AMD Radeon Vega 8 vs Intel Arc A380 Comparison

AMD Radeon Vega 8 is an integrated graphics processor from AMD’s Raven Ridge generation, built on the 14 nm GCN 5.0 architecture. Intel Arc A380 is a discrete dual-slot graphics card from Intel’s Alchemist generation, built on the 6 nm Xe-HPG architecture. The data shows a clear performance hierarchy between the two, but the comparison is more nuanced than raw benchmark scores alone.

Head-to-Head Benchmarks

The head-to-head benchmark data is limited to two tests, and Intel Arc A380 wins both decisively. In Geekbench OpenCL, the Arc A380 scores 38224 against the Vega 8’s 8822, a delta of -76.9% from the Arc A380’s perspective. This means the Vega 8 trails by roughly three-quarters in raw compute throughput. The Geekbench Vulkan result is similar: Arc A380 scores 36736 while Vega 8 scores 8134, a delta of -77.9%. The Vega 8 is not competitive in these cross-API compute workloads.

However, the average benchmark scores tell a slightly different story. The Vega 8 has an average benchmark score of 9221, while the Arc A380 averages 8558. This is a reversal of the head-to-head results, and it stems from the different benchmark suites each product was tested with. The Vega 8’s average is pulled up by its Geekbench Metal score of 10706, a test the Arc A380 does not have a result for. The Arc A380’s average is dragged down by several Passmark scores, including a Passmark DirectX 10 score of 37, DirectX 11 score of 38, and DirectX 12 score of 35. These low DirectX scores are outliers in an otherwise strong profile, but they weigh heavily in the average.

Looking at the nearest rivals, the Vega 8 sits at the 45th percentile among all GPUs. Its nearest rival is the AMD Radeon 890M, which scores 9210 on average — a delta of just 0.1%. The NVIDIA GeForce GTX 960 is also close at 9273 (-0.6%), and the GTX 465 at 9294 (-0.8%). The Arc A380 sits at the 44th percentile, just one point lower. Its nearest rival is the AMD FirePro W5170M at 8595 (-0.4%), followed by the AMD Radeon HD 8870M at 8462 (1.1%), and the NVIDIA GeForce MX330 at 8458 (1.2%). In percentile terms, the two are virtually tied — 45 vs 44 — despite the massive head-to-head deltas.

Architecture Differences

The architectural gap between these two is generational. The Vega 8 uses GCN 5.0 on a 14 nm process from GlobalFoundries, while the Arc A380 uses Xe-HPG on a 6 nm process from TSMC. The transistor counts reflect this: the Vega 8 has 4,940 million transistors on a 210 mm² die, giving a density of 23.5M per mm². The Arc A380 has 7,200 million transistors on a smaller 157 mm² die, achieving 45.9M per mm² — nearly double the density.

The compute resources are also starkly different. Vega 8 has 512 shading units, 32 texture mapping units, and 8 ROPs. The Arc A380 doubles or quadruples these figures: 1024 shading units, 64 TMUs, and 32 ROPs. The Arc A380 also has 8 ray tracing cores, which the Vega 8 lacks entirely. In terms of raw throughput, the Vega 8 delivers 1,126.4 GFLOPS FP32 and 2.253 TFLOPS FP16 (2:1). The Arc A380 delivers 4.198 TFLOPS FP32 and 8.397 TFLOPS FP16 (2:1) — roughly 3.7x the FP32 throughput.

Memory is another fundamental divide. The Vega 8 uses system-shared memory with bandwidth described as "System Dependent." The Arc A380 has 6 GB of dedicated GDDR6 memory on a 96-bit bus, delivering 186.0 GB/s of bandwidth. That dedicated memory is a major advantage for the discrete card, as it does not compete with the CPU for system RAM. The Vega 8’s pixel rate is 8.800 GPixel/s and texture rate is 35.20 GTexel/s; the Arc A380’s are 65.60 GPixel/s and 131.2 GTexel/s, respectively.

API support differs as well. The Vega 8 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. The Arc A380 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The Arc A380’s DirectX 12 Ultimate support includes features like ray tracing and mesh shaders, which the Vega 8’s older 12_1 feature level cannot offer. Both support OpenGL 4.6, but the Vulkan version is newer on the Intel side.

The Verdict

The data is unambiguous on raw performance: the Intel Arc A380 is in a different class. Its compute scores are over 4x higher in Geekbench OpenCL and Vulkan, and its FP32 throughput is 3.7x higher. For any workload that uses compute shaders, ray tracing, or dedicated VRAM, the Arc A380 is the clear choice. The Vega 8’s single win is its average benchmark score of 9221 vs 8558, but that is a function of benchmark selection, not actual performance superiority. The Vega 8’s Geekbench Metal score of 10706 is its strongest result, but the Arc A380 has no Metal test to compare against.

For users who need a discrete GPU with dedicated memory, the Arc A380 is the only option between these two. It offers 6 GB of GDDR6, a 186.0 GB/s memory bandwidth, and a 250 W suggested PSU requirement. It carries a launch MSRP of 149 USD. The Vega 8 is an integrated part with no power connectors, no slot width, and motherboard-dependent display outputs. It consumes 25 W TDP, versus the Arc A380’s 75 W TDP.

The verdict depends on the use case. If the system requires a compact, low-power integrated solution with no add-in card, the Vega 8 is the only viable choice. If the system can accommodate a dual-slot, 222 mm long, 114 mm tall, 42 mm wide card with a 1x 8-pin power connector, the Arc A380 is overwhelmingly faster. The percentile rankings — 45th for Vega 8, 44th for Arc A380 — are misleading; they reflect different benchmark pools, not comparable performance.

Specification Differences

The two differ in nearly every specification. The process node is 14 nm for Vega 8 vs 6 nm for Arc A380. The foundry is GlobalFoundries for AMD, TSMC for Intel. Transistors: 4,940 million vs 7,200 million. Die size: 210 mm² vs 157 mm². Transistor density: 23.5M / mm² vs 45.9M / mm². Base clock: 300 MHz vs 2000 MHz. Boost clock: 1100 MHz vs 2050 MHz. Memory size: System Shared vs 6 GB. Memory type: System Shared vs GDDR6. Bus width: System Shared vs 96 bit. Bandwidth: System Dependent vs 186.0 GB/s.

Shading units: 512 vs 1024. TMUs: 32 vs 64. ROPs: 8 vs 32. Ray tracing cores: none vs 8. Pixel rate: 8.800 GPixel/s vs 65.60 GPixel/s. Texture rate: 35.20 GTexel/s vs 131.2 GTexel/s. FP32: 1,126.4 GFLOPS vs 4.198 TFLOPS. FP16: 2.253 TFLOPS (2:1) vs 8.397 TFLOPS (2:1). TDP: 25 W vs 75 W. Slot width: IGP vs Dual-slot. Power connectors: None vs 1x 8-pin. Suggested PSU: none vs 250 W. Bus interface: IGP vs PCIe 4.0 x8. Display outputs: Motherboard Dependent vs 1x HDMI 2.13x DisplayPort 2.0. DirectX: 12 (12_1) vs 12 Ultimate (12_2). Vulkan: 1.3 vs 1.4. Release date: 2018-02-11 vs 2022-06-13. Predecessor: GCN 3.0 IGP vs Xe Graphics. Successor: Vega II IGP vs Battlemage.

FAQ

Q: Which GPU has higher raw compute performance?

A: The Intel Arc A380. It scores 38224 in Geekbench OpenCL vs the Vega 8’s 8822, and 36736 in Geekbench Vulkan vs 8134. Its FP32 throughput is 4.198 TFLOPS vs 1,126.4 GFLOPS.

Q: Does the Vega 8 have any ray tracing capability?

A: No. The Vega 8 has zero ray tracing cores. The Arc A380 has 8 ray tracing cores.

Q: How does memory configuration differ?

A: The Vega 8 uses system-shared memory with bandwidth described as "System Dependent." The Arc A380 has 6 GB of dedicated GDDR6 memory on a 96-bit bus with 186.0 GB/s bandwidth.

Q: Why does the Vega 8 have a higher average benchmark score?

A: The Vega 8 averages 9221 vs the Arc A380’s 8558. This is because the Vega 8’s average includes a Geekbench Metal score of 10706, a test the Arc A380 lacks, while the Arc A380’s average includes several low Passmark DirectX scores (37, 38, 35).

Q: What are the power requirements for each?

A: The Vega 8 is an IGP with 25 W TDP and no power connectors. The Arc A380 is a dual-slot card with 75 W TDP, requires a 1x 8-pin power connector, and has a suggested PSU of 250 W.

Q: Which has better API support?

A: The Arc A380 supports DirectX 12 Ultimate (12_2) and Vulkan 1.4. The Vega 8 supports DirectX 12 (12_1) and Vulkan 1.3. Both support OpenGL 4.6.

Where Each One Wins

The Intel Arc A380 wins in every compute-heavy scenario. Its 4.198 TFLOPS FP32 and 8.397 TFLOPS FP16 make it suitable for GPU compute tasks, and its 8 ray tracing cores enable hardware-accelerated ray tracing, which the Vega 8 cannot do. The 6 GB GDDR6 memory with 186.0 GB/s bandwidth means textures and geometry can be stored locally, avoiding system memory bottlenecks. The Arc A380’s 65.60 GPixel/s pixel rate and 131.2 GTexel/s texture rate are 7.5x and 3.7x higher than the Vega 8’s, respectively. For gaming, DirectX 12 Ultimate support allows access to modern features like mesh shaders and variable rate shading.

The AMD Radeon Vega 8 wins in power efficiency and system simplicity. At 25 W TDP with no power connectors and no slot requirement, it integrates directly into a motherboard. It has no length, height, or width dimensions because it has no physical card. It is end-of-life and was released in 2018, while the Arc A380 was released in 2022 and is also end-of-life. The Vega 8’s Geekbench Metal score of 10706 suggests it performs well in Apple’s Metal API, though this is an outlier in its benchmark profile.

For a builder choosing between these, the deciding factor is whether the system can accept a discrete card. If it can — and the PSU supports 250 W — the Arc A380 is the only sensible pick. If the system is constrained to integrated graphics, the Vega 8 is the available option, but the data shows it is not competitive with even a low-end discrete card like the Arc A380. The percentile difference of 45 vs 44 is within noise; the actual benchmark deltas of -76.9% and -77.9% are not.