Intel HD Graphics P4600 vs NVIDIA GeForce GTX 860M Comparison

The benchmark data is unequivocal: the NVIDIA GeForce GTX 860M is in a completely different performance class than the Intel HD Graphics P4600. In the only shared benchmark, the disparity is massive, with the discrete mobile GPU delivering roughly three times the compute performance of the integrated Intel solution. This is not a close contest; it is a demonstration of the architectural gulf between a dedicated graphics processor and a CPU-integrated graphics core.

Head-to-Head Benchmarks

The sole direct comparison available is the Geekbench OpenCL test, and it is a decisive victory for the NVIDIA GeForce GTX 860M. The NVIDIA part scores 10461 points, while the Intel HD Graphics P4600 manages only 3389 points. This translates to a 67.6% deficit for the Intel solution, meaning the GTX 860M is more than three times faster in this compute-oriented workload.

This result is consistent with the raw specifications of each chip. The GTX 860M is built on a discrete architecture with 1152 shading units, 96 texture mapping units (TMUs), and 16 raster output units (ROPs). Its peak FP32 throughput is 2.108 TFLOPS with a texture fill rate of 87.84 GTexel/s and a pixel rate of 21.96 GPixel/s. In stark contrast, the Intel HD Graphics P4600 is a minimalist integrated GPU with only 160 shading units, 20 TMUs, and a mere 2 ROPs, resulting in a peak FP32 of 384.0 GFLOPS, a texture rate of 24.00 GTexel/s, and a pixel rate of 2.400 GPixel/s.

The gap in pixel throughput is particularly illustrative: the GTX 860M can fill pixels at a rate roughly nine times higher than the P4600, which directly impacts resolution and detail settings in rendering tasks. The GTX 860M's advantage in texture rate is also substantial, at over 3.5 times the P4600's capability. These architectural numbers underpin the benchmark result, confirming that the OpenCL score is not an anomaly but a reflection of the fundamental processing power available in each chip. The Intel P4600’s average score of 3389 places it near the NVIDIA GeForce GT 740 (3431, a -1.2% delta) and the NVIDIA Quadro 2000M (3434, a -1.3% delta), while the GTX 860M’s average of 2967 puts it in a different league, closer to the NVIDIA GeForce GTX 750 Ti (2953, a 0.5% delta).

Where Each One Wins

The performance split is stark and leaves little room for ambiguity.

NVIDIA GeForce GTX 860M: This GPU wins in every measurable aspect of performance. In the Geekbench OpenCL test, it is the clear victor. Its dedicated memory subsystem, featuring 2 GB of GDDR5 on a 128-bit bus with 80.00 GB/s of bandwidth, provides a massive advantage over the P4600's "System Shared" memory, which has "System Dependent" bandwidth. This dedicated high-speed memory allows the GTX 860M to feed its 1152 shading units effectively, making it suitable for any task that leverages GPU compute, from gaming to content creation. Its memory clock of 1250 MHz (or 5 Gbps effective) and base/boost clocks of 797 MHz / 915 MHz are all far higher than the P4600's 350 MHz base and 1200 MHz boost, further solidifying its performance lead.

Intel HD Graphics P4600: The Intel solution's only advantage is in its integration and power profile, though the data on power is nuanced. While the GTX 860M has a lower TDP at 75 W, the P4600's TDP is listed at 84 W; however, this is a system-level figure for the CPU package it's part of, not the GPU alone. The P4600's win is purely logistical: it requires no separate memory, uses the "Ring Bus" interface, and is "Motherboard Dependent" for display outputs, making it a simple, low-cost addition to a workstation CPU. It wins on the basis of simplicity, not performance. For any benchmark that measures raw graphics or compute output, the GTX 860M is the definitive choice.

The Verdict

The data dictates a simple conclusion: choose the NVIDIA GeForce GTX 860M if you need any meaningful graphics or compute performance. The 67.6% lead in the only shared benchmark is a chasm that no amount of driver optimization can bridge. The GTX 860M's dedicated 2 GB GDDR5 memory and 2.108 TFLOPS of compute power make it a capable solution for gaming and GPU-accelerated applications, placing it in the 19th percentile of all GPUs.

The Intel HD Graphics P4600, sitting in the 20th percentile, is not a competitor in this space. It is a display-out solution. Its performance is comparable to entry-level discrete GPUs from a previous generation, like the NVIDIA GeForce GT 730M (3316 average score, a 2.2% delta in the P4600's favor), but it lacks the dedicated resources to handle modern workloads. Choose the P4600 only when its performance is irrelevant, such as in a server or basic workstation where the CPU's integrated graphics are sufficient for 2D desktop output and the system requires no add-in card. For any task where the GPU is expected to work, the GTX 860M is the only logical choice based on this data.

FAQ

Q: Which GPU is faster in the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce GTX 860M is significantly faster, scoring 10461 points compared to the Intel HD Graphics P4600's 3389 points.

Q: By what percentage is the GTX 860M faster than the P4600 in the shared benchmark?

A: The GTX 860M outperforms the P4600 by 67.6% in the Geekbench OpenCL test.

Q: Does the Intel HD Graphics P4600 have its own dedicated video memory?

A: No, its memory size, type, and bus width are all listed as "System Shared," meaning it relies on the system's main memory.

Q: What is the memory bandwidth of the NVIDIA GeForce GTX 860M?

A: The GTX 860M has a memory bandwidth of 80.00 GB/s from its 2 GB of GDDR5 memory on a 128-bit bus.

Q: How many shading units does each GPU have?

A: The NVIDIA GeForce GTX 860M has 1152 shading units, while the Intel HD Graphics P4600 has only 160.

Q: What is the FP32 compute performance of each GPU?

A: The GTX 860M delivers 2.108 TFLOPS, whereas the Intel HD Graphics P4600 delivers 384.0 GFLOPS.

Architecture Differences

The two GPUs represent fundamentally different design philosophies and hardware generations. The Intel HD Graphics P4600 is an integrated graphics processor, part of the Haswell GT2 chip, built on Intel's Generation 7.5 architecture. It is fabricated on a 22 nm process at Intel's own foundry. Its design is minimal, with 160 shading units, and its output is routed through a "Ring Bus" interface. The GPU relies entirely on system memory, with bandwidth being "System Dependent," and its display outputs are "Motherboard Dependent." It supports DirectX 12 (11_1), OpenGL 4.3, and Vulkan 1.0.

Conversely, the NVIDIA GeForce GTX 860M is a discrete mobile GPU based on the GK104 chip, using the Kepler architecture from the GeForce 800M generation. It is built on a 28 nm process at TSMC, and it is a much more complex chip, containing 3,540 million transistors on a 294 mm² die with a transistor density of 12.0M / mm². It connects to the rest of the system via a PCIe 3.0 x16 interface and has its own dedicated memory subsystem. The GTX 860M supports a newer set of APIs, including DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The GTX 860M also has a predecessor, the GeForce 700M, and a successor, the GeForce 900M, while the Intel part has no listed predecessor or successor.

Specification Differences

The specification sheet reveals a total mismatch in core resources and memory architecture.

| Specification | Intel HD Graphics P4600 | NVIDIA GeForce GTX 860M |

| :--- | :--- | :--- |

| Process Node | 22 nm | 28 nm |

| Foundry | Intel | TSMC |

| Transistors | Not listed | 3,540 million |

| Die Size | Not listed | 294 mm² |

| Base Clock | 350 MHz | 797 MHz |

| Boost Clock | 1200 MHz | 915 MHz |

| Memory Size | System Shared | 2 GB |

| Memory Type | System Shared | GDDR5 |

| Memory Bus Width | System Shared | 128 bit |

| Memory Bandwidth | System Dependent | 80.00 GB/s |

| Shading Units | 160 | 1152 |

| TMUs | 20 | 96 |

| ROPs | 2 | 16 |

| Pixel Rate | 2.400 GPixel/s | 21.96 GPixel/s |

| Texture Rate | 24.00 GTexel/s | 87.84 GTexel/s |

| FP32 Performance | 384.0 GFLOPS | 2.108 TFLOPS |

| TDP | 84 W | 75 W |

| Bus Interface | Ring Bus | PCIe 3.0 x16 |

| Power Connectors | Not listed | None |

| Display Outputs | Motherboard Dependent | Portable Device Dependent |

| DirectX Support | 12 (11_1) | 12 (11_0) |

| OpenGL Support | 4.3 | 4.6 |

| Vulkan Support | 1.0 | 1.2.175 |

The most critical differences are the memory and the execution resources. The GTX 860M's dedicated 2 GB GDDR5 with 80.00 GB/s bandwidth is a stark contrast to the P4600's shared memory. Similarly, the 7.2x difference in shading units and 8x difference in ROPs directly explain the massive performance gap seen in the benchmark results. The GTX 860M has a lower TDP (75 W) than the P4600 (84 W), but this is likely a full-package figure for the Intel CPU, making the dedicated NVIDIA GPU remarkably power-efficient for its performance level.