AMD Radeon HD 8790M vs NVIDIA GeForce GT 555M Comparison

Head-to-Head Benchmarks

The recorded data shows only one direct head-to-head benchmark between these two mobile graphics processors, and it is a decisive victory for the NVIDIA GeForce GT 555M. In the Geekbench OpenCL test, the GT 555M scored 6,493 points against the AMD Radeon HD 8790M’s 5,017 points, a performance advantage of 29.4%. This is a substantial margin, placing the NVIDIA part clearly ahead in this compute-oriented workload. The AMD GPU’s OpenCL score is its weaker result; its Vulkan score of 6,365 is far closer to the NVIDIA part’s OpenCL figure, though no direct Vulkan comparison was recorded for the GT 555M. The database’s average benchmark score reflects this disparity: the GT 555M averages 6,493, while the HD 8790M averages 5,691, a gap of roughly 14%.

Context from the nearest rivals table reinforces the NVIDIA part’s standing. The GT 555M sits just 0.2% ahead of the NVIDIA Quadro M5000M (6,481) and 0.3% ahead of the AMD Radeon Vega 10 Mobile (6,476), while trailing the GeForce GTX 670M by only 0.3% (6,513) and the Intel UHD Graphics P750 by 0.9% (6,554). The GT 555M’s 37th percentile among all GPUs is therefore a credible mid-pack position. The HD 8790M, by contrast, ranks at the 33rd percentile. Its nearest rivals include the Intel Iris Pro Graphics P6300 (5,712, 0.4% ahead), the NVIDIA GeForce GTX 670MX (5,721, 0.5% ahead), and the NVIDIA GeForce GTX 550 Ti (5,731, 0.7% ahead), while the NVIDIA Quadro M500M trails it by 1.6% (5,604). The AMD part’s average score is thus competitive with those older desktop and mobile parts, but it is not in the same league as the GT 555M on the only directly comparable test.

Architecture Differences

The two GPUs come from different architectural generations and are built on different process nodes. The NVIDIA GeForce GT 555M uses the GF106 chip, which is based on the Fermi architecture. It is fabricated by TSMC on a 40 nm process. The AMD Radeon HD 8790M uses the Mars chip, which belongs to the GCN 1.0 architecture, and is also built by TSMC but on a smaller 28 nm node. This process shrink is significant: the AMD chip integrates 950 million transistors on a die size of 77 mm², yielding a transistor density of 12.3 million transistors per mm². The NVIDIA chip, in contrast, packs 1,170 million transistors onto a much larger 238 mm² die, giving a density of only 4.9 million transistors per mm². The smaller process node allows AMD to achieve higher density despite fewer total transistors.

Core configuration also differs markedly. The GT 555M has 144 shading units, 24 texture mapping units, and 16 raster output units. The HD 8790M has 384 shading units, 24 TMUs, and just 8 ROPs. The AMD part has more than double the shading units, but half the ROPs. Clock behavior separates the two as well: the GT 555M’s base and boost clocks are not recorded in the database, while the HD 8790M has a base clock of 850 MHz and a boost clock of 900 MHz. These clock figures, combined with the higher shading unit count, produce a large FP32 compute gap. The GT 555M delivers 339.8 GFLOPS, while the HD 8790M delivers 691.2 GFLOPS, more than double. Texture rate also favors AMD at 21.60 GTexel/s versus 14.16 GTexel/s, but the pixel rate is unusual: the GT 555M achieves 3.540 GPixel/s, while the HD 8790M achieves 7.200 GPixel/s, again more than double, despite having fewer ROPs, due to its higher clocks and architecture efficiency.

Memory subsystems are entirely different. The GT 555M uses 1024 MB of DDR3 memory on a 128-bit bus, yielding a bandwidth of 28.80 GB/s. The HD 8790M uses 2 GB of GDDR5 memory on the same 128-bit bus, which more than doubles bandwidth to 64.00 GB/s. The memory clock is also recorded differently: the GT 555M runs at 900 MHz with 1800 Mbps effective, while the HD 8790M runs at 1000 MHz with 4 Gbps effective. API support diverges as well. The GT 555M lists DirectX 12 (11_0), OpenGL 4.6, and no Vulkan support. The HD 8790M lists DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The AMD part therefore has a slightly newer DirectX feature level and explicit Vulkan support, while the NVIDIA part lacks a recorded Vulkan capability. Bus interfaces also differ: the GT 555M uses PCIe 2.0 x16, while the HD 8790M uses MXM-A (3.0), and the AMD card is documented as an MXM module with no power connectors required.

FAQ

Q: Which GPU has the higher OpenCL benchmark score?

A: The NVIDIA GeForce GT 555M scores 6,493 in Geekbench OpenCL, while the AMD Radeon HD 8790M scores 5,017, giving the NVIDIA part a 29.4% lead in that test.

Q: Does the AMD Radeon HD 8790M support Vulkan?

A: Yes, the database records Vulkan 1.2.170 support for the HD 8790M, and it has a Geekbench Vulkan score of 6,365. The GT 555M has no Vulkan version listed and no Vulkan benchmark score.

Q: How do their memory bandwidth figures compare?

A: The GT 555M has 28.80 GB/s of bandwidth from 1024 MB of DDR3 on a 128-bit bus, while the HD 8790M has 64.00 GB/s from 2 GB of GDDR5 on the same 128-bit bus. The AMD part’s bandwidth is more than double.

Q: Which GPU has more shading units?

A: The AMD Radeon HD 8790M has 384 shading units, compared to 144 on the NVIDIA GeForce GT 555M. This contributes to the AMD part’s higher FP32 compute of 691.2 GFLOPS versus 339.8 GFLOPS.

Q: What are the process nodes for these two chips?

A: The GT 555M, based on the Fermi architecture, is built on a 40 nm process. The HD 8790M, based on GCN 1.0, is built on a 28 nm process. Both are fabricated by TSMC.

Q: What is the average benchmark score for each GPU?

A: The GT 555M has an average benchmark score of 6,493, placing it at the 37th percentile of all GPUs. The HD 8790M has an average score of 5,691, placing it at the 33rd percentile.

Specification Differences

The following fields differ between the NVIDIA GeForce GT 555M and the AMD Radeon HD 8790M:

• Chip: GF106 (NVIDIA) versus Mars (AMD)
• Architecture: Fermi versus GCN 1.0
• Process Node: 40 nm versus 28 nm
• Transistors: 1,170 million versus 950 million
• Die Size: 238 mm² versus 77 mm²
• Transistor Density: 4.9M / mm² versus 12.3M / mm²
• Base Clock: Not recorded versus 850 MHz
• Boost Clock: Not recorded versus 900 MHz
• Memory Clock: 900 MHz / 1800 Mbps effective versus 1000 MHz / 4 Gbps effective
• Memory Size: 1024 MB versus 2 GB
• Memory Type: DDR3 versus GDDR5
• Bandwidth: 28.80 GB/s versus 64.00 GB/s
• Shading Units: 144 versus 384
• ROPs: 16 versus 8
• Pixel Rate: 3.540 GPixel/s versus 7.200 GPixel/s
• Texture Rate: 14.16 GTexel/s versus 21.60 GTexel/s
• FP32: 339.8 GFLOPS versus 691.2 GFLOPS
• TDP: 35 W versus not recorded
• Slot Width: Not recorded versus MXM Module
• Power Connectors: Not recorded versus None
• Bus Interface: PCIe 2.0 x16 versus MXM-A (3.0)
• DirectX: 12 (11_0) versus 12 (11_1)
• Vulkan: Not recorded versus 1.2.170
• Release Date: 2011-07-01 versus 2013-03-31
• Predecessor: GeForce 400M versus London
• Successor: GeForce 600M versus Gem System

Where Each One Wins

The NVIDIA GeForce GT 555M wins the only recorded head-to-head benchmark, the Geekbench OpenCL test, by 29.4%. Its average benchmark score is also higher at 6,493 versus 5,691. The GT 555M’s percentile ranking is better as well, at the 37th percentile versus the 33rd percentile. For any workload that relies on the specific OpenCL performance measured in the database, the GT 555M is the clear choice. Its higher ROP count of 16 versus 8 also suggests an advantage in pixel-bound scenarios, though this is not directly benchmarked.

The AMD Radeon HD 8790M wins on several architectural and feature fronts. It has more than double the shading units (384 versus 144), more than double the FP32 compute (691.2 GFLOPS versus 339.8 GFLOPS), and more than double the texture rate (21.60 GTexel/s versus 14.16 GTexel/s). Its pixel rate is also higher at 7.200 GPixel/s versus 3.540 GPixel/s. The HD 8790M has double the memory capacity (2 GB versus 1024 MB) and more than double the bandwidth (64.00 GB/s versus 28.80 GB/s), using faster GDDR5 memory. It also supports Vulkan 1.2.170, which the GT 555M does not, and its DirectX feature level is 11_1 versus 11_0. For compute-heavy tasks that leverage the GCN architecture’s higher shader throughput, the AMD part is superior.

The Verdict

The data points to a split decision based on workload type. The NVIDIA GeForce GT 555M is the winner in the only direct comparison available, the Geekbench OpenCL test, where its 6,493 score beats the HD 8790M’s 5,017 by 29.4%. This is a substantial margin, and it is supported by the GT 555M’s higher average benchmark score (6,493 versus 5,691) and higher percentile rank (37th versus 33rd). Buyers or system integrators prioritizing that specific compute workload should select the GT 555M.

However, the AMD Radeon HD 8790M offers a more modern feature set and superior raw specifications in several categories. Its 28 nm process node, 384 shading units, 691.2 GFLOPS of FP32 compute, and 64.00 GB/s of memory bandwidth all exceed the GT 555M’s corresponding figures. The HD 8790M also adds Vulkan support and a newer DirectX feature level, making it more future-proof in API compatibility. The GT 555M does have a lower TDP at 35 W, while the HD 8790M’s TDP is not recorded, so power efficiency cannot be directly compared. For users who value shader throughput, memory capacity, and modern API support over the single recorded benchmark result, the HD 8790M is the better option. The final choice hinges on whether the measured OpenCL advantage or the architectural feature advantage matters more for the intended use case.