AMD Instinct MI300 vs Intel Arc Pro B370 Comparison

Head-to-Head Benchmarks

The recorded database contains no benchmark scores for either the AMD Instinct MI300 or the Intel Arc Pro B370. Both GPUs have an average benchmark score of 0 and a percentile ranking of 50 against all GPUs. As such, there are no direct performance comparisons, no delta percentages, and no per-test wins to report for either product in this head-to-head matchup.

The absence of measured data does not diminish the significance of the specifications on file. The AMD Instinct MI300 delivers FP32 compute at 47.87 TFLOPS and FP16 at 47.87 TFLOPS with a 1:1 ratio. The Intel Arc Pro B370 delivers FP32 at 6.144 TFLOPS and FP16 at 12.29 TFLOPS with a 2:1 ratio. These figures show the MI300 holds a 7.79x advantage in FP32 throughput and a 3.89x advantage in FP16 throughput. The MI300 also posts a texture rate of 1,496.0 GTexel/s versus 96.00 GTexel/s for the Arc Pro B370, a 15.58x gap. The Arc Pro B370 counters with a pixel rate of 48.00 GPixel/s, while the MI300 records 0 MPixel/s, meaning the AMD part cannot rasterize pixels at all in its current configuration.

Memory capacity differs enormously. The MI300 ships with 128 GB of HBM3 across an 8192-bit bus, delivering 5.32 TB/s of bandwidth. The Arc Pro B370 uses system shared memory with system dependent bandwidth, so no fixed capacity or throughput figure exists in the database. Clock behavior also differs: the MI300 runs at a 1000 MHz base and 1700 MHz boost, while the Arc Pro B370 runs at a 300 MHz base and 2400 MHz boost. The Intel part has a higher boost clock by 700 MHz, but the AMD part starts from a much higher base.

Rasterization hardware shows a stark split. The MI300 has 14,080 shading units, 880 TMUs, and 0 ROPs. The Arc Pro B370 has 1,280 shading units, 40 TMUs, 20 ROPs, and 10 ray tracing cores. The MI300 has 11x more shading units and 22x more TMUs, but it lacks ROPs entirely, so its pixel rate is zero. The Arc Pro B370 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, while the MI300 reports N/A for all three APIs. This indicates the MI300 is not designed for graphics rendering workloads in the traditional sense, while the Arc Pro B370 carries full modern graphics API support.

Power and physical requirements also contrast sharply. The MI300 has a TDP of 600 W with a suggested PSU of 1000 W and uses 2x 8-pin power connectors. The Arc Pro B370 has a TDP of 25 W, uses no power connectors, and requires no suggested PSU. The MI300 measures 267 mm in length and 111 mm in height, while the Arc Pro B370 is an integrated graphics processor (IGP) with no listed dimensions. The MI300 connects via PCIe 5.0 x16 and has no display outputs; the Arc Pro B370 also uses an IGP bus interface and lists display outputs as portable device dependent.

The Verdict

The data indicates two fundamentally different products aimed at different workloads. The AMD Instinct MI300 is a compute accelerator. Its 128 GB HBM3 memory, 5.32 TB/s bandwidth, and 47.87 TFLOPS FP32 output point to large-scale data center tasks such as AI training and scientific simulation. Its lack of ROPs, N/A graphics APIs, and no display outputs confirm it is not a rendering card.

The Intel Arc Pro B370 is an integrated GPU for mobile or compact platforms. Its 25 W TDP, 300 MHz base clock, and system shared memory fit a low-power, portable device context. It supports DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, includes 10 ray tracing cores, and has a pixel rate of 48.00 GPixel/s. This makes it suitable for mainstream graphics, media playback, and light content creation on devices where discrete GPUs are impractical.

Who should pick which comes down to workload. Users running high-performance compute workloads that require massive memory bandwidth and FP32 throughput should select the MI300. Users needing a power-efficient integrated GPU with modern API support and ray tracing capability should select the Arc Pro B370. The MI300 offers 7.79x higher FP32 throughput and 15.58x higher texture rate, but it cannot render pixels or output video. The Arc Pro B370 offers graphics features the MI300 lacks entirely, but its compute throughput is far lower. There is no overlap in intended use cases.

FAQ

Q: Which GPU has higher FP32 compute performance?

A: The AMD Instinct MI300 has FP32 performance of 47.87 TFLOPS, versus 6.144 TFLOPS for the Intel Arc Pro B370. The MI300 leads by 7.79x.

Q: Does the Intel Arc Pro B370 support ray tracing?

A: Yes, the Arc Pro B370 includes 10 ray tracing cores. The AMD Instinct MI300 has no listed ray tracing cores.

Q: Which GPU has more memory bandwidth?

A: The AMD Instinct MI300 has 5.32 TB/s bandwidth from 128 GB of HBM3 on an 8192-bit bus. The Intel Arc Pro B370 uses system shared memory with system dependent bandwidth, so no fixed bandwidth figure is recorded.

Q: Can the AMD Instinct MI300 output video to a display?

A: No. The MI300 lists no display outputs and reports N/A for DirectX, OpenGL, and Vulkan support. Its pixel rate is 0 MPixel/s.

Q: What is the power draw difference between these two GPUs?

A: The MI300 has a TDP of 600 W and requires a 1000 W suggested PSU. The Arc Pro B370 has a TDP of 25 W and uses no power connectors.

Q: Which GPU is smaller or easier to install?

A: The Arc Pro B370 is an integrated GPU (IGP) with no listed dimensions. The MI300 is a discrete card measuring 267 mm in length and 111 mm in height, requiring a PCIe 5.0 x16 slot.

Specification Differences

The two products differ across nearly every recorded specification field.

• Process node: The AMD Instinct MI300 uses a 5 nm TSMC process. The Intel Arc Pro B370 uses a 3 nm Intel process.
• Transistors: The MI300 has 153,000 million transistors on a 1017 mm² die, with a density of 150.4M per mm². The Arc Pro B370 lists transistors and die size as unknown.
• Base clock: MI300 runs at 1000 MHz. Arc Pro B370 runs at 300 MHz.
• Boost clock: MI300 runs at 1700 MHz. Arc Pro B370 runs at 2400 MHz.
• Memory: MI300 has 128 GB HBM3 on an 8192-bit bus with 5.32 TB/s bandwidth. Arc Pro B370 uses system shared memory with system dependent bandwidth.
• Memory clock: MI300 lists 1300 MHz (5.2 Gbps effective). Arc Pro B370 lists system shared memory.
• Shading units: MI300 has 14,080. Arc Pro B370 has 1,280.
• TMUs: MI300 has 880. Arc Pro B370 has 40.
• ROPs: MI300 has 0. Arc Pro B370 has 20.
• Ray tracing cores: MI300 has none listed. Arc Pro B370 has 10.
• Pixel rate: MI300 is 0 MPixel/s. Arc Pro B370 is 48.00 GPixel/s.
• Texture rate: MI300 is 1,496.0 GTexel/s. Arc Pro B370 is 96.00 GTexel/s.
• FP16: MI300 is 47.87 TFLOPS (1:1). Arc Pro B370 is 12.29 TFLOPS (2:1).
• TDP: MI300 is 600 W. Arc Pro B370 is 25 W.
• Power connectors: MI300 uses 2x 8-pin. Arc Pro B370 uses none.
• Suggested PSU: MI300 lists 1000 W. Arc Pro B370 lists none.
• Bus interface: MI300 uses PCIe 5.0 x16. Arc Pro B370 uses IGP.
• Display outputs: MI300 has none. Arc Pro B370 lists portable device dependent.
• APIs: MI300 lists N/A for DirectX, OpenGL, and Vulkan. Arc Pro B370 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.
• Dimensions: MI300 is 267 mm long, 111 mm tall. Arc Pro B370 has no dimensions listed.
• Production status: MI300 has none listed. Arc Pro B370 is active.
• Release date: MI300 released 2023-01-03. Arc Pro B370 releases 2026-01-26.
• Predecessor: MI300 lists Radeon Instinct. Arc Pro B370 lists HD Graphics-WM.

Architecture Differences

The AMD Instinct MI300 uses the CDNA 3.0 architecture on a 5 nm TSMC process, with 153,000 million transistors on a 1017 mm² die. The chip name is Aqua Vanjaram. It belongs to the Instinct (MIx) generation. The architecture prioritizes compute throughput with 14,080 shading units and 880 TMUs, but it omits ROPs entirely. The FP16 ratio is 1:1, meaning FP16 and FP32 run at the same throughput. The memory subsystem uses 128 GB of HBM3 across an 8192-bit bus, yielding 5.32 TB/s bandwidth. The MI300 has no display outputs and no graphics API support, confirming a compute-only design.

The Intel Arc Pro B370 uses the Xe3-LPG architecture on a 3 nm Intel process, with the chip named Panther Lake. It belongs to the Arc Graphics-WM (Panther Lake) generation. The design integrates graphics into the processor package, as shown by the IGP bus interface and system shared memory. It includes 1,280 shading units, 40 TMUs, 20 ROPs, and 10 ray tracing cores. The FP16 ratio is 2:1, so FP16 throughput is double FP32. The GPU supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. It has no dedicated memory; bandwidth is system dependent. The Arc Pro B370 is an active production part with a release date in 2026, while the MI300 has no production status listed.

The architectural philosophies diverge clearly. CDNA 3.0 on the MI300 is built for raw compute scale, with massive memory bandwidth and high FP32/FP16 throughput. Xe3-LPG on the Arc Pro B370 is built for integrated graphics efficiency, with a 25 W TDP, modern graphics APIs, and ray tracing support. The MI300 uses a 5 nm process with a 1017 mm² die, while the Arc Pro B370 uses a 3 nm process with unknown die size. The MI300 has no rasterization hardware, while the Arc Pro B370 has full ROP and ray tracing support.

Where Each One Wins

The AMD Instinct MI300 wins in compute-heavy workloads that demand massive parallel throughput and memory bandwidth. Its 47.87 TFLOPS FP32 and FP16 performance is 7.79x higher than the Arc Pro B370 in FP32 and 3.89x higher in FP16. Its 1,496.0 GTexel/s texture rate is 15.58x higher. The 128 GB HBM3 memory pool with 5.32 TB/s bandwidth supports large datasets that would exhaust system shared memory. The 600 W TDP and 1000 W suggested PSU indicate a designed-for-sustained compute environment. The MI300 wins in AI training, scientific simulation, and any workload that can use its 14,080 shading units.

The Intel Arc Pro B370 wins in graphics rendering and power-constrained scenarios. Its 20 ROPs and 48.00 GPixel/s pixel rate enable actual rasterization output, which the MI300 cannot do at all. Its 10 ray tracing cores provide hardware ray tracing, absent from the MI300. DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4 support make it compatible with modern graphics software, while the MI300 reports N/A for all APIs. The 25 W TDP and no power connectors allow operation in compact systems without PSU upgrades. The 300 MHz base clock and 2400 MHz boost clock show a wide dynamic range, useful for power management. The Arc Pro B370 wins in portable devices, everyday graphics, and any task requiring display output.

For raw compute throughput, the MI300 is the clear choice. For rendering, ray tracing, and low-power integration, the Arc Pro B370 is the only option between the two. The data shows no scenario where both compete directly, as the MI300 lacks graphics functionality and the Arc Pro B370 lacks comparable compute scale.