Intel Arc Pro B390 vs Intel Data Center GPU Max 1350 Comparison

Intel Arc Pro B390 and Intel Data Center GPU Max 1350 occupy opposite ends of the Intel graphics spectrum. The B390 is a compact integrated graphics solution built for mobile and embedded platforms, while the Max 1350 is a massive accelerator designed for data center compute workloads. The recorded data shows two products with different architectures, different memory strategies, and different performance targets. This analysis uses only the database entries to compare where each part wins and who should consider it.

Where Each One Wins

The Intel Arc Pro B390 wins in any scenario where power draw and physical footprint are primary constraints. Its 80 W TDP and IGP slot width mean it can operate without a dedicated power connector, making it suitable for thin, portable devices. The B390 also supports a full modern feature set: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. That combination of low power, integrated design, and current API support gives it a clear edge for client-side graphics, media playback, and light rendering tasks on the move.

The Intel Data Center GPU Max 1350 wins in raw compute throughput and memory capacity. Its 44.44 TFLOPS FP32 and 44.44 TFLOPS FP16 (1:1) dwarf the B390's 7.680 TFLOPS FP32 and 15.36 TFLOPS FP16 (2:1). The Max 1350 also carries 96 GB of HBM2e memory across an 8192-bit bus, delivering 2.46 TB/s of bandwidth. That is a compute and memory combination built for large-scale data center workloads such as AI training, scientific simulation, and high-performance computing. The Max 1350 has no display outputs, confirming it is not intended for graphics output.

The split is clean: the B390 wins on efficiency, integration, and client API support; the Max 1350 wins on absolute performance, memory capacity, and bandwidth.

FAQ

Q: Which GPU has higher FP32 performance?

A: The Intel Data Center GPU Max 1350 delivers 44.44 TFLOPS FP32, compared to the Arc Pro B390's 7.680 TFLOPS FP32. The Max 1350 provides roughly 5.8 times the FP32 throughput.

Q: How much memory does each GPU have?

A: The Arc Pro B390 uses System Shared memory, meaning its memory size, type, and bus width are all system-dependent. The Data Center GPU Max 1350 has 96 GB of dedicated HBM2e memory on an 8192-bit bus.

Q: Which GPU supports newer graphics APIs?

A: The Arc Pro B390 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The Data Center GPU Max 1350 supports DirectX 12 (12_1) and OpenGL 4.6, but has no Vulkan support listed.

Q: What are the TDP differences?

A: The Arc Pro B390 has an 80 W TDP and uses no power connectors. The Data Center GPU Max 1350 has a 450 W TDP and requires a suggested PSU of 850 W.

Q: What is the process node for each chip?

A: The Arc Pro B390 uses Intel's 3 nm process, while the Data Center GPU Max 1350 uses Intel's 10 nm process.

Q: Does either GPU have display outputs?

A: The Arc Pro B390 has display outputs described as Portable Device Dependent. The Data Center GPU Max 1350 has no display outputs at all.

Architecture Differences

The two GPUs are built on entirely different architectures. The Arc Pro B390 uses the Xe3-LPG architecture on a chip codenamed Panther Lake, part of the Arc Graphics-WM (Panther Lake) generation. It is manufactured on Intel's 3 nm process. The Data Center GPU Max 1350 uses the Generation 12.5 architecture on a chip codenamed Ponte Vecchio, part of the Data Center GPU (Ponte Vecchio) generation, manufactured on Intel's 10 nm process.

The transistor counts tell a story of scale. The Max 1350 packs 100,000 million transistors on a 1280 mm² die, with a transistor density of 78.1M per mm². The B390 lists its transistor count and die size as unknown, but its 3 nm process and integrated design suggest a much smaller chip.

The memory architectures are fundamentally different. The B390 uses System Shared memory, where the GPU borrows from the system's main memory, making bandwidth and capacity system-dependent. The Max 1350 uses 96 GB of dedicated HBM2e memory with an 8192-bit bus and 2.46 TB/s bandwidth.

The execution resources differ sharply. The B390 has 1536 shading units, 48 TMUs, 24 ROPs, and 12 ray tracing cores. The Max 1350 has 14336 shading units, 896 TMUs, 0 ROPs, and 112 ray tracing cores. The Max 1350 has no ROPs because it is not designed for rasterization output. The B390 also lists no tensor cores, and the Max 1350 lists none either.

The bus interfaces also differ. The B390 is an IGP with no slot width beyond that, while the Max 1350 uses a PCIe 5.0 x16 interface and comes as an OAM Module.

Head-to-Head Benchmarks

The database contains no direct head-to-head benchmark results between these two GPUs, and no nearest rival data is listed for either part. The comparison must therefore rely on the recorded specification data.

The largest single performance gap is in FP32 throughput. The Max 1350 delivers 44.44 TFLOPS, which is 5.8 times the B390's 7.680 TFLOPS. In FP16, the Max 1350 again delivers 44.44 TFLOPS at a 1:1 ratio, while the B390 delivers 15.36 TFLOPS at a 2:1 ratio. The Max 1350's FP16 performance is 2.9 times the B390's.

Memory bandwidth is another enormous gap. The Max 1350's 2.46 TB/s is orders of magnitude beyond the B390's system-dependent shared memory. The B390's memory bandwidth is listed as System Dependent, meaning it has no fixed figure, and its memory type is System Shared. The Max 1350's 96 GB of HBM2e on an 8192-bit bus is a class of memory the B390 cannot approach.

Texture rate shows the Max 1350 at 1,388.8 GTexel/s versus the B390's 120.0 GTexel/s, a difference of roughly 11.6 times. Pixel rate is a different story: the B390 records 60.00 GPixel/s while the Max 1350 records 0 MPixel/s. The Max 1350 has no ROPs and no display outputs, so it cannot rasterize pixels for display purposes.

Clock speeds also differ. The B390 has a base clock of 300 MHz and a boost clock of 2500 MHz. The Max 1350 has a base clock of 750 MHz and a boost clock of 1550 MHz. The B390 boosts higher, but the Max 1350 has far more execution units working in parallel.

The Max 1350's FP32 advantage is consistent with a data center compute part. The B390's higher boost clock and pixel rate reflect a client graphics part. The 12 ray tracing cores in the B390 support real-time ray tracing workloads, while the Max 1350's 112 ray tracing cores are present but serve compute-oriented rendering, not display output.

The Verdict

The data shows two parts with no overlap in intended use. The Intel Arc Pro B390 is for systems where the GPU must live inside the processor package, draw 80 W or less, and still support modern graphics APIs including DirectX 12 Ultimate and Vulkan 1.4. Its 7.680 TFLOPS FP32, 60.00 GPixel/s pixel rate, and 2500 MHz boost clock make it a sensible choice for portable devices needing capable integrated graphics. The B390's System Shared memory means its effective performance depends on the host system's memory configuration.

The Intel Data Center GPU Max 1350 is for systems that need maximum compute throughput and memory bandwidth. Its 44.44 TFLOPS FP32, 44.44 TFLOPS FP16, 96 GB HBM2e, and 2.46 TB/s bandwidth place it in a different performance class entirely. The 450 W TDP and 850 W suggested PSU indicate a machine designed for sustained compute loads, not battery operation. The lack of display outputs and 0 MPixel/s pixel rate confirm it is not a graphics output device.

Users who need a mobile or embedded GPU with current API support should choose the Arc Pro B390. Users who need a data center accelerator for compute-heavy workloads should choose the Data Center GPU Max 1350. The data does not support any other conclusion.

Specification Differences

The two parts differ across nearly every specification field. The B390 uses the Xe3-LPG architecture on a 3 nm process, while the Max 1350 uses Generation 12.5 on a 10 nm process. The B390 has 1536 shading units, 48 TMUs, 24 ROPs, and 12 ray tracing cores. The Max 1350 has 14336 shading units, 896 TMUs, 0 ROPs, and 112 ray tracing cores.

Clocks differ: the B390 runs at 300 MHz base and 2500 MHz boost; the Max 1350 runs at 750 MHz base and 1550 MHz boost. The B390's memory is System Shared with System Dependent bandwidth; the Max 1350 uses 96 GB HBM2e with 2.46 TB/s bandwidth on an 8192-bit bus.

Compute rates differ: the B390 delivers 7.680 TFLOPS FP32 and 15.36 TFLOPS FP16 (2:1); the Max 1350 delivers 44.44 TFLOPS FP32 and 44.44 TFLOPS FP16 (1:1). The B390 records 60.00 GPixel/s pixel rate and 120.0 GTexel/s texture rate; the Max 1350 records 0 MPixel/s pixel rate and 1,388.8 GTexel/s texture rate.

Power and physical specs differ: the B390 is an 80 W IGP with no power connectors; the Max 1350 is a 450 W OAM Module with an 850 W suggested PSU. The B390 uses an IGP bus interface; the Max 1350 uses PCIe 5.0 x16. The B390 has Portable Device Dependent display outputs; the Max 1350 has no display outputs.

API support differs: the B390 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4; the Max 1350 supports DirectX 12 (12_1) and OpenGL 4.6, with no Vulkan listed. The Max 1350 has a known transistor count of 100,000 million and die size of 1280 mm², while the B390 lists these as unknown. The Max 1350 also has a listed transistor density of 78.1M per mm², a figure the B390 does not provide.