Intel Arc A580 vs NVIDIA Quadro P6000 Comparison

The NVIDIA Quadro P6000 and Intel Arc A580 represent two very different generations of GPU design, separated by seven years of architectural evolution. The database records only two shared benchmark tests between them, and the Intel part wins both, but the full specification comparison reveals a more complex story than simple victory or defeat. The Quadro P6000 is a professional workstation card from 2016, built on the Pascal architecture with a focus on compute and rendering stability. The Arc A580 is a 2023 consumer-oriented GPU from Intel’s Alchemist generation, aiming at mainstream gaming and modern API support. Their average benchmark scores, percentile rankings, and architectural details show that each card has distinct strengths that matter depending on the workload.

Head-to-Head Benchmarks

The recorded head-to-head data covers two tests: Geekbench OpenCL and Geekbench Vulkan. In the OpenCL test, the Intel Arc A580 scores 91,657, while the NVIDIA Quadro P6000 scores 66,382. That gives Intel a 27.6% lead. This is a substantial margin, and it reflects the Arc A580’s higher memory bandwidth and newer architecture. The Quadro P6000’s OpenCL result is closer to its nearest rivals in the database: the AMD Radeon Pro WX 8200 sits at 69,870, just 0.2% ahead of the Quadro, and the NVIDIA RTX A3000 Mobile is at 70,140, which is 0.2% behind. The Intel card, by contrast, sits in a different performance class for this workload, outpacing the Quadro by a margin that none of the Quadro’s nearest rivals come close to matching.

In the Vulkan test, the gap narrows considerably. The Arc A580 scores 79,381, and the Quadro P6000 scores 73,590. The Intel card wins again, but only by 7.3%. This suggests that the Vulkan API reduces the architectural advantage of the newer Intel GPU, possibly due to the Quadro’s mature driver stack or the specific nature of the test workload. The Quadro’s Vulkan score is also much higher than its OpenCL score relative to the Intel card, indicating that the Pascal architecture handles Vulkan more efficiently than OpenCL when compared to Xe-HPG.

Looking at the broader database context, the Quadro P6000 has an average benchmark score of 69,986 across all recorded tests. Its nearest rivals are tightly clustered: the AMD Radeon Pro WX 8200 is 0.2% ahead, the NVIDIA RTX A3000 Mobile is 0.2% behind, the AMD Radeon RX 6600 LE is 1.2% behind, and the NVIDIA CMP 90HX is 1.4% ahead. The Arc A580 has an average score of 57,756, which places it in a different competitive tier. Its nearest rivals include the AMD Radeon RX 5600 OEM at 0.6% behind, the AMD Radeon RX 9070 GRE at 0.7% ahead, the Intel Arc A570M at 0.8% behind, and the AMD Radeon RX 6950 XT at 1.1% behind.

The average benchmark score tells a different story than the head-to-head results. The Quadro P6000’s average of 69,986 is higher than the Arc A580’s 57,756, despite losing both shared tests. This is because the database includes a wider range of benchmarks for each card, and the Quadro’s professional-oriented drivers and compute capabilities boost its overall standing. The Arc A580 has a 3DMark Steel Nomad DX12 score of 2,229, a test that the Quadro has no recorded result for, and this lower score drags down its average. The percentile ranking reflects this split: the Quadro P6000 sits at the 90th percentile of all GPUs in the database, while the Arc A580 sits at the 87th percentile. The Quadro is, on average, a faster card across the full database, even though the Arc wins the two tests where both have recorded data.

Where Each One Wins

The Intel Arc A580 wins in raw OpenCL compute performance. Its 91,657 OpenCL score is 27.6% higher than the Quadro’s, which is a decisive margin for any workload that relies heavily on general-purpose GPU compute. This includes tasks like physics simulations, video encoding, and certain machine learning inference workloads that use OpenCL. The Arc also wins in Vulkan, with a 7.3% lead, making it the better choice for modern games and applications that use Vulkan as their primary rendering API. The Arc’s support for DirectX 12 Ultimate (12_2) and its 24 ray tracing cores give it a feature set that the Quadro simply cannot match, since the Quadro only supports DirectX 12 (12_1) and has no dedicated ray tracing hardware.

The Quadro P6000 wins in overall database average score, sitting at 69,986 compared to the Arc’s 57,756. This suggests that in the broader mix of recorded benchmarks, which likely includes professional applications, OpenGL workloads, and older API tests, the Quadro holds its own and even surpasses the Intel card. The Quadro’s 24 GB of GDDR5X memory is a massive advantage for large datasets, rendering scenes with high-resolution textures, and compute tasks that need to keep more data resident on the GPU. The Arc’s 8 GB of GDDR6 memory is half the capacity, even though its bandwidth is higher at 512.0 GB/s compared to the Quadro’s 432.8 GB/s. For workloads that exceed 8 GB of VRAM, the Quadro is the only viable option between these two.

The Quadro also wins on memory capacity and ecosystem maturity. Its Pascal architecture has been in the field since 2016, and its production status is marked as end-of-life, but its driver support for professional applications is a known quantity. The Arc A580 is active in production, but its architecture is newer and its driver maturity is less proven in professional contexts. The Quadro’s predecessor is Quadro Maxwell and its successor is Quadro Volta, placing it in a long line of professional cards. The Arc’s predecessor is Xe Graphics, its successor is Battlemage, and it represents Intel’s first serious push into discrete GPUs in over two decades.

Architecture Differences

The NVIDIA Quadro P6000 uses the GP102 chip, built on the Pascal architecture, manufactured on a 16 nm process at TSMC. It packs 11,800 million transistors into a die size of 471 mm², giving it a transistor density of 25.1 million transistors per square millimeter. The Intel Arc A580 uses the DG2-512 chip, built on the Xe-HPG architecture, manufactured on a 6 nm process, also at TSMC. It contains 21,700 million transistors in a 406 mm² die, resulting in a transistor density of 53.4 million transistors per square millimeter. The Arc’s density is more than double the Quadro’s, which is a direct result of the smaller process node and the seven years of manufacturing improvements between them.

The Quadro’s compute configuration consists of 3,840 shading units, 240 texture mapping units, and 96 raster output units. The Arc A580 has 3,072 shading units, 192 texture mapping units, and 96 raster output units. The Quadro has more shading units and TMUs, but the same number of ROPs. The Arc adds 24 ray tracing cores, which the Quadro lacks entirely. This is the most significant architectural feature difference: the Pascal architecture predates dedicated ray tracing hardware, while Xe-HPG includes it as a core feature. The Quadro’s FP32 throughput is 12.63 TFLOPS, and its FP16 throughput is 197.4 GFLOPS at a 1:64 ratio. The Arc’s FP32 throughput is 12.29 TFLOPS, slightly lower, but its FP16 throughput is 24.58 TFLOPS at a 2:1 ratio. That is a massive difference in half-precision compute, with the Arc offering more than 100 times the FP16 performance of the Quadro.

Clock speeds also differ. The Quadro runs at a base clock of 1506 MHz and a boost clock of 1645 MHz, with memory clocked at 1127 MHz, which translates to 9 Gbps effective. The Arc runs at a base clock of 1700 MHz and a boost clock of 2000 MHz, with memory at 2000 MHz, or 16 Gbps effective. The Arc’s higher clocks contribute to its pixel rate of 192.0 GPixel/s, compared to the Quadro’s 157.9 GPixel/s. The texture rates are closer: the Arc produces 384.0 GTexel/s, and the Quadro produces 394.8 GTexel/s, with the Quadro holding a slight edge despite its lower clocks.

The memory subsystems are fundamentally different. The Quadro uses 24 GB of GDDR5X on a 384-bit bus, yielding 432.8 GB/s of bandwidth. The Arc uses 8 GB of GDDR6 on a 256-bit bus, yielding 512.0 GB/s of bandwidth. The Arc has higher bandwidth but far less capacity. The Quadro’s power draw is 250 W, and it requires a single 8-pin power connector with a suggested 600 W power supply. The Arc draws 175 W, needs two 8-pin connectors, and has a suggested 450 W power supply. Both cards are dual-slot designs. The Quadro uses a PCIe 3.0 x16 interface, while the Arc uses PCIe 4.0 x16, doubling the available bandwidth to the host system. The Quadro’s display outputs are 1x DVI and 4x DisplayPort 1.4a. The Arc offers 1x HDMI 2.1 and 3x DisplayPort 2.0, which are newer standards with higher bandwidth. The Quadro measures 267 mm in length and 111 mm in height, while the Arc’s dimensions are not recorded in the database.

Both cards support DirectX 12, but at different feature levels. The Quadro supports DirectX 12 (12_1), while the Arc supports DirectX 12 Ultimate (12_2). Both support OpenGL 4.6 and Vulkan 1.4. The Quadro’s API support is the older feature level, which means it cannot use certain modern rendering techniques like mesh shaders or variable rate shading that require DirectX 12 Ultimate. The Arc’s support for these features, combined with its ray tracing cores, gives it a clear advantage for modern game engines.

The Verdict

The data points to a clear split in use cases. If the workload is modern gaming, especially with ray tracing or DirectX 12 Ultimate features, the Intel Arc A580 is the correct choice. It wins both shared benchmarks, has dedicated ray tracing hardware, supports newer display outputs, and offers significantly higher FP16 compute. Its higher memory bandwidth also helps in scenarios where the data fits within 8 GB. The Arc wins in OpenCL by 27.6% and Vulkan by 7.3%, making it the faster card in the two tests where direct comparison is possible.

If the workload is professional content creation, large-scale rendering, or compute tasks that need more than 8 GB of memory, the Quadro P6000 is the better option. Its 24 GB of GDDR5X memory is three times the capacity of the Arc, and its higher average benchmark score of 69,986 versus 57,756 indicates stronger overall performance across the wider database. The Quadro’s 90th percentile ranking, compared to the Arc’s 87th, reinforces this. The Quadro also has a higher texture rate at 394.8 GTexel/s, and its FP32 throughput is marginally higher at 12.63 TFLOPS versus 12.29 TFLOPS, which matters for single-precision compute tasks.

Neither card is a universal winner. The Quadro is end-of-life, while the Arc is active in production. The Quadro launched with an MSRP of 5,999 USD, which reflects its professional positioning. The Arc has no recorded launch MSRP. The Quadro’s power draw is higher at 250 W, but it requires only one 8-pin connector, while the Arc needs two. The Arc is more power-efficient per unit of performance, with a lower TDP of 175 W and a smaller suggested power supply of 450 W versus 600 W. The Arc’s PCIe 4.0 interface is a generation ahead of the Quadro’s PCIe 3.0, which matters for data transfer in bandwidth-sensitive applications.

FAQ

Q: Which GPU has better OpenCL performance?

A: The Intel Arc A580 scores 91,657 in Geekbench OpenCL, which is 27.6% higher than the NVIDIA Quadro P6000’s 66,382.

Q: Does the Quadro P6000 support ray tracing?

A: No, the Quadro P6000 has no ray tracing cores. The Intel Arc A580 includes 24 ray tracing cores.

Q: Which card has more memory bandwidth?

A: The Intel Arc A580 has 512.0 GB/s of bandwidth, while the NVIDIA Quadro P6000 has 432.8 GB/s.

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA Quadro P6000 has an average benchmark score of 69,986, compared to the Intel Arc A580’s 57,756.

Q: What is the difference in FP16 performance?

A: The Intel Arc A580 delivers 24.58 TFLOPS of FP16 performance at a 2:1 ratio, while the Quadro P6000 delivers 197.4 GFLOPS at a 1:64 ratio.

Q: Which card uses less power?

A: The Intel Arc A580 has a TDP of 175 W and a suggested 450 W power supply, while the Quadro P6000 has a TDP of 250 W and a suggested 600 W power supply.

Specification Differences

| Specification | NVIDIA Quadro P6000 | Intel Arc A580 |

| --- | --- | --- |

| Chip | GP102 | DG2-512 |

| Architecture | Pascal | Xe-HPG |

| Process Node | 16 nm | 6 nm |

| Transistors | 11,800 million | 21,700 million |

| Die Size | 471 mm² | 406 mm² |

| Transistor Density | 25.1M / mm² | 53.4M / mm² |

| Base Clock | 1506 MHz | 1700 MHz |

| Boost Clock | 1645 MHz | 2000 MHz |

| Memory Clock | 1127 MHz, 9 Gbps effective | 2000 MHz, 16 Gbps effective |

| Memory Size | 24 GB | 8 GB |

| Memory Type | GDDR5X | GDDR6 |

| Memory Bus Width | 384 bit | 256 bit |

| Memory Bandwidth | 432.8 GB/s | 512.0 GB/s |

| Shading Units | 3840 | 3072 |

| TMUs | 240 | 192 |

| ROPs | 96 | 96 |

| Ray Tracing Cores | None | 24 |

| Pixel Rate | 157.9 GPixel/s | 192.0 GPixel/s |

| Texture Rate | 394.8 GTexel/s | 384.0 GTexel/s |

| FP32 | 12.63 TFLOPS | 12.29 TFLOPS |

| FP16 | 197.4 GFLOPS (1:64) | 24.58 TFLOPS (2:1) |

| TDP | 250 W | 175 W |

| Power Connectors | 1x 8-pin | 2x 8-pin |

| Suggested PSU | 600 W | 450 W |

| Bus Interface | PCIe 3.0 x16 | PCIe 4.0 x16 |

| Display Outputs | 1x DVI, 4x DisplayPort 1.4a | 1x HDMI 2.1, 3x DisplayPort 2.0 |

| DirectX | 12 (12_1) | 12 Ultimate (12_2) |

| Production Status | End-of-life | Active |

| Release Date | 2016-09-30 | 2023-10-09 |

| Predecessor | Quadro Maxwell | Xe Graphics |

| Successor | Quadro Volta | Battlemage |

| Launch MSRP | 5,999 USD | Not recorded |