AMD Radeon 760M vs Intel Arc A380E Comparison

Where Each One Wins

The AMD Radeon 760M and Intel Arc A380E occupy distinctly different performance tiers, and the benchmark data reflects this clearly. The Radeon 760M is an integrated graphics processor (IGP) built on the Phoenix chip, designed to share system memory and operate within a 15 W power envelope. The Arc A380E is a discrete, single-slot card with 6 GB of dedicated GDDR6 memory and a 75 W power draw. These fundamental positioning differences dictate where each part shows its strengths.

The Radeon 760M delivers its wins in compute-oriented workloads. Its recorded Geekbench OpenCL score of 20,255 and Geekbench Vulkan score of 30,336 are substantial figures for an integrated solution. The PassMark GPU Compute score of 2,840 further indicates that the 760M handles general-purpose compute tasks competently. The 760M also shows balanced DirectX performance across legacy and modern APIs, with PassMark scores of 65 in DirectX 9, 52 in DirectX 11, 25 in DirectX 12, and 19 in DirectX 10. Its 3DMark Steel Nomad DX12 score of 400 rounds out the picture of a part that can engage with modern graphics workloads despite its integrated nature.

The Arc A380E, by contrast, has no recorded benchmark scores in the database. Its wins are structural rather than measured. It offers a dedicated 6 GB memory pool with 186.0 GB/s of bandwidth over a 96-bit bus, which is a categorical advantage over the 760M's system-shared memory with system-dependent bandwidth. The A380E also doubles the shading units, TMUs, and ROPs of the 760M: 1,024 shading units versus 512, 64 TMUs versus 32, and 32 ROPs versus 16. The pixel rate of 64.00 GPixel/s and texture rate of 128.0 GTexel/s both exceed the 760M's 41.58 GPixel/s and 83.17 GTexel/s. In raw throughput terms, the A380E is the stronger graphics processor on paper, but without benchmark scores in the database, its measured wins cannot be quantified.

Architecture Differences

The two GPUs come from different architectural lineages and manufacturing processes. The Radeon 760M uses AMD's RDNA 3.0 architecture on the Phoenix chip, fabricated on a 4 nm process at TSMC. The die contains 25,390 million transistors across 178 mm², yielding a transistor density of 142.6 million per mm². The Arc A380E uses Intel's Xe-HPG architecture on the DG2-128 chip, fabricated on a 6 nm process at TSMC. Its die has 7,200 million transistors across 157 mm², for a density of 45.9 million per mm². The 760M is the denser chip by a wide margin, reflecting the newer process node.

Clock behavior diverges sharply. The 760M has an 800 MHz base clock and a 2,599 MHz boost clock, a wide dynamic range typical of an IGP that must scale with thermal and power headroom. The A380E runs at a fixed 2,000 MHz for both base and boost, a locked clock strategy suited to a discrete card with a stable power budget. Memory also differs categorically: the 760M uses system-shared memory with system-dependent bandwidth, while the A380E has 6 GB of GDDR6 at 1,937 MHz with 15.5 Gbps effective transfer, delivering 186.0 GB/s on a 96-bit bus.

Compute resources show a clear split. The A380E has 1,024 shading units, 64 TMUs, and 32 ROPs, exactly double each of the 760M's 512 shading units, 32 TMUs, and 16 ROPs. Both have 8 ray tracing cores. Floating-point throughput tells a more nuanced story: the 760M reaches 5.323 TFLOPS FP32 and 5.323 TFLOPS FP16 at a 1:1 ratio, while the A380E reaches 4.096 TFLOPS FP32 and 8.192 TFLOPS FP16 at a 2:1 ratio. The 760M leads in FP32, the A380E leads in FP16, and the A380E's FP16 advantage is exactly double its FP32 rate. The 760M's FP16 and FP32 rates are identical, indicating a different execution path for half-precision work.

The physical and interface profiles differ as well. The 760M is an IGP with no slot width, no power connectors, and motherboard-dependent display outputs, using a PCIe 4.0 x8 interface. The A380E is a single-slot card measuring 254 mm by 127 mm by 20 mm, has no power connectors, lists a 250 W suggested PSU, and provides four DisplayPort 2.0 outputs, also on PCIe 4.0 x8. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The 760M is marked Active in production status with a release date of 2024-01-30, while the A380E is End-of-life with a release date of 2024-03-31 and a listed successor of Battlemage.

Head-to-Head Benchmarks

The database contains no direct head-to-head benchmark entries between the Radeon 760M and the Arc A380E, and the A380E has no individual benchmark scores recorded at all. The comparison must therefore rest on the 760M's absolute scores and the A380E's architectural specifications.

The Radeon 760M's strongest recorded result is the Geekbench Vulkan score of 30,336, which stands well above its OpenCL score of 20,255. This gap of roughly 10,000 points suggests that the 760M's RDNA 3.0 architecture is particularly responsive to Vulkan's execution model. The PassMark G3D score of 5,310 provides a second independent measure of general 3D capability, while the PassMark G2D score of 890 covers 2D workloads. The 3DMark Steel Nomad DX12 score of 400 is the lowest absolute number among the modern-API tests, but Steel Nomad is a demanding workload and the score should be read relative to the 760M's integrated-class positioning.

The 760M's nearest rivals in the database place it in a tight cluster. The AMD Radeon RX 6400 sits 0.3% above the 760M with an average score of 6,001 versus 6,019. The NVIDIA GeForce GTX 770M also sits 0.3% above with 6,000. The NVIDIA RTX PRO 6000 Blackwell Server is 0.4% above at 5,996, and the NVIDIA Quadro P2000 is 0.5% below at 6,049. These deltas are all within one percentage point, meaning the 760M's average benchmark score of 6,019 places it in a statistically tight performance band. The 760M's percentile ranking of 35 across all GPUs indicates that it outperforms roughly a third of the database's tracked graphics processors.

The A380E's architectural specifications suggest where its measured performance would land if scores were recorded. Its FP32 throughput of 4.096 TFLOPS trails the 760M's 5.323 TFLOPS by roughly 23%. Its FP16 throughput of 8.192 TFLOPS leads the 760M's 5.323 TFLOPS by roughly 54%. Its pixel rate of 64.00 GPixel/s leads the 760M's 41.58 GPixel/s by roughly 54%, and its texture rate of 128.0 GTexel/s leads the 760M's 83.17 GTexel/s by roughly 54%. The consistent 54% lead in pixel and texture throughput reflects the doubled ROP and TMU counts. The A380E's memory bandwidth of 186.0 GB/s is a fixed, dedicated resource, whereas the 760M's bandwidth is system-dependent and cannot be stated as a fixed number.

The Verdict

The data indicates two different design philosophies. The Radeon 760M is a high-density, low-power integrated GPU that punches above its class in compute benchmarks. Its 5.323 TFLOPS FP32, 35th percentile ranking, and average benchmark score of 6,019 place it in the same performance band as the RX 6400, GTX 770M, RTX PRO 6000 Blackwell Server, and Quadro P2000, all within 0.5% of each other. The 760M's 1:1 FP16 ratio and strong Vulkan score suggest a balanced compute architecture that does not sacrifice half-precision performance.

The Arc A380E is a discrete card with double the shading units, TMUs, and ROPs, a fixed 2,000 MHz clock, and dedicated 6 GB GDDR6 memory. Its architectural limits are clear: lower FP32 than the 760M, but higher FP16, pixel rate, and texture rate. The 2:1 FP16 ratio indicates that Intel prioritized half-precision throughput in the Xe-HPG design. With no recorded benchmarks, the A380E's practical performance cannot be verified from the database, but its 50th percentile ranking across all GPUs and end-of-life production status frame it as a part that is no longer in active development.

For compute-heavy workloads that favor FP32 and Vulkan, the recorded data favors the Radeon 760M. For workloads that require dedicated memory bandwidth, fixed clocks, and higher pixel and texture throughput, the Arc A380E's specifications point in its direction. The 760M is the only part with measured results, so any comparison involving the A380E must be treated as specification-based rather than benchmark-based. The 760M's active production status and higher transistor density also suggest a more modern design. The A380E's end-of-life status and successor designation of Battlemage indicate that Intel has moved on.

FAQ

Q: Which GPU has more shading units?

A: The Intel Arc A380E has 1,024 shading units, exactly double the Radeon 760M's 512.

Q: What is the Radeon 760M's average benchmark score?

A: The recorded average benchmark score is 6,019, placing it 0.3% above the AMD Radeon RX 6400's average of 6,001.

Q: How does FP16 performance differ between the two?

A: The Arc A380E delivers 8.192 TFLOPS FP16 at a 2:1 ratio, while the Radeon 760M delivers 5.323 TFLOPS FP16 at a 1:1 ratio, meaning the A380E leads half-precision throughput by roughly 54%.

Q: What memory configuration does the Arc A380E use?

A: The A380E uses 6 GB of GDDR6 memory on a 96-bit bus with 186.0 GB/s bandwidth and an effective 15.5 Gbps transfer rate.

Q: What is the Radeon 760M's percentile ranking?

A: The 760M ranks in the 35th percentile across all GPUs in the database, with its nearest rival, the NVIDIA Quadro P2000, sitting 0.5% below its average score.

Q: Does the database have any benchmark scores for the Arc A380E?

A: The database records no benchmark scores for the A380E, so its performance must be assessed from its architectural specifications rather than measured results.