AMD Instinct MI308X vs Intel Arc B390 Comparison

Head-to-Head Benchmarks

The recorded data contains no direct head-to-head benchmark comparisons between the AMD Instinct MI308X and the Intel Arc B390. The two products occupy entirely different segments, and their performance profiles are defined by separate measurement categories. The AMD Instinct MI308X has no benchmark scores recorded in the database, leaving its percentile ranking at 50 among all GPUs. The Intel Arc B390, by contrast, has a single recorded 3DMark Steel Nomad DX12 score of 1482, placing it at the 9th percentile among all GPUs.

The Intel Arc B390's benchmark result places it in a tightly clustered group of rival products. It sits 1.3% ahead of the NVIDIA GeForce GT 520MX, which scores 1463. It is 1.5% ahead of the NVIDIA GeForce 800M at 1460, 2.5% ahead of the NVIDIA GeForce GT 625 OEM at 1446, and 2.7% ahead of the NVIDIA GeForce GT 710 at 1443. These deltas are narrow, indicating that the Arc B390 lands at the upper edge of a performance tier dominated by entry-level discrete and integrated graphics solutions. The data shows a product that barely edges past its nearest competitors in raw 3DMark performance, with the largest gap being a modest 2.7% over the GT 710.

The AMD Instinct MI308X presents a fundamentally different situation. With no recorded benchmarks, its percentile score of 50 is a neutral placement, not a performance assertion. The absence of benchmark data for the MI308X means the database cannot quantify its standing against the Arc B390 or any other GPU. Instead, the MI308X's performance characteristics must be derived from its architectural specifications, which describe a compute accelerator rather than a consumer graphics card.

Where Each One Wins

The AMD Instinct MI308X wins decisively in raw compute throughput. Its FP32 performance is recorded at 81.72 TFLOPS, and its FP16 performance is also 81.72 TFLOPS at a 1:1 ratio. The Intel Arc B390 delivers 7.680 TFLOPS in FP32 and 15.36 TFLOPS in FP16 at a 2:1 ratio. The FP32 gap is substantial: the MI308X provides more than ten times the single-precision throughput. In FP16, the MI308X still holds a large advantage, delivering 81.72 TFLOPS versus the Arc B390's 15.36 TFLOPS, a factor of roughly 5.3.

Memory capacity and bandwidth also favor the MI308X overwhelmingly. The MI308X carries 192 GB of HBM3 memory on an 8192-bit bus, yielding 5.32 TB/s of bandwidth. The Arc B390 uses system shared memory, with its size, bus width, and bandwidth all listed as system dependent. The MI308X's memory subsystem is a dedicated, high-bandwidth design intended for large-scale data movement, while the Arc B390 relies on the host system's memory pool, a configuration typical of integrated graphics.

Texture and pixel throughput tell a similar story. The MI308X achieves 2,553.6 GTexel/s, while the Arc B390 manages 120.0 GTexel/s. Pixel rate is a different matter: the MI308X records 0 MPixel/s, reflecting its lack of a traditional raster output stage, while the Arc B390 delivers 60.00 GPixel/s. This is the one area where the Arc B390 posts a nonzero result, and it aligns with the product's role as a graphics-capable processor with display outputs.

The Intel Arc B390 wins in areas tied to graphics feature support and practical deployment. It supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The MI308X lists N/A for DirectX, OpenGL, and Vulkan, indicating that it is not designed for consumer graphics APIs. The Arc B390 also has 24 ROPs and 12 ray tracing cores, while the MI308X records 0 ROPs and no ray tracing core count. The MI308X has no display outputs, while the Arc B390's display outputs are portable device dependent.

Architecture Differences

The two processors come from different manufacturers, fabs, and design philosophies. The AMD Instinct MI308X uses the Aqua Vanjaram chip, built on the CDNA 3.0 architecture, and is manufactured on a 5 nm process at TSMC. It contains 153,000 million transistors on a 1017 mm² die, yielding a transistor density of 150.4 million transistors per square millimeter. The Intel Arc B390 uses the Panther Lake chip, built on the Xe3-LPG architecture, and is manufactured on a 3 nm process at Intel. Its transistor count and die size are listed as unknown.

The generation labels differ accordingly. The MI308X belongs to the Instinct (MIx) generation, while the Arc B390 belongs to the Arc Graphics-M (Panther Lake) generation. The MI308X's predecessor is listed as Radeon Instinct, and its release date is 2023-12-05. The Arc B390 has no predecessor listed and a release date of 2026-01-26. The MI308X's production status is not recorded, while the Arc B390 is marked as active.

Core counts differentiate the two sharply. The MI308X has 19,456 shading units and 1,216 texture mapping units. The Arc B390 has 1,536 shading units and 48 texture mapping units. The MI308X has no ROPs, while the Arc B390 has 24. The MI308X has no ray tracing cores listed, while the Arc B390 has 12. Neither product lists tensor core counts.

Clock behavior also diverges. The MI308X runs at a base clock of 1000 MHz and a boost clock of 2100 MHz, with memory clocked at 1300 MHz or 5.2 Gbps effective. The Arc B390 runs at a base clock of 300 MHz and a boost clock of 2500 MHz, with system shared memory. The MI308X's boost clock is lower, but its massive core count and HBM3 memory compensate.

Power and physical configuration present further contrasts. The MI308X has a TDP of 750 W, a slot width of OAM Module, no power connectors, and a suggested PSU of 1150 W. The Arc B390 has a TDP of 80 W, a slot width of IGP, no power connectors, and no suggested PSU. The MI308X uses a PCIe 5.0 x16 bus interface, while the Arc B390 uses an integrated graphics processor bus interface. The MI308X has no display outputs, while the Arc B390's display outputs are portable device dependent.

FAQ

Q: Which processor has higher FP32 compute performance?

A: The AMD Instinct MI308X records 81.72 TFLOPS in FP32, while the Intel Arc B390 records 7.680 TFLOPS. The MI308X delivers more than ten times the FP32 throughput.

Q: How much memory bandwidth does each processor provide?

A: The MI308X has 192 GB of HBM3 memory with 5.32 TB/s of bandwidth. The Arc B390 uses system shared memory, and its bandwidth is listed as system dependent.

Q: What graphics API support does each processor offer?

A: The Arc B390 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The MI308X lists N/A for DirectX, OpenGL, and Vulkan.

Q: What are the TDP ratings for both processors?

A: The MI308X has a TDP of 750 W, while the Arc B390 has a TDP of 80 W.

Q: How does the Intel Arc B390 compare to its nearest rivals in 3DMark Steel Nomad DX12?

A: The Arc B390 scores 1482, which is 1.3% ahead of the NVIDIA GeForce GT 520MX, 1.5% ahead of the NVIDIA GeForce 800M, 2.5% ahead of the NVIDIA GeForce GT 625 OEM, and 2.7% ahead of the NVIDIA GeForce GT 710.

Q: Which processor has a higher boost clock?

A: The Arc B390 boosts to 2500 MHz, while the MI308X boosts to 2100 MHz.

The Verdict

The data describes two products with no overlap in intended function. The AMD Instinct MI308X is a compute accelerator built on the CDNA 3.0 architecture with 153,000 million transistors, 19,456 shading units, 192 GB of HBM3 memory, and 5.32 TB/s of bandwidth. Its FP32 throughput of 81.72 TFLOPS and texture rate of 2,553.6 GTexel/s position it for heavy data-parallel workloads. The absence of display outputs, raster operations, and consumer graphics API support confirms that it is not a graphics card in the conventional sense.

The Intel Arc B390 is an integrated graphics processor based on the Xe3-LPG architecture, with 1,536 shading units, 12 ray tracing cores, 24 ROPs, and support for DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4. Its benchmark score of 1482 places it at the 9th percentile among all GPUs, just ahead of several NVIDIA entry-level parts by margins of 1.3% to 2.7%. Its 80 W TDP and system shared memory reflect a low-power, portable-oriented design.

The choice between these two comes down to workload type. For compute-heavy tasks such as large-scale matrix operations or high-bandwidth data processing, the MI308X's specifications dominate in every measurable compute category. For graphics rendering, display output, and API compatibility, the Arc B390 is the only option of the two, as the MI308X has no graphics pipeline support. The MI308X's percentile rank of 50 with no benchmarks indicates an unverified performance profile in the database, while the Arc B390's percentile rank of 9 is backed by a single recorded benchmark. Each product serves a distinct segment, and the data does not suggest any meaningful overlap in their respective domains.