NVIDIA P104-100 vs NVIDIA RTX A5000 Comparison

The NVIDIA RTX A5000 and the NVIDIA P104-100 are two very different GPUs that happen to share a manufacturer. The A5000 is a workstation-class Ampere part, while the P104-100 is a Pascal-based mining card with no display outputs. Benchmark data shows a significant performance gulf between them, with the RTX A5000 winning every head-to-head test by a wide margin. The data indicates that these cards are not direct competitors, but rather serve entirely different purposes in the hardware landscape.

Head-to-Head Benchmarks

The performance disparity between the RTX A5000 and the P104-100 is stark and consistent across all three benchmark tests. The most dramatic difference appears in the Geekbench Vulkan test, where the RTX A5000 scores 137,828 against the P104-100's 45,165. This represents a 205.2% advantage for the A5000, meaning it delivers more than triple the performance of the older card in this API. The Vulkan result is particularly telling because it reflects the A5000's architectural efficiency in modern, low-level graphics workloads.

The Geekbench OpenCL benchmark tells a similar story. The RTX A5000 produces a score of 157,905, while the P104-100 manages only 52,368. The delta here is 201.5%, again showing the A5000 at roughly three times the compute performance of the P104-100. OpenCL is often used for general-purpose GPU compute tasks, and this result indicates that the A5000 is substantially better suited for non-graphics workloads, which aligns with its workstation positioning.

Even in the 3DMark Steel Nomad DX12 test, which is a modern gaming-oriented benchmark, the gap remains enormous. The RTX A5000 scores 3,783, while the P104-100 scores 1,413. The 167.7% delta here is the smallest of the three tests but still represents a massive performance advantage. This result is noteworthy because the P104-100, despite being a mining card, was based on gaming hardware; yet the newer architecture and far higher resource counts of the A5000 overpower it completely.

Looking at the broader benchmark picture, the RTX A5000's average benchmark score is 33,622, placing it in the 78th percentile of all GPUs. The P104-100's average score is 32,982, which puts it in the 77th percentile. Despite the head-to-head wins being lopsided, their aggregate scores are surprisingly close. This is because the P104-100's percentile is buoyed by a different set of comparison points; its nearest rivals include the NVIDIA T600 Mobile at 32,849 (0.4% lower) and the T550 Mobile at 33,161 (0.5% higher). The A5000's nearest rivals, meanwhile, include the GeForce GTX 1060 5 GB at 33,694 (0.2% lower) and the Radeon RX 7700S at 33,849 (0.7% lower). The average scores mask the fact that the A5000 has a much higher peak capability, as evidenced by its triple-digit wins in every direct comparison.

The Verdict

The benchmark data makes it clear that the RTX A5000 is the superior GPU in every measurable way. It wins all three head-to-head tests with deltas ranging from 167.7% to 205.2%. For any workload that relies on graphics, compute, or modern API support, the A5000 is overwhelmingly faster. The P104-100, by contrast, is a relic of the cryptocurrency mining era, and its performance profile reflects that narrow purpose.

Who should pick which? Strictly from the data, the RTX A5000 is the only choice for professional graphics work, 3D rendering, or any task requiring modern DirectX 12 Ultimate or Vulkan features. Its 24 GB memory and 768.0 GB/s bandwidth, combined with its 27.77 TFLOPS FP32 compute, make it a capable workstation tool. The P104-100, with its 4 GB memory and 6.655 TFLOPS FP32, is simply outclassed in every scenario that the benchmarks measure. Its only potential advantage would be in a scenario where its specific compute characteristics are relevant, but the data does not show any such scenario; it loses every test.

Users who already own a P104-100 and are considering an upgrade will find the RTX A5000 to be a massive step forward. The 205.2% Vulkan delta alone suggests that any modern application will run dramatically better on the A5000. However, the A5000 is a workstation card with a 230 W TDP and requires a 550 W PSU, whereas the P104-100 has no TDP listed but suggests a 200 W PSU. The power requirements are higher for the A5000, but the performance return is proportionally much larger.

FAQ

Q: Which GPU wins the most head-to-head benchmarks?

A: The NVIDIA RTX A5000 wins all three head-to-head benchmarks against the NVIDIA P104-100. The wins are in 3DMark Steel Nomad DX12, Geekbench OpenCL, and Geekbench Vulkan, with the A5000 taking 3 wins and the P104-100 taking 0.

Q: What is the largest performance gap between the two cards?

A: The largest gap is in the Geekbench Vulkan test, where the RTX A5000 scores 137,828 compared to the P104-100's 45,165. This represents a 205.2% delta in favor of the A5000.

Q: How does the RTX A5000 compare in the Geekbench OpenCL benchmark?

A: The RTX A5000 scores 157,905 in Geekbench OpenCL, while the P104-100 scores 52,368. The A5000 leads by 201.5%, indicating substantially higher compute performance for general-purpose workloads.

Q: Are the average benchmark scores of these two GPUs similar?

A: Yes, the average benchmark scores are close. The RTX A5000 has an average score of 33,622, and the P104-100 has an average score of 32,982. However, this aggregate metric hides the fact that the A5000 wins every direct head-to-head test by a wide margin.

Q: What are the nearest rivals for each GPU based on average score?

A: The RTX A5000's nearest rival is the NVIDIA GeForce GTX 1060 5 GB with an average score of 33,694, which is 0.2% higher. The P104-100's nearest rival is the NVIDIA T600 Mobile with an average score of 32,849, which is 0.4% lower.

Q: Does the P104-100 win any benchmark against the RTX A5000?

A: No. Across the three head-to-head tests listed (3DMark Steel Nomad DX12, Geekbench OpenCL, and Geekbench Vulkan), the P104-100 does not win a single test. The RTX A5000 wins all three.

Specification Differences

The two cards diverge sharply on almost every hardware specification. The RTX A5000 uses the GA102 chip built on an 8 nm process at Samsung, containing 28,300 million transistors on a 628 mm² die. The P104-100 uses the GP104 chip built on a 16 nm process at TSMC, containing 7,200 million transistors on a 314 mm² die. The A5000's transistor density is 45.1M per mm², while the P104-100 has a density of 22.9M per mm², reflecting the newer manufacturing process.

Memory configurations are completely different. The RTX A5000 has 24 GB of GDDR6 memory on a 384-bit bus, yielding 768.0 GB/s of bandwidth. The P104-100 has 4 GB of GDDR5X memory on a 256-bit bus, yielding 320.3 GB/s of bandwidth. The A5000's memory operates at 2000 MHz (16 Gbps effective), while the P104-100's runs at 1251 MHz (10 Gbps effective).

Compute resources also differ massively. The RTX A5000 has 8192 shading units, 256 TMUs, and 96 ROPs. The P104-100 has 1920 shading units, 120 TMUs, and 64 ROPs. The A5000 also includes 64 RT cores and 256 tensor cores, while the P104-100 has neither. Pixel rate on the A5000 is 162.7 GPixel/s versus 110.9 GPixel/s on the P104-100. Texture rate is 433.9 GTexel/s versus 208.0 GTexel/s.

Clock speeds tell a different story. The P104-100 has a higher base clock at 1607 MHz versus 1170 MHz on the A5000, and a slightly higher boost clock at 1733 MHz versus 1695 MHz. However, the A5000's vastly larger execution resources more than compensate for its lower clocks. The A5000's FP32 throughput is 27.77 TFLOPS versus 6.655 TFLOPS on the P104-100. The FP16 comparison is even starker: the A5000 achieves 27.77 TFLOPS (1:1 ratio), while the P104-100 manages only 104.0 GFLOPS (1:64 ratio).

The bus interface and outputs also differ. The RTX A5000 uses PCIe 4.0 x16 and has 4x DisplayPort 1.4a outputs. The P104-100 uses PCIe 1.0 x4 and has no display outputs, confirming its mining-only design. Both cards are dual-slot and use a single 8-pin power connector. The A5000 has a 230 W TDP and suggests a 550 W PSU, while the P104-100 lists no TDP but suggests a 200 W PSU. Both are 267 mm long, though the A5000 has a listed height of 112 mm while the P104-100 does not list a height.

Architecture Differences

The architecture gap between these two GPUs is generational. The RTX A5000 is based on the Ampere architecture, which is a workstation-generation part released in the Ax000 series. The P104-100 is based on the older Pascal architecture and belongs to the Mining GPUs generation. This architectural difference explains many of the performance disparities seen in the benchmarks.

The process node is a major differentiator. Ampere on the A5000 uses an 8 nm process from Samsung, while Pascal on the P104-100 uses a 16 nm process from TSMC. The smaller process allows the A5000 to pack 28,300 million transistors, nearly four times the 7,200 million in the P104-100, onto a die that is only twice as large (628 mm² versus 314 mm²). This results in a transistor density of 45.1M per mm² for the A5000 versus 22.9M per mm² for the P104-100.

Cache and core features differ as well. The RTX A5000 includes dedicated ray tracing cores (64) and tensor cores (256), which are entirely absent from the P104-100. This makes the A5000 capable of hardware-accelerated ray tracing and AI workloads, while the P104-100 can only handle traditional rasterization and compute. The A5000 also supports DirectX 12 Ultimate (12_2), while the P104-100 is limited to DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4.

The FP16 compute ratio highlights a major architectural change. The A5000 delivers FP16 at a 1:1 ratio with FP32, meaning it can process half-precision data at the same rate as full-precision. The P104-100 has a 1:64 ratio, meaning its FP16 throughput is a tiny fraction of its FP32 rate. This makes the A5000 far more suitable for machine learning inference and other half-precision workloads. The release dates reflect the generational gap: the A5000 launched in April 2021, while the P104-100 launched in December 2017. The A5000's predecessor is Quadro Turing, and its successor is Workstation Ada, while the P104-100 has no listed predecessor or successor, underscoring its status as a one-off mining product.

Where Each One Wins

The RTX A5000 wins everywhere that the benchmarks measure. In 3DMark Steel Nomad DX12, it leads by 167.7%, making it the clear choice for modern DirectX 12 gaming or rendering workloads. In Geekbench OpenCL, its 201.5% advantage makes it the better option for general-purpose compute, including tasks like physics simulation, image processing, and scientific calculations. In Geekbench Vulkan, its 205.2% lead makes it superior for Vulkan-based applications, which are common in emulation and modern game engines.

The P104-100 has no benchmark wins. However, its specification profile suggests theoretical use cases that the data does not directly measure. Its high base clock of 1607 MHz, higher than the A5000's 1170 MHz, could make it more responsive in very simple, latency-bound tasks that do not scale with core count. Its lower power suggestion of 200 W PSU versus 550 W for the A5000 means it could fit into systems with smaller power supplies. Its lack of display outputs makes it unsuitable for any interactive graphics work, but for a headless compute server, it could potentially handle lighter workloads.

Yet the data is unambiguous. The RTX A5000's nearest rivals are mainstream gaming and mobile GPUs like the GTX 1060 5 GB and RX 7700S, with average scores near 33,700. The P104-100's nearest rivals are mobile workstation parts like the T600 Mobile and RTX 3050 Mobile, with average scores near 33,000. In direct comparison, the A5000 achieves 167.7% to 205.2% higher scores. For any user prioritizing performance in the measured benchmarks, the RTX A5000 is the only rational choice. The P104-100 remains a niche product whose only advantage is its specific mining-oriented design, which is not reflected in any of the head-to-head tests.