NVIDIA Quadro P6000 vs NVIDIA RTX A2000 Comparison

Head-to-Head Benchmarks

The recorded database shows a split decision between these two workstation cards. The NVIDIA Quadro P6000 takes one benchmark, the NVIDIA RTX A2000 takes the other, and the margins tell a clear story about each card's design priorities.

In Geekbench Vulkan, the Quadro P6000 posts a score of 73590 against the RTX A2000's 69089. That is a 6.5% advantage for the older Pascal card. The Vulkan result gauges graphics and compute performance through a modern low-level API, and this is the Quadro's strongest showing. It is a meaningful lead, not a trivial one, and it suggests the P6000 retains real rendering muscle despite its age.

The RTX A2000 fights back in Geekbench OpenCL, scoring 67695 versus the P6000's 66382. The delta is 1.9% in favor of the Ampere card. OpenCL is a common compute workload across professional applications, and this result shows the A2000 is essentially on par with the P6000 in that area, edging ahead by a small but measurable margin.

Each card wins exactly one head-to-head test. The database records one win for the Quadro P6000 and one win for the RTX A2000. The P6000's Vulkan lead is roughly three times larger than the A2000's OpenCL lead, which makes the P6000 the more decisive winner in its preferred benchmark. However, the A2000's OpenCL victory matters because it demonstrates that the newer architecture can match the older flagship in a widely used compute environment.

The average benchmark scores place these cards in different neighborhoods. The P6000 posts an average benchmark score of 69986, while the A2000 averages 46043. That is a substantial gap, but it is driven by the fact that the A2000 has fewer benchmark entries in the database, including a low 3DMark Steel Nomad DX12 result of 1345. That single score drags down the A2000's overall average. The head-to-head comparison, which uses only the shared Geekbench tests, is the fairer apples-to-apples view, and it shows near parity.

Percentile rankings reinforce the picture. The P6000 sits at the 90th percentile among all GPUs, while the A2000 sits at the 85th percentile. The P6000 is slightly higher in the overall distribution, which aligns with its larger lead in Vulkan and its older but enormous memory footprint.

Architecture Differences

These two cards come from different architectural eras, and the data in the database highlights the key shifts.

The Quadro P6000 uses the GP102 chip built on NVIDIA's Pascal architecture, fabricated on a 16 nm process at TSMC. It packs 11,800 million transistors into a die size of 471 mm², for a transistor density of 25.1 million per square millimeter. The RTX A2000 uses the GA106 chip with the Ampere architecture, built on an 8 nm process at Samsung. It has 12,000 million transistors in a 276 mm² die, yielding a much higher density of 43.5 million per square millimeter.

The A2000 is a smaller chip with almost the same transistor count, which explains its higher density. That density enables more modern features in a smaller physical package.

The P6000 has 3840 shading units, 240 texture mapping units, and 96 raster output pipelines. The A2000 has 3328 shading units, 104 TMUs, and 48 ROPs. The P6000 has more raw shading and texture throughput, but the A2000 introduces hardware that the P6000 lacks entirely: 26 ray tracing cores and 104 tensor cores. These are absent from the Pascal architecture, which predates NVIDIA's dedicated RT and tensor hardware.

Clock speeds also differ substantially. The P6000 runs at a base clock of 1506 MHz and boosts to 1645 MHz. The A2000 runs much lower, with a base clock of 562 MHz and a boost of 1200 MHz. The P6000's higher clocks contribute to its higher pixel rate of 157.9 GPixel/s and texture rate of 394.8 GTexel/s. The A2000 manages 57.60 GPixel/s and 124.8 GTexel/s.

The FP32 compute figures reflect the same trend. The P6000 delivers 12.63 TFLOPS of FP32, while the A2000 delivers 7.987 TFLOPS. However, the FP16 picture is completely different. The P6000 has FP16 at 197.4 GFLOPS, which is a 1:64 ratio relative to FP32, meaning it is heavily nerfed for half-precision work. The A2000 has FP16 at 7.987 TFLOPS, a 1:1 ratio, making it fully capable for FP16 compute. That is a massive architectural change between Pascal and Ampere.

Memory technology also differs. The P6000 uses 24 GB of GDDR5X on a 384-bit bus, hitting a bandwidth of 432.8 GB/s. The A2000 uses 6 GB of GDDR6 on a 192-bit bus, with 288.0 GB/s of bandwidth. The P6000 has four times the capacity and 50% more bandwidth.

The process node, foundry, memory type, bus width, and feature set all mark distinct generational steps. The P6000 is a wide, high-clock GPU from 2016. The A2000 is a dense, lower-clock GPU from 2021 with dedicated AI and ray tracing hardware.

Where Each One Wins

The data suggests a clear use-case split. For raw FP32 compute and rasterization throughput, the P6000 is the stronger card. Its FP32 figure of 12.63 TFLOPS is 58% higher than the A2000's 7.987 TFLOPS. Its pixel rate of 57.9 GP/s is 174% higher, and its texture rate is 216% higher. Any workload that relies heavily on these traditional graphics metrics will favor the P6000.

The P6000 also wins in Vulkan, which correlates with its higher fill rates and raw shader count. It is a better fit for tasks that use Vulkan for rendering or compute and that can tolerate its lower FP16 performance.

The A2000 wins in OpenCL, which points to broader compute compatibility. More importantly, the A2000 has dedicated tensor cores and RT cores. Workloads that use these features, such as AI inference or ray-traced rendering, will only run properly on the A2000. The P6000 cannot execute these features at all.

The A2000 also has a much more efficient power profile. It has a 70 W TDP versus the P6000's 250 W, and it requires no external power connector. The P6000 needs a single 8-pin connector and a 600 W suggested power supply, while the A2000 only suggests a 250 W PSU. For compact systems or multi-GPU setups, the A2000 is the only realistic choice.

The A2000 also supports PCIe 4.0 x16, while the P6000 is limited to PCIe 3.0 x16. This affects data transfer speeds for CPU-GPU communication, especially in scenarios that frequently move data across the bus.

FAQ

Q: Which card is faster in Vulkan?

A: The NVIDIA Quadro P6000 scores 73590 in Geekbench Vulkan, which is 6.5% ahead of the RTX A2000's 69089.

Q: Does the RTX A2000 win any benchmark?

A: Yes, it wins Geekbench OpenCL. The A2000 scores 67695 against the P6000's 66382, which is a 1.9% lead.

Q: Which card has more memory?

A: The Quadro P6000 has 24 GB of GDDR5X, while the RTX A2000 has 6 GB of GDDR6. The P6000 also has a wider 384-bit bus and higher bandwidth of 432.8 GB/s compared to 288.0 GB/s.

Q: Does the RTX A2000 support ray tracing?

A: Yes, the Ampere architecture includes 26 ray tracing cores and 104 tensor cores. The Pascal-based Quadro P6000 has no ray tracing or tensor cores.

Q: What are the FP16 capabilities of each card?

A: The RTX A2000 has 7.987 TFLOPS of FP16 performance, which is 1:1 with its FP32 rate. The P6000 has only 197.4 GFLOPS of FP16, a 1:64 ratio, making it far less capable for half-precision workloads.

Q: Which card has a higher overall percentile ranking?

A: The Quadro P6000 is in the 90th percentile among all GPUs, while the RTX A2000 is in the 85th percentile.

Specification Differences

| Specification | NVIDIA Quadro P6000 | NVIDIA RTX A2000 |

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

| Architecture | Pascal | Ampere |

| Process Node | 16 nm (TSMC) | 8 nm (Samsung) |

| Die Size | 471 mm² | 276 mm² |

| Transistor Density | 25.1M / mm² | 43.5M / mm² |

| Base Clock | 1506 MHz | 562 MHz |

| Boost Clock | 1645 MHz | 1200 MHz |

| Memory | 24 GB GDDR5X | 6 GB GDDR6 |

| Memory Bus | 384 bit | 192 bit |

| Bandwidth | 432.8 GB/s | 288.0 GB/s |

| Shading Units | 3840 | 3328 |

| TMUs | 240 | 104 |

| ROPs | 96 | 48 |

| RT Cores | 0 | 26 |

| Tensor Cores | 0 | 104 |

| Pixel Rate | 157.9 GPixel/s | 57.60 GPixel/s |

| Texture Rate | 394.8 GTexel/s | 124.8 GTexel/s |

| FP32 | 12.63 TFLOPS | 7.987 TFLOPS |

| FP16 | 197.4 GFLOPS (1:64) | 7.987 TFLOPS (1:1) |

| TDP | 250 W | 70 W |

| Power Connectors | 1x 8-pin | None |

| Suggested PSU | 600 W | 250 W |

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

| Display Outputs | 1x DVI, 4x DisplayPort 1.4a | 4x mini-DisplayPort 1.4a |

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

| Length | 267 mm | 167 mm |

| Height | 111 mm | 69 mm |

| Release Date | 2016-09-30 | 2021-08-09 |

| Launch MSRP | 5,999 USD | 449 USD |

The Verdict

The data supports a simple conclusion: these are cards for different jobs, and the choice depends on workload priorities.

Pick the NVIDIA Quadro P6000 if you need maximum raw rasterization throughput, the highest FP32 (12.63 TFLOPS), and a huge 24 GB frame buffer with 432.8 GB/s of bandwidth. It wins the Vulkan benchmark by a 6.5% margin, and its pixel and texture rates are far above the A2000. It is also the only one with a DVI output, which may matter for legacy display compatibility. The P6000 sits in the 90th percentile of all GPUs, higher than the A2000's 85th.

Pick the NVIDIA RTX A2000 if you need modern features: ray tracing cores, tensor cores, and full-rate FP16. It wins the OpenCL benchmark, and it is far more efficient at 70 W with no power connector. It is a smaller card, at 167 mm long versus the P6000's 267 mm, and it supports PCIe 4.0. The A2000 also has a much lower launch MSRP (449 USD) which reflects its lower absolute performance, but it delivers features the P6000 cannot offer.

The verdict is clear. For pure legacy compute and high-end rasterization with massive memory, the Quadro P6000 is the stronger choice. For modern workstations that require ray tracing, AI acceleration, or FP16 compute, the RTX A2000 is the only one of the two that can do the job. The 1.9 % OpenCL win for the A2000 shows that the newer architecture has closed a significant part of the performance gap, but the P6000's advantages in raw throughput and memory capacity are still decisive in its own workload domains.