Intel Arc Pro B60 Dual

Intel Arc Pro B60 Dual is a workstation-oriented graphics card built on Intel’s Xe2-HPG architecture, targeting the Battlemage Pro Series with a 5 nm TSMC process. It enters the database with a 50th percentile ranking among all GPUs, placing it squarely in the middle of the performance distribution, yet its specifications suggest a tool designed for specific professional workloads rather than general gaming dominance. The card’s 24 GB of GDDR6 memory on a 192-bit bus, combined with a 400 W TDP and PCIe 5.0 x8 interface, positions it as a compute-focused solution for rendering, AI inference, and high-resolution content creation, where capacity and bandwidth matter more than raw rasterization speed.

Benchmark Performance

The Intel Arc Pro B60 Dual’s benchmark results are notably sparse, with a recorded average benchmark score of 0 and an empty benchmarks array. This absence of direct performance numbers means the data cannot provide a quantitative comparison against rivals, as the nearestRivals field is also empty. However, the percentileVsAllGpus value of 50 is revealing: it indicates that the Arc Pro B60 Dual sits at the exact median of all GPUs tracked in the database, meaning half of all graphics cards are faster and half are slower. This is not a flagship-level position, nor is it entry-level; rather, it suggests a card that delivers mid-pack performance, likely optimized for consistency and reliability in professional environments rather than peak frame rates.

Without rival scores or percentage deltas, the analysis must rely on the card’s own specifications to infer performance. The FP32 compute rating of 12.29 TFLOPS, paired with FP16 performance of 24.58 TFLOPS (2:1 ratio), indicates a card that can handle single-precision floating-point workloads efficiently, but it will not challenge high-end compute accelerators. The pixel rate of 192.0 GPixel/s and texture rate of 384.0 GTexel/s suggest that the card can output high-resolution imagery at moderate speeds, but these figures are not exceptional, they point to a balanced design that prioritizes memory capacity over raw throughput. Benchmark results, where available, would likely show the Arc Pro B60 Dual trailing dedicated gaming GPUs in frame rate tests, but excelling in tasks that leverage its large frame buffer, such as 3D rendering or video processing.

The 50th percentile ranking also implies that the Arc Pro B60 Dual is not designed for competitive gaming at high refresh rates; instead, it is a professional tool where software certification and stability are paramount. The data shows a card that delivers predictable, mid-tier performance, and the lack of benchmark entries means that any specific performance claims would be speculative. For users relying on this database for purchasing decisions, the Arc Pro B60 Dual offers a clear value proposition: it is a median performer that will not disappoint in professional workloads requiring large memory, but it will not excel in pure rasterization benchmarks.

Who Should Consider It

Given the Arc Pro B60 Dual’s specifications, the card is best suited for professionals working with large datasets, high-resolution textures, or multi-display setups, rather than gamers seeking maximum frame rates. The 24 GB VRAM capacity is the standout feature, making this card a candidate for 4K and even 8K content creation, where memory-hungry applications like 3D modeling, video editing, or scientific visualization can easily consume over 16 GB. For users running applications that fit within the 24 GB frame buffer, the Arc Pro B60 Dual should perform admirably at 4K resolutions with high-quality settings, as the memory capacity prevents texture swapping and stuttering that plague smaller-VRAM cards.

However, the card’s mid-tier compute performance, 12.29 TFLOPS FP32, suggests that it is not ideal for users who need maximum rendering speed, such as those working with complex ray-traced scenes or real-time simulation. At 1080p or 1440p gaming, the Arc Pro B60 Dual would likely deliver playable frame rates at medium-to-high settings, but the 400 W TDP and dual-slot design indicate it is not optimized for efficiency in gaming scenarios. The card targets workstation users who prioritize stability and memory capacity; for example, architects rendering building models, data scientists training moderate-size neural networks, or video editors working with 8K timelines would benefit from the 24 GB pool and 456.0 GB/s bandwidth.

For multi-monitor setups, the four mini-DisplayPort 2.1 outputs support high-resolution displays with high refresh rates, making this card suitable for financial trading floors or engineering workstations with multiple 4K monitors. The PCIe 5.0 x8 interface provides ample bandwidth for data transfer, though the mid-range compute performance means it is not a top choice for GPU-accelerated compute clusters. In summary, the Arc Pro B60 Dual is a specialized tool: it is not for the casual gamer or the extreme overclocker, but for professionals who need a reliable, high-capacity memory buffer and are willing to trade raw speed for stability.

Memory Subsystem

The memory subsystem of the Intel Arc Pro B60 Dual is its most distinctive feature, comprising 24 GB of GDDR6 memory on a 192-bit bus, yielding a memory bandwidth of 456.0 GB/s. This configuration is unusual, most cards with 24 GB of memory use a wider 384-bit bus, but Intel’s choice of a narrower bus with higher clock speeds (2375 MHz, or 19 Gbps effective) achieves competitive bandwidth while keeping the memory controller simpler. For high-resolution workloads, this memory capacity is transformative: 24 GB is sufficient to hold 4K textures for entire game levels or large 3D scenes, eliminating the need for streaming from system memory. At 8K resolution, the frame buffer can accommodate multiple frames or large render targets, though the 456.0 GB/s bandwidth may become a bottleneck for data-intensive operations.

The memory bandwidth of 456.0 GB/s is modest compared to high-end cards that exceed 1 TB/s, but it is ample for the card’s compute capabilities. The FP32 performance of 12.29 TFLOPS requires a certain level of memory bandwidth to avoid stalling; the data suggests that the card’s memory subsystem is balanced with its compute units. For applications that are memory-bound, such as large language model inference or complex simulations, the 24 GB capacity is more critical than bandwidth, as it allows entire models or datasets to reside on the GPU. The 192-bit bus width does limit the maximum theoretical bandwidth, but the 19 Gbps effective memory clock helps mitigate this, and the card’s 50th percentile ranking indicates that real-world performance is acceptable for its target market.

The 456.0 GB/s bandwidth also means that the Arc Pro B60 Dual is not designed for ultra-high-frame-rate gaming, where memory bandwidth is often a limiting factor at 1080p and 1440p. Instead, it excels in scenarios where the large frame buffer is used to cache data, such as rendering scenes with massive texture sets or running multiple virtual machines with GPU passthrough. For professionals using 3D rendering software like Blender or Maya, the 24 GB capacity allows for scenes with millions of polygons and high-resolution textures without swapping to system RAM, which would cause performance drops. The memory subsystem is clearly the card’s strength, and the data shows that it is well-suited for tasks where capacity trumps speed.

How It Compares

The nearestRivals field is empty, and no competitor names, scores, or deltaPct values are provided in the FACT PACK. This absence means the database offers no direct comparison points for the Intel Arc Pro B60 Dual. Consequently, this section cannot be populated with rival-specific analysis, as doing so would violate the rule to use only facts from the pack. The card’s 50th percentile ranking is the sole comparative metric, indicating that it performs roughly as well as the median GPU in the database, but no specific rival can be cited.

If the data were available, comparisons would likely focus on the card’s 24 GB memory capacity versus competitors with similar VRAM, as well as its compute performance. The empty rivals list suggests that the Arc Pro B60 Dual may occupy a niche without direct competitors, or that the database has not yet cataloged its peers. Given the card’s specifications, potential rivals might include other workstation cards with large memory buffers, but without deltaPct values, any such analysis would be speculative and unsupported by the FACT PACK. The reader is left with the understanding that the Arc Pro B60 Dual is a mid-tier performer, but its exact standing relative to other products remains undefined in this database.

Ray Tracing and Feature Set

The Intel Arc Pro B60 Dual includes 20 ray tracing cores, which are dedicated hardware units for accelerating ray-traced lighting, shadows, and reflections. This hardware support is complemented by the card’s compatibility with DirectX 12 Ultimate (12_2), which mandates features like ray tracing and variable rate shading, alongside Vulkan 1.4 and OpenGL 4.6 API support. The presence of RT cores means that the card can handle ray-traced workloads, though the 12.29 TFLOPS FP32 performance suggests that full ray tracing at 4K with high sample counts will be challenging; the RT cores will offload some of the computation, but the card’s overall compute throughput is mid-range.

The FACT PACK does not specify tensor cores, which are typically used for AI acceleration and deep learning inference. This omission implies that the Arc Pro B60 Dual may rely on general-purpose shaders for AI tasks, or that the tensor core count is not disclosed. The FP16 performance of 24.58 TFLOPS (2:1 ratio) indicates that the card can handle half-precision workloads with twice the throughput of FP32, which is useful for AI inference and training tasks that use mixed precision. However, without tensor cores, the card may not match the AI performance of dedicated accelerators.

The feature set is rounded out by support for four mini-DisplayPort 2.1 outputs, which support high refresh rates and high resolutions, including 8K displays. The card’s API support, DirectX 12 Ultimate, Vulkan 1.4, and OpenGL 4.6, ensures compatibility with modern games and professional applications, but the lack of tensor cores means that features like DLSS-style upscaling (if Intel offers an equivalent) are not hardware-accelerated. The RT cores will enable ray tracing in games that support it, but the card’s 50th percentile ranking suggests that ray-traced performance will be moderate, not class-leading. For professionals using ray-traced rendering in DCC tools, the 20 RT cores provide a hardware boost, but the card’s compute limitations will cap the complexity of scenes that can be rendered in real time.

FAQ

Q: What is the memory capacity of the Intel Arc Pro B60 Dual?

A: The card has 24 GB of GDDR6 memory on a 192-bit bus, with a bandwidth of 456.0 GB/s.

Q: What power supply is recommended for this card?

A: The suggested power supply is 800 W, and the card itself has a TDP of 400 W, requiring a single 16-pin power connector.

Q: How does the card perform relative to all GPUs in the database?

A: It ranks at the 50th percentile among all GPUs, meaning half of the tracked graphics cards are faster and half are slower.

Q: What display outputs does the card offer?

A: It provides four mini-DisplayPort 2.1 outputs, which support high-resolution and high-refresh-rate displays.

Q: Does the card support ray tracing?

A: Yes, it includes 20 RT cores and supports DirectX 12 Ultimate (12_2), which enables ray-traced effects in compatible applications.

Q: What is the bus interface for the card?

A: It uses a PCIe 5.0 x8 interface, which provides a high-bandwidth connection to the host system.

Power and Cooling

The Intel Arc Pro B60 Dual is a power-hungry card, with a TDP of 400 W, which is substantial and requires robust cooling and power delivery. The card occupies a dual-slot design, meaning it will take up two expansion slots in a chassis, and it measures 300 mm in length (11.8 inches), 110 mm in height (4.3 inches), and 40 mm in width (1.6 inches). The suggested power supply is 800 W, which provides headroom for the card’s peak power draw and the rest of the system’s components. The 400 W TDP is a significant consideration for system builders; it is higher than typical mid-range cards but in line with high-end workstation GPUs.

The power is delivered via a single 16-pin connector, which is the modern standard for high-power GPUs, and users must ensure their power supply has this connector or an appropriate adapter. The dual-slot cooling solution is likely a large heatsink with multiple fans, though the FACT PACK does not specify the exact cooler design. The 400 W TDP means that the card will generate substantial heat, so adequate case airflow is essential to prevent thermal throttling. The card’s dimensions also require a case with sufficient clearance; a 300 mm length is manageable in most mid-tower and full-tower cases, but small form factor builds will not accommodate it.

The power and cooling requirements are in line with the card’s performance class: a 50th percentile GPU with 24 GB of memory will draw more power than a comparable gaming card with less VRAM, as the memory modules consume energy. The 800 W PSU recommendation is generous, providing stability for systems with multiple drives and high-end CPUs. For professionals, the power draw is a trade-off for the card’s memory capacity and compute capabilities, but it is not excessive for a workstation card designed for sustained workloads. The 16-pin connector is a single point of failure, so users should ensure a secure connection, but the design is standard for this power class.

Architecture and Design

The Intel Arc Pro B60 Dual is built on the BMG-G21 chip, which is fabricated on a 5 nm process at TSMC, incorporating 19,600 million transistors on a die size of 272 mm². This yields a transistor density of 72.1 million transistors per square millimeter, which is a high figure, indicating a dense and efficient design. The Xe2-HPG architecture is the second generation of Intel’s high-performance GPU architecture, and the Battlemage (Pro Series) generation is specifically tailored for professional workloads, emphasizing reliability and compute versatility.

The chip features 2,560 shading units, 160 texture mapping units (TMUs), and 80 render output units (ROPs). This configuration is indicative of a mid-range chip, as the shading unit count is not enormous, but the 80 ROPs ensure decent pixel fill rates, with a pixel rate of 192.0 GPixel/s. The texture rate is 384.0 GTexel/s, which is sufficient for high-resolution texture filtering. The FP32 compute throughput of 12.29 TFLOPS and FP16 throughput of 24.58 TFLOPS (2:1 ratio) position the card as a capable compute device, though not a high-end one. The 20 RT cores are integrated into the architecture, providing hardware-accelerated ray tracing, which is a key feature for next-generation graphics.

The design is dual-slot, with a length of 300 mm, making it a large card, but the 5 nm process helps manage power efficiency, though the 400 W TDP is still considerable. The transistor count of 19,600 million is high, but the die size of 272 mm² is moderate, reflecting the dense packing of the 5 nm node. The architecture supports DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, ensuring broad software compatibility. The card’s production status is active, and it was released on 2025-09-04, with a launch MSRP of 1,199 USD. The BMG-G21 chip is a testament to Intel’s commitment to the professional GPU market, offering a balanced set of features that prioritize memory capacity and compute versatility over raw gaming performance.