GPU Comparison

Intel HD Graphics P4600 and Intel Arc Pro B60 occupy opposite ends of Intel's graphics spectrum, separated by over a decade of silicon evolution. The P4600 is a 2013 integrated graphics solution embedded in Haswell processors, while the Arc Pro B60 is a 2025 discrete workstation card built on the Xe2-HPG architecture. The data reveals a generational chasm in raw compute, memory bandwidth, and feature support, yet both share a curious similarity in their overall benchmark percentile ranking.

Head-to-Head Benchmarks

Direct head-to-head benchmark results between the two are unavailable, so the comparison rests on their respective benchmark suites. The P4600 posts a single Geekbench OpenCL score of 3389, placing it in the 20th percentile of all GPUs. Its nearest rivals include the NVIDIA GeForce GT 740 (avg score 3431, delta -1.2%), the NVIDIA Quadro 2000M (avg score 3434, delta -1.3%), the Intel HD Graphics 530 (avg score 3332, delta 1.7%), and the NVIDIA GeForce GT 730M (avg score 3316, delta 2.2%). This places the P4600 within a tight 2.2% band of its competitors, indicating it performs essentially on par with entry-level discrete GPUs from its era, despite being an integrated part.

The Arc Pro B60's benchmark profile is far more extensive, featuring eight distinct tests. Its Passmark G3D score of 14580 is the standout, dwarfing its Passmark DirectX 9 score of 179, DirectX 10 score of 61, DirectX 11 score of 122, and DirectX 12 score of 76. The GPU also achieves a Passmark G2D score of 763 and a Passmark GPU Compute score of 7029. Its 3DMark Steel Nomad DX12 score is 2646. Interestingly, the Arc Pro B60's average benchmark score is 3182, which is actually lower than the P4600's single Geekbench score of 3389, yet this is misleading due to the different test methodologies and the Arc Pro B60's low DirectX legacy scores dragging down its average.

The Arc Pro B60's nearest rivals paint a different picture. It sits at 0.6% ahead of the NVIDIA Quadro P1000 (avg score 3163), 0.9% behind the NVIDIA GeForce GT 640 (avg score 3210), 3.2% behind the NVIDIA GeForce 920M (avg score 3287), and 3.5% ahead of the NVIDIA GeForce RTX 5080 SUPER (avg score 3075). The presence of the RTX 5080 SUPER in this list is striking, suggesting that the Arc Pro B60's average score places it in a strange performance tier that mixes very old and very new hardware, likely due to the weighting of its legacy DirectX scores. Both GPUs share the same 20th percentile ranking, but the Arc Pro B60's raw compute numbers are orders of magnitude higher.

Architecture Differences

The architectural gulf between these two is immense. The P4600 is built on Intel's 22 nm process with the Haswell GT2 chip, part of the Generation 7.5 architecture. The Arc Pro B60 uses a 5 nm process from TSMC, featuring the BMG-G21 chip based on the Xe2-HPG architecture, and packs 19,600 million transistors on a 272 mm² die with a transistor density of 72.1M / mm². The P4600 has no listed transistor count or die size, underscoring its simpler integrated design.

Shading resources differ by an order of magnitude. The P4600 has 160 shading units, 20 texture mapping units (TMUs), and just 2 render output units (ROPs). The Arc Pro B60 has 2560 shading units, 160 TMUs, and 80 ROPs. The Arc Pro B60 also introduces 20 ray tracing cores, a feature completely absent from the older P4600. Pixel rate jumps from 2.400 GPixel/s on the P4600 to 192.0 GPixel/s on the Arc Pro B60, while texture rate leaps from 24.00 GTexel/s to 384.0 GTexel/s. FP32 compute goes from 384.0 GFLOPS on the P4600 to 12.29 TFLOPS on the Arc Pro B60, with the latter also offering FP16 at 24.58 TFLOPS (2:1).

Clock speeds reflect the different design philosophies. The P4600 runs at a base of 350 MHz with a 1200 MHz boost, typical for a power-efficient integrated GPU. The Arc Pro B60 runs at a base of 2000 MHz and boosts to 2400 MHz, with memory clocked at 2375 MHz (19 Gbps effective). The P4600 uses system-shared memory with system-dependent bandwidth, while the Arc Pro B60 has 24 GB of dedicated GDDR6 on a 192-bit bus, delivering 456.0 GB/s of bandwidth. The API support also diverges: the P4600 supports DirectX 12 (11_1), OpenGL 4.3, and Vulkan 1.0, while the Arc Pro B60 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

FAQ

Q: How do the two GPUs compare in raw compute performance?

A: The Arc Pro B60 delivers 12.29 TFLOPS of FP32 compute, which is 32 times the P4600's 384.0 GFLOPS. The Arc Pro B60 also supports FP16 at 24.58 TFLOPS, a capability the P4600 lacks entirely.

Q: What is the difference in their benchmark percentile rankings?

A: Both GPUs sit at the 20th percentile of all GPUs, despite the Arc Pro B60 having vastly superior raw specifications. This is because the Arc Pro B60's average score of 3182 is dragged down by low legacy DirectX scores (61 in DX10, 76 in DX12), while the P4600's single Geekbench score of 3389 is relatively higher.

Q: How does memory configuration differ between the two?

A: The P4600 uses system-shared memory with system-dependent bandwidth, meaning it borrows from the main system RAM. The Arc Pro B60 has 24 GB of dedicated GDDR6 memory on a 192-bit bus, providing 456.0 GB/s of bandwidth.

Q: What are the power and physical differences?

A: The P4600 has a TDP of 84 W and is an integrated GPU (IGP) with no slot width, relying on the motherboard for display output. The Arc Pro B60 has a 200 W TDP, is a dual-slot card measuring 167 mm x 69 mm x 40 mm, requires a 1x 8-pin power connector, and suggests a 550 W PSU.

Q: Which GPU has better API support?

A: The Arc Pro B60 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The P4600 only supports DirectX 12 (11_1), OpenGL 4.3, and Vulkan 1.0, making the Arc Pro B60 more future-proof for modern applications.

Q: What is the production status and release timeline?

A: The P4600 is end-of-life and was released on 2013-05-31. The Arc Pro B60 is active and was released on 2025-09-04, with a launch MSRP of 499 USD.

Specification Differences

The two GPUs differ in nearly every measurable specification. The process node jumps from 22 nm (Intel) to 5 nm (TSMC), and the Arc Pro B60 has a transistor count of 19,600 million on a 272 mm² die, while the P4600 has no listed transistor data. Clock speeds differ significantly: the P4600 runs at 350 MHz base / 1200 MHz boost, while the Arc Pro B60 runs at 2000 MHz base / 2400 MHz boost. Memory is the most dramatic difference: the P4600 uses system-shared memory with system-dependent bandwidth, while the Arc Pro B60 has 24 GB of GDDR6 on a 192-bit bus with 456.0 GB/s bandwidth.

The compute units scale up: shading units go from 160 to 2560, TMUs from 20 to 160, and ROPs from 2 to 80. The Arc Pro B60 adds 20 ray tracing cores, which the P4600 lacks entirely. Pixel rate multiplies from 2.400 GPixel/s to 192.0 GPixel/s, texture rate from 24.00 GTexel/s to 384.0 GTexel/s, and FP32 from 384.0 GFLOPS to 12.29 TFLOPS. The Arc Pro B60 also adds FP16 at 24.58 TFLOPS. TDP rises from 84 W to 200 W, and the form factor changes from an IGP to a dual-slot card with a 1x 8-pin power connector and a 550 W suggested PSU. The bus interface changes from Ring Bus to PCIe 5.0 x8, and display outputs go from motherboard-dependent to 4x mini-DisplayPort 2.1. API support improves across the board, and the production status shifts from end-of-life to active.

The Verdict

The data makes the choice clear for most use cases. The Arc Pro B60 is the superior GPU in virtually every raw performance metric, delivering 32 times the FP32 compute, 80 times the pixel rate, and 16 times the texture rate compared to the P4600. Its dedicated 24 GB of GDDR6 memory with 456.0 GB/s bandwidth eliminates the system-shared bottleneck that plagues the P4600. The Arc Pro B60 also brings modern API support, ray tracing cores, and a much higher 200 W power envelope to sustain its performance.

The P4600, however, is not without its niche. Its 84 W TDP and integrated design mean it requires no additional power connectors and draws its power from the motherboard, making it suitable for legacy systems where a discrete GPU is not an option. Its single Geekbench score of 3389 is comparable to the Arc Pro B60's average of 3182, though this is an artifact of the Arc Pro B60's weak legacy DirectX performance rather than a real equivalence. The P4600 is end-of-life and offers no path forward for modern workloads, while the Arc Pro B60 is active and built for current and future professional applications.

Where Each One Wins

The Arc Pro B60 wins decisively in compute-heavy workloads. Its 12.29 TFLOPS FP32 and 24.58 TFLOPS FP16 performance, combined with 20 ray tracing cores, make it suited for rendering, simulation, and AI-adjacent tasks that can leverage its parallel architecture. The Passmark G3D score of 14580 and Passmark GPU Compute score of 7029 indicate strong performance in 3D rendering and general-purpose compute. Its 24 GB of GDDR6 memory with 456.0 GB/s bandwidth allows it to handle large datasets and textures that the P4600's system-shared memory cannot accommodate. The Arc Pro B60's modern API support (DirectX 12 Ultimate, Vulkan 1.4) also means it can run the latest graphics features, whereas the P4600 is stuck with DirectX 12 (11_1) and Vulkan 1.0.

The P4600 wins in power efficiency per watt of thermal design. Its 84 W TDP is less than half of the Arc Pro B60's 200 W, and since it is an integrated GPU, it requires no additional cooling or power delivery infrastructure. The P4600's single Geekbench score of 3389, while low in absolute terms, is comparable to the Arc Pro B60's average of 3182, suggesting that in legacy OpenCL workloads, the integrated GPU holds its own against the newer discrete card. The P4600 also wins in physical footprint, as an IGP takes zero slot space, while the Arc Pro B60 is a dual-slot card measuring 167 mm in length. For users with legacy Haswell systems who cannot upgrade their platform, the P4600 remains a functional, if limited, graphics solution. For anyone needing modern performance, the Arc Pro B60 is the only choice, despite its heavier power draw and physical size.