AMD Radeon Vega 3 vs Intel UHD Graphics 710 Comparison

# AMD Radeon Vega 3 vs Intel UHD Graphics 710: A Benchmark Analysis

The database comparison between the AMD Radeon Vega 3 and the Intel UHD Graphics 710 reveals two integrated graphics solutions with notably different performance profiles. The AMD part, built on the older GCN 5.0 architecture, posts a higher average benchmark score, while the Intel part, based on the more recent Generation 12.2 architecture, counters with a strong win in one of the two direct head-to-head tests. Both are end-of-life products, with the AMD launching in late 2019 and the Intel part following in early 2022. The data shows a clear split: AMD wins in raw compute throughput, while Intel takes the lead in a specific API workload.

FAQ

Q: What are the average benchmark scores for these two GPUs?

A: The AMD Radeon Vega 3 averages a score of 4268 across its three recorded benchmarks. The Intel UHD Graphics 710 averages 3792 across its two recorded benchmarks.

Q: How do they compare in the OpenCL benchmark?

A: In the geekbench_opencl test, the AMD Radeon Vega 3 scores 3963, which is 13.4% higher than the Intel UHD 710's 3496.

Q: Which GPU wins the Vulkan benchmark?

A: The Intel UHD Graphics 710 wins the geekbench_vulkan test, scoring 4088 against the AMD's 3961, a margin of 3.1%.

Q: What is the performance percentile of each part?

A: The AMD Radeon Vega 3 sits in the 25th percentile of all GPUs, meaning it outperforms a quarter of the database. The Intel UHD 710 sits in the 22nd percentile.

Q: What are the closest rival scores?

A: For AMD, the closest rival is the NVIDIA GeForce GTX 460M with an average score of 4282, which is 0.3% higher. For Intel, the NVIDIA GeForce MX110 is closest with an average score of 3834, which is 1.1% higher.

Architecture Differences

The AMD Radeon Vega 3 uses the GCN 5.0 architecture built on a 12 nm process node at GlobalFoundries, a notable difference from the Intel part. The AMD chip integrates 4,940 million transistors on a 210 mm² die, resulting in a transistor density of 23.5 million per mm². Intel's UHD Graphics 710 is built on the Generation 12.2 architecture using a 10 nm process at Intel fabs. The die size and transistor count are not published for the Intel part.

The execution resources differ substantially. The AMD part carries 192 shading units, 12 texture mapping units, and 4 ROPs. The Intel part has 128 shading units, 8 texture mapping units, and 8 ROPs. This gives the AMD part a 50% advantage in shading unit count and a 50% advantage in both TMUs and ROPs, which directly feeds into its higher peak pixel and texture rates.

Clock behavior also separates the two. The AMD Radeon Vega 3 has a base clock of 300 MHz and a boost clock of 1100 MHz. Intel UHD 710 has a base of 300 MHz and a higher boost of 1300 MHz. Both parts draw memory from the system, with system shared memory and bandwidth, and both connect over an integrated graphics interface. The AMD part uses a ring bus interface.

API support is identical in coverage: both list DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. The AMD part also reports a pixel rate of 4.400 GPixel/s and a texture rate of 13.20 GTexel/s. Intel's figures are 10.40 GPixel/s and 10.40 GTexel/s. These differences are a direct result of the higher core counts and clock speeds of the AMD design.

Head-to-Head Benchmarks

The head-to-head results show a clear split based on the API. In the geekbench OpenCL test, the AMD Radeon Vega 3 takes a decisive lead, scoring 3963 versus Intel's 3496. That is a 13.4% advantage for AMD. The Vulkan test goes the other way: Intel UHD 710 wins with 4088 against AMD's 3961, a 3.1% margin. The two GPUs are therefore trading wins, each taking one test.

The OpenCL result is the larger margin by far and shows the AMD's compute architecture at a clear advantage in general-purpose compute workloads. The Vulkan test, while still a win for Intel, is a much smaller gap, suggesting that in real-world graphics workloads, the two parts are comparable, with Intel having a slight edge.

The average score of the AMD part, 4268, is 11.9% higher than the Intel average of 3792. This is also reflected in the percentile positions: the AMD at the 25th percentile, the Intel at the 22nd percentile. The results show the AMD architecture offers a higher raw compute throughput in the database, which is particularly visible in the OpenCL test.

Specification Differences

| Field | AMD Radeon Vega 3 | Intel UHD 710 |

|---|---|---|

| Architecture | GCN 5.0 | Generation 12.2 |

| Process node | 12 nm | 10 nm |

| Foundry | GlobalFoundries | Intel |

| Transistors | 4,940 million | Not published |

| Die size | 210 mm² | Not published |

| Transistor density | 23.5M / mm² | Not published |

| Base clock | 300 MHz | 300 MHz |

| Boost clock | 1100 MHz | 1300 MHz |

| Shading units | 192 | 128 |

| TMUs | 12 | 8 |

| ROPs | 4 | 8 |

| Pixel rate | 4.400 GPixel/s | 10.40 GPixel/s |

| Texture rate | 13.20 GTexel/s | 10.40 GTexel/s |

| FP32 | 422.4 GFLOPS | 332.8 GFLOPS |

| FP16 | 844.8 GFLOPS (2:1) | 665.6 GFLOPS (2:1) |

| TDP | 15 W | 15 W |

The two parts show a generational shift. The Intel chip is built on a smaller 10nm node and has a higher boost clock, but the AMD chip has 50% more of the shader cores, texture units, and a much higher memory performance. The AMD's 15W TDP matches Intel's 15W TDP, but the AMD has a higher transistor count and die area, suggesting more silicon dedicated to the compute work.

Where Each One Wins

AMD Radeon Vega 3 wins in OpenCL compute workloads. The strongest single test result comes from the geekbench opencl test, where AMD leads by 13.4%. This is a pure general-purpose compute test that uses the shader cores and FP32 units. The AMD's 192 shader units at a 1100 MHz boost clock provide a raw FP32 throughput of 422.4 GFLOPS, versus 332.8 GFLOPS for the Intel. This is a direct performance advantage for any workload that uses general compute, such as productivity applications, encoding, and some game physics.

Intel UHD 710 wins in the Vulkan graphics workload. The geekbench_vulkan test is an Intel win, scoring 4088 vs 3961. Although the margin is only 3.1%, it is still a win. Vulkan is a low-level, modern graphics API, and the Intel result indicates that the Intel part is better at handling this particular API's characteristics, which can be a lower driver overhead or a better geometry pipeline in this specific test. This is an important result for users who run Vulkan titles or applications, as the Intel part shows a measurable advantage here.

AMD Radeon Vega 3 wins in average performance. The average score of 4268 places it 11.9% above the Intel average of 3792. It is also the 25th percentile vs the 22nd percentile, so the AMD part is positioned for a slightly larger part of the overall GPU performance distribution. The AMD's larger shader core count and larger memory subsystem, combined with a much higher boost clock, are the reasons why it is a better choice for a general purpose compute and many graphics loads.

Intel UHD 710 wins in power efficiency parity. Both parts have the same 15W TDP, but the Intel is a smaller die with fewer transistors and a more efficient 10nm process. It delivers the Vulkan win with lower power draw in that test. At the same TDP, the Intel part achieves a win in a modern API, which suggests that in thermally constrained systems, it can hold its own or exceed the AMD in certain workloads while consuming the same power budget.

For users who need OpenCL compute performance, the AMD Radeon Vega 3 is the better choice. This is a clear, measured win. It is 13.4% faster in compute workloads, and its higher shader core count and shading rate, texture rate make it the better part for any workload that is not pure graphics.

For users who prioritize Vulkan graphics performance, the Intel UHD Graphics 710 is the better choice. This is the only head-to-head win for Intel. It is a smaller margin (3.1%), but it is still a win. The better Vulkan performance of the Intel chip makes it the better choice for Vulkan users, which is a growing segment of the game market.

Both parts have different strengths: AMD wins the compute performance, Intel wins the Vulkan graphics workload. The choice between them depends on the API and workload that matters most to you.