AMD Radeon HD 8790M vs NVIDIA GeForce GTX 770M Comparison

# Head-to-Head Benchmarks

The only direct comparison available in the data is Geekbench OpenCL, and it is a decisive victory for the NVIDIA GeForce GTX 770M. The GTX 770M scores 8,552 points, while the AMD Radeon HD 8790M scores 5,017 points. That yields a deltaPct of 70.5 percent in favor of the NVIDIA part. In practical terms, the GTX 770M completes OpenCL workloads more than 70 percent faster than the HD 8790M, a margin that would translate into significantly shorter render times or faster compute tasks in any application leveraging OpenCL.

The gap is not merely a function of one benchmark, either. The average benchmark score across all tested workloads tells a similar story. The GTX 770M averages 6,000 points, while the HD 8790M averages 5,691 points. That is a 5.4 percent advantage for the NVIDIA part in aggregate performance, though the single head-to-head result suggests the OpenCL-specific advantage is far larger than the average would imply. The average score includes different workload mixes, so the 70.5 percent OpenCL margin may overstate the overall performance gap, but it remains the clearest signal in the data.

Looking at percentile rankings, the GTX 770M sits at the 35th percentile of all GPUs, while the HD 8790M sits at the 33rd percentile. These are close overall positions, but the GTX 770M edges ahead. The GTX 770M also has a Geekbench Metal score of 3,448, a test the HD 8790M does not appear in the data for, while the HD 8790M has a Geekbench Vulkan score of 6,365, a test the GTX 770M does not appear in. These complementary benchmark results suggest each card has strengths in different API environments, though neither can be directly compared on those tests.

The Verdict

From the data, the NVIDIA GeForce GTX 770M is the stronger performer in the one workload where both cards were tested head-to-head. The 70.5 percent OpenCL advantage is substantial and unambiguous. For any user whose primary applications rely on OpenCL compute, the GTX 770M is the clear choice. The GTX 770M also holds a higher average benchmark score (6,000 vs 5,691) and a higher percentile rank (35th vs 33rd), reinforcing its overall edge.

However, the HD 8790M is not without merit. Its Vulkan score of 6,365 is noteworthy, and it comes with a DirectX 12 (11_1) API support level versus the GTX 770M's DirectX 12 (11_0). For users targeting applications that leverage Vulkan or the newer DirectX feature level, the AMD card may offer better compatibility. The HD 8790M also has a higher base clock (850 MHz vs 706 MHz) and boost clock (900 MHz vs 797 MHz), though these clock advantages do not translate into a win in the OpenCL benchmark.

The practical verdict: if compute performance in OpenCL is the priority, choose the GTX 770M. If API compatibility with Vulkan or DirectX 12 (11_1) matters more, the HD 8790M has a case, but its lower average score and lower percentile rank suggest it is generally the weaker card. The data does not support calling the HD 8790M a superior product on any measured metric except API feature level and clock speed.

Architecture Differences

The two GPUs come from different architectural families. The NVIDIA GeForce GTX 770M is built on the Kepler architecture, using the GK106 chip, while the AMD Radeon HD 8790M uses the GCN 1.0 architecture with the Mars chip. Both are manufactured on a 28 nm process at TSMC, but their transistor budgets and die sizes differ significantly.

The GTX 770M packs 2,540 million transistors into a 221 mm² die, yielding a transistor density of 11.5M per mm². The HD 8790M, in contrast, has 950 million transistors on a 77 mm² die, with a slightly higher transistor density of 12.3M per mm². The NVIDIA chip is nearly three times larger in transistor count and die area, which explains much of its performance advantage. The higher density of the AMD chip is interesting but does not compensate for the sheer scale of the GK106.

Clock speeds differ notably. The GTX 770M runs at a base of 706 MHz and boosts to 797 MHz, while the HD 8790M runs at 850 MHz base and 900 MHz boost. The AMD card's higher clocks are a direct consequence of its smaller, denser die, but the NVIDIA card's larger shader count and memory subsystem overcome the clock disadvantage.

The GTX 770M has 960 shading units, 80 texture mapping units, and 24 ROPs. The HD 8790M has 384 shading units, 24 TMUs, and 8 ROPs. These are large differences: the NVIDIA card has 2.5 times the shading units, 3.3 times the TMUs, and 3 times the ROPs of the AMD card. Pixel rate is 15.94 GPixel/s for the GTX 770M versus 7.20 GPixel/s for the HD 8790M, and texture rate is 63.76 GTexel/s versus 21.60 GTexel/s. FP32 compute is 1.530 TFLOPS versus 691.2 GFLOPS, a 2.2x advantage for NVIDIA.

API support differs as well. The GTX 770M supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The HD 8790M supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The AMD card has a slightly higher DirectX feature level, while the NVIDIA card has a marginally newer Vulkan version.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 770M has an average benchmark score of 6,000, compared to 5,691 for the AMD Radeon HD 8790M. The GTX 770M leads by 309 points, or about 5.4 percent.

Q: What is the performance difference in OpenCL?

A: In the Geekbench OpenCL test, the GTX 770M scores 8,552 against 5,017 for the HD 8790M. This is a 70.5 percent advantage for the NVIDIA card.

Q: Does the AMD card have any benchmark where it wins?

A: The data shows no head-to-head benchmark wins for the HD 8790M. The winsA field is 1 for the GTX 770M and 0 for the AMD card. However, the HD 8790M does have a Geekbench Vulkan score of 6,365, which is not directly comparable to any GTX 770M result.

Q: Which card has more memory?

A: The GTX 770M has 3 GB of GDDR5 memory on a 192-bit bus, while the HD 8790M has 2 GB of GDDR5 on a 128-bit bus. The GTX 770M's memory bandwidth is 96.19 GB/s versus 64.00 GB/s for the AMD card.

Q: What are the transistor counts of each chip?

A: The GTX 770M's GK106 chip contains 2,540 million transistors, while the HD 8790M's Mars chip contains 950 million transistors. The NVIDIA chip is built on a 221 mm² die, and the AMD chip on a 77 mm² die.

Q: How do their percentile rankings compare?

A: The GTX 770M is at the 35th percentile of all GPUs, while the HD 8790M is at the 33rd percentile. Both are in the lower-middle range of GPU performance.

Where Each One Wins

The NVIDIA GeForce GTX 770M wins in every measured head-to-head comparison. It leads in OpenCL by 70.5 percent, in average benchmark score by 5.4 percent, and in percentile ranking by 2 points. It also wins decisively on raw specifications: 960 shading units versus 384, 80 TMUs versus 24, 24 ROPs versus 8, 3 GB memory versus 2 GB, and 96.19 GB/s bandwidth versus 64.00 GB/s. Pixel rate, texture rate, and FP32 compute are all substantially higher on the GTX 770M. The GTX 770M is the card for OpenCL compute, high-resolution textures, and any workload that benefits from a wide memory bus and high shader count.

The AMD Radeon HD 8790M wins in a few specific areas, though none show up as benchmark victories. It has higher base and boost clocks (850/900 MHz vs 706/797 MHz), a higher transistor density (12.3M per mm² vs 11.5M per mm²), and a slightly higher DirectX feature level (12 (11_1) vs 12 (11_0)). Its Vulkan score of 6,365 is a standalone data point that suggests reasonable Vulkan performance, though no GTX 770M Vulkan score exists for comparison. The HD 8790M also has a lower transistor count and die size, which may imply lower power draw, though the TDP is not listed for the AMD card (the GTX 770M is rated at 75 W). For users constrained by motherboard space, both use MXM modules, but the HD 8790M uses the MXM-A (3.0) interface while the GTX 770M uses MXM-B (3.0).

Specification Differences

The two cards differ across nearly every specification category. The GTX 770M uses the Kepler architecture with a GK106 chip, while the HD 8790M uses GCN 1.0 with a Mars chip. Transistor count: 2,540 million versus 950 million. Die size: 221 mm² versus 77 mm². Transistor density: 11.5M per mm² versus 12.3M per mm². Base clock: 706 MHz versus 850 MHz. Boost clock: 797 MHz versus 900 MHz. Memory clock is essentially identical at 1002 MHz versus 1000 MHz, both with 4 Gbps effective.

Memory configuration differs: 3 GB versus 2 GB GDDR5, 192-bit versus 128-bit bus, and 96.19 GB/s versus 64.00 GB/s bandwidth. Shading units: 960 versus 384. TMUs: 80 versus 24. ROPs: 24 versus 8. Pixel rate: 15.94 GPixel/s versus 7.20 GPixel/s. Texture rate: 63.76 GTexel/s versus 21.60 GTexel/s. FP32: 1.530 TFLOPS versus 691.2 GFLOPS. TDP: 75 W for the GTX 770M, not listed for the HD 8790M.

Bus interface: MXM-B (3.0) versus MXM-A (3.0). Display outputs are portable device dependent for both. DirectX support: 12 (11_0) versus 12 (11_1). OpenGL: 4.6 for both. Vulkan: 1.2.175 versus 1.2.170. Release dates: 2013-05-29 for the GTX 770M versus 2013-03-31 for the HD 8790M. Predecessors: GeForce 600M versus London. Successors: GeForce 800M versus Gem System. Both are end-of-life products with no launch MSRP listed.