AMD Instinct MI350X vs Intel Arc Graphics 48EU Mobile Comparison

Head-to-Head Benchmarks

The recorded data contains no direct head-to-head benchmark comparisons for the AMD Instinct MI350X and the Intel Arc Graphics 48EU Mobile. The head-to-head benchmark list is empty, with zero recorded wins for either product. This absence of comparative measurements means the two products cannot be ranked against each other through direct performance testing in the database. The percentile rankings for both products stand at 50, indicating they occupy the median position within the database's overall GPU performance distribution, though this figure does not reflect any direct comparison between them.

The AMD Instinct MI350X delivers a peak FP32 throughput of 72.09 TFLOPS, while the Intel Arc Graphics 48EU Mobile delivers 1,382.4 GFLOPS in FP32. This represents a substantial gap in raw compute capability, with the AMD part exceeding the Intel part by a factor of roughly 52 in single-precision floating-point operations. The texture rate differential is similarly pronounced: the MI350X achieves 2,252.8 GTexel/s versus the Arc's 43.20 GTexel/s, a difference of more than 50-fold. Pixel rate tells a different story, however, as the MI350X records 0 MPixel/s while the Arc manages 14.40 GPixel/s, a consequence of the AMD accelerator having no ROPs and no display outputs.

Memory bandwidth separates the two decisively. The MI350X uses 288 GB of HBM3e across an 8192-bit bus, delivering 8.19 TB/s. The Arc Graphics 48EU Mobile relies on system shared memory with bandwidth described as "System Dependent." The AMD memory subsystem operates at 2000 MHz with 8 Gbps effective data rate, while the Intel part uses system memory without a dedicated bandwidth figure. Clock speeds show the MI350X running at a 1000 MHz base and 2200 MHz boost, compared to the Arc's 300 MHz base and 1800 MHz boost, though these figures reflect different design philosophies for different workloads.

FAQ

Q: What is the primary architectural difference between the two products?

A: The AMD Instinct MI350X uses the CDNA 4.0 architecture with 16,384 shading units and 1,024 texture mapping units, while the Intel Arc Graphics 48EU Mobile uses the Xe-LPG architecture with 384 shading units and 24 texture mapping units. The AMD part has no ROPs, while the Intel part has 8 ROPs.

Q: Which product has higher memory bandwidth?

A: The MI350X delivers 8.19 TB/s of bandwidth through 288 GB of HBM3e memory on an 8192-bit bus. The Arc Graphics 48EU Mobile uses system shared memory with bandwidth listed as "System Dependent" and no dedicated bus width.

Q: What are the power consumption figures?

A: The MI350X has a TDP of 1000 W and a suggested PSU of 1400 W, whereas the Arc Graphics 48EU Mobile has a TDP of 28 W and no suggested PSU listed.

Q: Do both products support DirectX?

A: The Intel Arc Graphics 48EU Mobile supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The AMD Instinct MI350X lists all APIs as N/A, reflecting its design as an accelerator without display outputs.

Q: What are the physical form factors?

A: The MI350X is an OAM Module measuring 102 mm in length and 165 mm in width, with no display outputs. The Arc Graphics 48EU Mobile is an IGP with no dimensions listed and display outputs described as "Portable Device Dependent."

Q: Which product was released earlier?

A: The Intel Arc Graphics 48EU Mobile released on 2023-12-13, while the AMD Instinct MI350X released on 2025-06-11. The Intel part's predecessor is HD Graphics-M, and the AMD part's predecessor is Radeon Instinct.

Architecture Differences

The AMD Instinct MI350X is built on CDNA 4.0 architecture, fabricated on a 3 nm process at TSMC. The chip contains 185,000 million transistors on a die size of 2380 mm², yielding a transistor density of 77.7 million per mm². The Intel Arc Graphics 48EU Mobile uses Xe-LPG architecture on a 10 nm process at Intel, with no transistor count, die size, or density figures recorded. The node difference represents a significant generational gap in manufacturing technology.

Compute resources differ dramatically. The MI350X has 16,384 shading units, 1,024 TMUs, and 0 ROPs. Its FP32 throughput is 72.09 TFLOPS, and its FP16 throughput is also 72.09 TFLOPS at a 1:1 ratio. The Arc Graphics 48EU Mobile has 384 shading units, 24 TMUs, and 8 ROPs. Its FP32 throughput is 1,382.4 GFLOPS, while FP16 reaches 2.765 TFLOPS at a 2:1 ratio. The FP16 handling differs fundamentally: the MI350X treats both precisions equally, while the Arc uses a 2:1 ratio that doubles FP16 throughput relative to FP32.

The MI350X has no display outputs, no ROPs, and lists all graphics APIs as N/A, indicating it is not designed for rendering to screens. The Arc Graphics 48EU Mobile has display outputs described as "Portable Device Dependent," supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, and includes 8 ROPs for pixel processing. The MI350X connects via PCIe 5.0 x16, while the Arc uses a Ring Bus interface. Power delivery also differs: the MI350X has no power connectors and requires a 1400 W suggested PSU, while the Arc has no power connector listed and no PSU requirement.

Specification Differences

The two products differ across nearly every recorded specification field. Process node: 3 nm for AMD versus 10 nm for Intel. Foundry: TSMC versus Intel. Transistors: 185,000 million for AMD, none recorded for Intel. Die size: 2380 mm² for AMD, none for Intel. Transistor density: 77.7M per mm² for AMD, none for Intel.

Clock speeds: the MI350X runs at 1000 MHz base and 2200 MHz boost, while the Arc runs at 300 MHz base and 1800 MHz boost. The MI350X has a memory clock of 2000 MHz with 8 Gbps effective data rate, while the Arc's memory clock is listed as "System Shared." Memory configuration: the MI350X has 288 GB of HBM3e with an 8192-bit bus and 8.19 TB/s bandwidth, while the Arc has "System Shared" memory, type, bus width, and bandwidth described as "System Dependent."

Shading units: 16,384 versus 384. TMUs: 1,024 versus 24. ROPs: 0 versus 8. Pixel rate: 0 MPixel/s versus 14.40 GPixel/s. Texture rate: 2,252.8 GTexel/s versus 43.20 GTexel/s. FP32: 72.09 TFLOPS versus 1,382.4 GFLOPS. FP16: 72.09 TFLOPS (1:1) versus 2.765 TFLOPS (2:1). TDP: 1000 W versus 28 W. Slot width: OAM Module versus IGP. Power connectors: None versus none recorded. Suggested PSU: 1400 W versus none. Bus interface: PCIe 5.0 x16 versus Ring Bus. Display outputs: No outputs versus "Portable Device Dependent." APIs: N/A versus DirectX 12 (12_1), OpenGL 4.6, Vulkan 1.4. Dimensions: 102 mm length and 165 mm width for AMD, none for Intel. Release date: 2025-06-11 versus 2023-12-13. Predecessor: Radeon Instinct versus HD Graphics-M.

Where Each One Wins

The AMD Instinct MI350X wins decisively in compute throughput, memory bandwidth, and memory capacity. Its FP32 performance of 72.09 TFLOPS and FP16 performance of 72.09 TFLOPS position it as a high-throughput accelerator for dense numerical workloads. The 288 GB HBM3e memory pool with 8.19 TB/s bandwidth provides massive data access capability, suitable for large models or datasets that exceed the memory capacity of typical graphics cards. The 1,024 TMUs and 2,252.8 GTexel/s texture rate indicate strong texel-processing capability for workloads that rely on texture sampling. The PCIe 5.0 x16 interface provides high-bandwidth host connectivity. The 3 nm process and 185,000 million transistors on a 2380 mm² die indicate a large, complex compute device.

The Intel Arc Graphics 48EU Mobile wins in areas related to rendering and integration. Its 8 ROPs and 14.40 GPixel/s pixel rate allow actual pixel output, which the MI350X cannot deliver at all (0 MPixel/s). The Arc supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, making it usable for graphics applications and games, whereas the MI350X lists all APIs as N/A. The Arc's display outputs, described as "Portable Device Dependent," enable connection to screens in mobile devices, while the MI350X has no outputs. The 28 W TDP makes the Arc suitable for power-constrained mobile environments, compared to the MI350X's 1000 W TDP and 1400 W suggested PSU. The Arc's IGP form factor and Ring Bus interface allow integration into a processor package, with no separate card or module needed. Its 2:1 FP16 ratio of 2.765 TFLOPS shows some accelerated half-precision capability relative to its FP32 throughput.

The Verdict

The data shows two products engineered for entirely different purposes. The AMD Instinct MI350X is a compute accelerator with no rendering capability: zero ROPs, zero pixel rate, no display outputs, and no graphics API support. Its strengths are raw compute (72.09 TFLOPS FP32), massive memory (288 GB HBM3e at 8.19 TB/s), and high clock speeds (2200 MHz boost) on a 3 nm TSMC process. It targets workloads that need large memory capacity and sustained floating-point throughput, such as AI training or scientific simulation, where pixel output and API compatibility are irrelevant.

The Intel Arc Graphics 48EU Mobile is an integrated GPU for portable devices. It provides actual graphics functionality: 8 ROPs, 14.40 GPixel/s pixel rate, DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4 support, plus display outputs. Its 28 W TDP fits mobile power budgets, and its IGP form factor eliminates the need for discrete card installation. The 10 nm Intel process and 384 shading units reflect a design optimized for efficiency and basic rendering, not extreme compute.

Users requiring a rendering device for a laptop or portable system should select the Intel Arc Graphics 48EU Mobile, as it supports display output and graphics APIs. Users requiring an accelerator for memory-intensive compute with no display needs should select the AMD Instinct MI350X, as it delivers substantially higher FP32 throughput, memory bandwidth, and memory capacity. The release dates reinforce this split: the Arc launched in 2023 as a mobile graphics solution, while the MI350X launched in 2025 as a data-center-scale compute module. Neither product can substitute for the other in its intended role.