHz. Memory clocks also differ: the 760M uses system-shared memory, while the HD 8790M uses 1000 MHz (4 Gbps effective) GDDR5. The HD 8790M has 2 GB of dedicated memory on a 128-bit bus with 64.00 GB/s bandwidth; the 760M's memory is system-shared with system-dependent bandwidth.

Compute resources favor the 760M: it has 512 shading units, 32 TMUs, 16 ROPs, and 8 ray-tracing cores, compared to the HD 8790M's 384 shading units, 24 TMUs, and 8 ROPs. The 760M has no tensor cores, and neither does the HD 8790M. Pixel and texture rates are higher on the 760M at 41.58 GPixel/s and 83.17 GTexel/s, versus 7.200 GPixel/s and 21.60 GTexel/s for the HD 8790M. FP32 compute is 5.323 TFLOPS for the 760M versus 691.2 GFLOPS for the HD 8790M.

Power and interface specs differ as well. The 760M has a 15 W TDP and is an IGP with a PCIe 4.0 x8 interface, while the HD 8790M is an MXM Module with an MXM-A (3.0) interface. The 760M's display outputs are motherboard-dependent; the HD 8790M's are portable-device-dependent. Neither has power connectors.

Architecture Differences

The architectural gap between these two GPUs is significant. The Radeon 760M is built on RDNA 3.0, a modern architecture from the Navi III IGP (Phoenix) generation, whereas the HD 8790M uses GCN 1.0 from the Solar System (HD 8700M) generation. This represents roughly a decade of architectural evolution, from 2013 to 2024.

The 760M's RDNA 3.0 architecture includes 8 ray-tracing cores, a feature entirely absent from the HD 8790M. This enables hardware-accelerated ray tracing on the 760M, which the older GCN 1.0 part cannot do. The 760M also supports DirectX 12 Ultimate (12_2), which includes features like variable rate shading and mesh shaders; the HD 8790M is limited to DirectX 12 (11_1). Vulkan support is also newer on the 760M (1.4) versus the HD 8790M (1.2.170).

The process node difference is stark: 4 nm versus 28 nm. This is reflected in transistor density, where the 760M packs 142.6M transistors per mm² compared to the HD 8790M's 12.3M per mm². The 760M's 25,390 million transistors dwarf the HD 8790M's 950 million, enabling far more compute units and features within a similar power envelope. The 760M's 15 W TDP is notably low despite its higher performance, a direct benefit of the advanced node.

The 760M supports FP16 compute at a 1:1 ratio, with 5.323 TFLOPS for both FP32 and FP16. The HD 8790M has no listed FP16 performance, only 691.2 GFLOPS FP32. This makes the 760M more versatile for workloads that can use reduced precision. The 760M's predecessor is listed as Navi II IGP, while the HD 8790M's predecessor is London and its successor is Gem System.

Where Each One Wins

The AMD Radeon 760M wins in every shared benchmark category, so the use-case split is largely about suitability rather than direct performance. The 760M is the clear choice for modern gaming and compute workloads that leverage DirectX 12 Ultimate features, ray tracing, or Vulkan 1.4. Its 8 ray-tracing cores and 5.323 TFLOPS FP32 performance make it capable of handling contemporary titles at playable settings, especially given its 15 W TDP fits in thin-and-light laptops.

The HD 8790M's advantage lies in its dedicated 2 GB GDDR5 memory with 64.00 GB/s bandwidth. In systems where the 760M's shared memory would compete with the CPU for bandwidth, the HD 8790M's dedicated pool could provide more consistent memory performance. This makes the HD 8790M a potential choice for older legacy applications that rely on fixed-function hardware or for systems where the host system's memory bandwidth is limited.

However, the HD 8790M's end-of-life status and lack of modern API support (capped at DirectX 12 11_1) limit its future-proofing. The 760M's active production status and support for DirectX 12 Ultimate and Vulkan 1.4 make it the better long-term investment for any new system. For users with existing MXM-equipped laptops, the HD 8790M might be a drop-in replacement, but benchmark results show it is 303.7% behind in OpenCL and 376.6% behind in Vulkan, so any upgrade path to the 760M would yield massive gains.

In summary, the 760M is the winner for nearly all use cases, with its only potential drawback being shared memory. The HD 8790M retains relevance only in niche scenarios where dedicated memory is essential and modern API support is unnecessary.