AMD Radeon Instinct MI300 vs Intel Arc Pro B60 Dual Comparison

Head-to-Head Benchmarks

The recorded database contains no benchmark scores for either the AMD Radeon Instinct MI300 or the Intel Arc Pro B60 Dual. Both cards hold a percentile ranking of 50 against all GPUs, and both carry an average benchmark score of 0. The head-to-head benchmark array is empty, and neither card registers a win in direct comparison. This absence of measured performance data means any attempt to rank one above the other on empirical grounds is unsupported by the current records. What the data does provide is a detailed specification profile for each card, which allows for a comparative analysis of their theoretical compute ceilings, memory subsystems, and architectural intent. The MI300 is a data center accelerator aimed at massive compute workloads, while the Arc Pro B60 Dual is a professional graphics card with display outputs and an active production status. Their respective roles in the market are clear from the specification sheets alone, even without benchmark validation.

FAQ

Q: Which card has the higher FP32 compute throughput?

A: The AMD Radeon Instinct MI300 delivers 47.87 TFLOPS of FP32 performance, while the Intel Arc Pro B60 Dual delivers 12.29 TFLOPS. The MI300 is roughly 3.9 times ahead in single-precision floating-point throughput based on these recorded figures.

Q: What are the memory capacities and types of each card?

A: The MI300 uses 128 GB of HBM3 memory on an 8192-bit bus, yielding 6.55 TB/s of bandwidth. The Arc Pro B60 Dual uses 24 GB of GDDR6 memory on a 192-bit bus, yielding 456.0 GB/s of bandwidth. The MI300 has over 14 times the memory bandwidth of the Intel card.

Q: Do both cards support PCIe 5.0?

A: Yes, both use the PCIe 5.0 interface. The MI300 connects via a x16 link, while the Arc Pro B60 Dual uses a x8 link. The MI300 has double the lane count for host communication.

Q: Which card has display outputs?

A: The Intel Arc Pro B60 Dual has 4x mini-DisplayPort 2.1 outputs, making it suitable for direct display connection. The AMD Radeon Instinct MI300 has no display outputs, confirming its role as a compute-only accelerator.

Q: What are the respective release dates?

A: The AMD Radeon Instinct MI300 was released on January 3, 2023. The Intel Arc Pro B60 Dual was released on September 4, 2025. The Intel card arrived roughly two and a half years later.

Q: What is the power connector requirement for each card?

A: The MI300 requires 2x 8-pin power connectors and a suggested power supply of 1000 W. The Arc Pro B60 Dual requires a single 16-pin connector and a suggested power supply of 800 W.

Architecture Differences

The two cards represent fundamentally different architectural philosophies. The AMD Radeon Instinct MI300 uses the CDNA 3.0 architecture, built on the Aqua Vanjaram chip, and belongs to the Radeon Instinct (MIx) generation. CDNA 3.0 is a compute-optimized design with no display outputs, no DirectX, OpenGL, or Vulkan API support listed, and a pixel rate recorded as 0 MPixel/s. It is a pure accelerator for data center workloads, with 14,080 shading units, 880 texture mapping units, and zero ROPs. The lack of raster output units and the absence of any graphics API support make it clear that this chip does not render graphics frames; it processes compute workloads in bulk.

The Intel Arc Pro B60 Dual uses the Xe2-HPG architecture, built on the BMG-G21 chip, and belongs to the Battlemage (Pro Series) generation. This is a graphics-capable architecture with 2,560 shading units, 160 TMUs, 80 ROPs, and 20 ray tracing cores. It supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, and it provides 4x mini-DisplayPort 2.1 outputs. The pixel rate of 192.0 GPixel/s and texture rate of 384.0 GTexel/s confirm that this card performs traditional graphics rasterization and ray tracing tasks. The MI300's texture rate of 1,496.0 GTexel/s is far higher, but this figure is tied to compute throughput rather than visual rendering.

Process node and manufacturing details show both cards are fabricated by TSMC on a 5 nm process. The MI300 die measures 1017 mm² with 153,000 million transistors, giving a transistor density of 150.4M per mm². The Arc Pro B60 Dual die measures 272 mm² with 19,600 million transistors, giving a transistor density of 72.1M per mm². The MI300 is nearly four times the die size and has roughly 7.8 times the transistor count. This scale difference reflects the MI300's role as a high-end accelerator with enormous memory and compute resources, versus the Intel card's more modest professional graphics positioning.

The MI300's FP16 throughput is 383.0 TFLOPS at an 8:1 ratio, while the Arc Pro B60 Dual delivers 24.58 TFLOPS at a 2:1 ratio. The MI300's FP16 advantage is substantial, but the ratio difference indicates the AMD card uses a more aggressive mixed-precision design. The Intel card's 2:1 FP16 ratio suggests a more balanced relationship between FP16 and FP32, which is typical for graphics-oriented architectures where both precision levels are used in rendering and compute tasks.

Specification Differences

The AMD Radeon Instinct MI300 and Intel Arc Pro B60 Dual differ across nearly every measurable specification field. The MI300 has a base clock of 1000 MHz and a boost clock of 1700 MHz. The Arc Pro B60 Dual has a base clock of 2000 MHz and a boost clock of 2400 MHz. The Intel card runs at substantially higher clock frequencies, but this does not compensate for the AMD card's much larger compute unit count.

Memory configurations are starkly different. The MI300 offers 128 GB of HBM3 with an 8192-bit bus and 6.55 TB/s bandwidth. The Arc Pro B60 Dual offers 24 GB of GDDR6 with a 192-bit bus and 456.0 GB/s bandwidth. The MI300's memory clock is listed as 1600 MHz with 6.4 Gbps effective speed. The Intel card's memory clock is 2375 MHz with 19 Gbps effective speed. While the Intel card has a faster memory clock in MHz terms, the AMD card's massive bus width gives it a bandwidth advantage of over 14 times.

The MI300 has 14,080 shading units and 880 TMUs, with zero ROPs. The Arc Pro B60 Dual has 2,560 shading units, 160 TMUs, 80 ROPs, and 20 ray tracing cores. The MI300's FP32 output is 47.87 TFLOPS, while the Intel card delivers 12.29 TFLOPS. The MI300's texture rate is 1,496.0 GTexel/s versus 384.0 GTexel/s for the Intel card. The MI300's pixel rate is 0 MPixel/s, while the Intel card manages 192.0 GPixel/s.

Power consumption differs as well. The MI300 has a TDP of 600 W, requires 2x 8-pin power connectors, and suggests a 1000 W power supply. The Arc Pro B60 Dual has a TDP of 400 W, requires a single 16-pin connector, and suggests an 800 W power supply. The Intel card is more power-efficient per watt on a raw TDP basis, but the AMD card delivers far more absolute compute.

Physical dimensions show the MI300 at 267 mm in length and 111 mm in height, while the Arc Pro B60 Dual measures 300 mm in length, 110 mm in height, and 40 mm in width. The Intel card is longer but similar in height. The Intel card is explicitly listed as a dual-slot card, while the MI300's slot width is not recorded. The bus interface differs: PCIe 5.0 x16 for the MI300 versus PCIe 5.0 x8 for the Arc Pro B60 Dual.

The MI300's predecessor is listed as FirePro Data Center, while no predecessor is recorded for the Intel card. The Intel card has an active production status and a launch MSRP of 1,199 USD. The MI300 has no launch MSRP recorded and no production status listed. The MI300's release date is January 3, 2023; the Intel card's release date is September 4, 2025.

The Verdict

The data indicates two cards with almost no overlap in intended use. The AMD Radeon Instinct MI300 is a high-end compute accelerator. Its 128 GB of HBM3 memory, 6.55 TB/s bandwidth, 47.87 TFLOPS FP32, and 383.0 TFLOPS FP16 position it for large-scale scientific computing, AI training, and inference workloads where massive memory capacity and bandwidth are critical. The absence of display outputs and graphics API support confirms this is not a card for desktop or workstation rendering. The 1000 W suggested power supply and 600 W TDP reflect the demands of such a high-throughput device.

The Intel Arc Pro B60 Dual is a professional graphics card. Its 24 GB of GDDR6 memory, 456.0 GB/s bandwidth, 12.29 TFLOPS FP32, 24.58 TFLOPS FP16, 20 ray tracing cores, and 4x mini-DisplayPort 2.1 outputs make it suitable for professional visualization, content creation, and graphics workloads that require both compute and rendering capabilities. The active production status and 1,199 USD launch MSRP indicate a current, purchasable product aimed at the professional graphics segment.

Benchmark results are absent from the database, so neither card can be declared a performance winner on measured data. The specification sheets, however, define clear roles. The MI300 is the compute powerhouse, with roughly 3.9 times the FP32 throughput, over 14 times the memory bandwidth, and over 5 times the memory capacity of the Intel card. The Arc Pro B60 Dual is the graphics-capable option, with ray tracing support, full graphics API coverage, and display outputs that the MI300 entirely lacks.

For workloads centered on raw compute throughput and vast memory pools, the MI300 is the only viable choice between these two. For tasks requiring graphics output, ray tracing, or DirectX and Vulkan support, the Arc Pro B60 Dual is the only card that can perform them. The choice between these cards is not a matter of one being better than the other; it is a matter of selecting the tool that matches the job. The data supports this binary conclusion: the MI300 for compute, the Arc Pro B60 Dual for graphics.