GPU Comparison

AMD Radeon RX 590 vs NVIDIA P106-100: the data shows a clear, if narrow, victory for the AMD card in the only directly comparable benchmark, but the two GPUs are far closer in overall average performance than their architectural differences suggest. The RX 590 holds a 20.2% lead in the 3DMark Steel Nomad DX12 test, yet its average benchmark score of 24,744 is only 6.4% higher than the P106-100's 23,249. This gap places the RX 590 at the 70th percentile of all GPUs versus the P106-100's 68th, meaning both sit in the same broad performance tier.

Head-to-Head Benchmarks

The single shared benchmark, 3DMark Steel Nomad DX12, delivers an unambiguous result: the AMD Radeon RX 590 scores 1,081 points against the NVIDIA P106-100's 899 points. That is a 20.2% advantage for the RX 590, a substantial margin that reflects its higher raw compute throughput. The RX 590's FP32 performance is 7.119 TFLOPS, while the P106-100 manages only 4.375 TFLOPS, a 62.7% difference in theoretical peak floating-point output. Yet the actual benchmark delta is smaller, suggesting the P106-100's architecture extracts more real-world efficiency from each FLOP.

The RX 590 also wins on texture throughput, posting 222.5 GTexel/s versus the P106-100's 136.7 GTexel/s. That 62.8% advantage in texture fill rate aligns with the RX 590's 144 texture mapping units (TMUs) compared to 80 on the P106-100. However, the pixel rate tells a different story: the P106-100 outputs 82.03 GPixel/s, which is 65.9% higher than the RX 590's 49.44 GPixel/s. This inversion comes from the P106-100's 48 raster operation units (ROPs) versus the RX 590's 32, meaning the NVIDIA card is better equipped for fill-rate-bound work despite losing the compute-heavy DX12 test.

In broader synthetic terms, the RX 590's average benchmark score of 24,744 puts it just 0.1% behind the NVIDIA RTX A5000 Mobile (24,763) and 0.4% ahead of the Intel Arc A350M (24,647). The P106-100's 23,249 average sits exactly level with the AMD Radeon Pro Vega 16 (23,250) and only 0.1% behind the AMD Radeon RX 6600M (23,273). These nearest-rival comparisons show that both cards are clustered with mid-range mobile and workstation parts, not high-end desktop flagships.

Where Each One Wins

The RX 590 dominates in compute-oriented workloads. Its 7.119 TFLOPS FP32 output is 62.7% higher than the P106-100's 4.375 TFLOPS, and its FP16 performance is identical to FP32 at 7.119 TFLOPS (1:1 ratio). The P106-100's FP16 output is a paltry 68.36 GFLOPS (1:64 ratio), making it effectively useless for half-precision compute tasks. For any application that leverages FP16, common in AI inference and some content creation, the RX 590 is categorically superior. The RX 590 also carries 2,304 shading units against 1,280, giving it 80% more shader hardware to throw at parallel workloads.

The P106-100 counters with higher clock speeds and better pixel throughput. Its boost clock of 1,709 MHz exceeds the RX 590's 1,545 MHz by 10.6%, and its base clock of 1,506 MHz beats the RX 590's 1,469 MHz. The pixel rate advantage of 82.03 GPixel/s versus 49.44 GPixel/s means the P106-100 is better suited to resolution-scaling and fragment-heavy rendering, where ROP throughput bottlenecks performance. Its 48 ROPs versus 32 provides a 50% advantage in that specific stage of the graphics pipeline.

The P106-100 also wins on power efficiency per the data, with a 120 W TDP versus the RX 590's 175 W, that is 31.4% less power draw for 6.4% less average performance. The P106-100's suggested PSU is 300 W versus the RX 590's 450 W, and it uses a single 6-pin connector rather than the RX 590's 8-pin. For systems with limited power delivery capacity, the P106-100 is the more accommodating option. However, the P106-100 has no display outputs, making it strictly a compute or mining accelerator, while the RX 590 offers 1x HDMI 2.0b and 3x DisplayPort 1.4a.

The Verdict

Choose the AMD Radeon RX 590 if your priority is raw compute performance or you need a functional display output. The 20.2% lead in 3DMark Steel Nomad DX12 is the only direct head-to-head evidence, and it favors AMD. The RX 590's 8 GB of GDDR5 memory on a 256-bit bus with 256.0 GB/s bandwidth provides 33.3% more capacity and 33.2% more bandwidth than the P106-100's 6 GB on a 192-bit bus with 192.2 GB/s. For gaming or general GPU workloads that require rendering output, the P106-100 is disqualified outright because it cannot connect to a monitor.

Choose the NVIDIA P106-100 if you are building a dedicated compute or mining rig where display output is irrelevant and power efficiency matters. The 120 W TDP versus 175 W is a meaningful reduction, and the higher pixel rate of 82.03 GPixel/s could benefit certain fill-rate-bound tasks. The P106-100's 48 ROPs and 1,709 MHz boost clock suggest it handles rasterization-heavy workloads better, despite losing the DX12 compute benchmark. Its PCIe 1.0 x16 interface, however, is a severe limitation, the RX 590 uses PCIe 3.0 x16, which offers substantially higher host bandwidth for data transfers.

The data does not support a clear winner for all scenarios. The RX 590 wins the only shared test and has superior memory and compute specs, but the P106-100's lower power draw and higher pixel throughput make it a viable alternative in constrained environments. Both cards are end-of-life products, with the RX 590 released on 2018-11-14 and the P106-100 on 2017-06-18.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon RX 590 scores 24,744 on average, which is 6.4% higher than the NVIDIA P106-100's 23,249.

Q: What is the performance difference in the 3DMark Steel Nomad DX12 test?

A: The RX 590 scores 1,081 points versus 899 for the P106-100, a 20.2% advantage for the AMD card.

Q: Does the P106-100 have any display outputs?

A: No, the NVIDIA P106-100 has no display outputs, while the RX 590 provides 1x HDMI 2.0b and 3x DisplayPort 1.4a.

Q: How do their power requirements compare?

A: The P106-100 has a 120 W TDP with a 300 W suggested PSU and a single 6-pin connector, whereas the RX 590 has a 175 W TDP with a 450 W suggested PSU and a single 8-pin connector.

Q: Which card has more memory bandwidth?

A: The RX 590 offers 256.0 GB/s over a 256-bit bus, while the P106-100 provides 192.2 GB/s over a 192-bit bus, a 33.2% difference.

Q: What is the FP16 performance comparison?

A: The RX 590 delivers 7.119 TFLOPS FP16 (1:1 ratio), while the P106-100 manages only 68.36 GFLOPS (1:64 ratio), making the AMD card far superior for half-precision workloads.

Architecture Differences

The AMD Radeon RX 590 uses the Polaris 30 chip built on GCN 4.0 architecture, manufactured on a 12 nm process at GlobalFoundries. The NVIDIA P106-100 uses the GP106 chip with Pascal architecture, manufactured on a 16 nm process at TSMC. These process nodes differ by 4 nm, with the RX 590's smaller node contributing to its higher transistor density of 24.6M per mm² versus 22.0M per mm² for the P106-100.

Transistor counts diverge significantly: the RX 590 packs 5,700 million transistors on a 232 mm² die, while the P106-100 has 4,400 million on a 200 mm² die. That is 29.5% more transistors for the AMD chip. The RX 590's GCN 4.0 architecture emphasizes wide compute arrays with 2,304 shading units, 144 TMUs, and 32 ROPs. The P106-100's Pascal design uses 1,280 shading units, 80 TMUs, and 48 ROPs, showing NVIDIA's preference for higher clock speeds and more ROPs per shader.

Memory architectures differ as well. The RX 590 uses 8 GB of GDDR5 on a 256-bit interface, while the P106-100 uses 6 GB of GDDR5 on a 192-bit interface. Both run memory at approximately 8 Gbps effective, but the wider bus gives the RX 590 a bandwidth advantage. The P106-100's API support includes DirectX 12 (12_1) and Vulkan 1.4, while the RX 590 supports DirectX 12 (12_0) and Vulkan 1.3, the NVIDIA card has a higher DX12 feature level and newer Vulkan version.

The P106-100 is explicitly a mining GPU, as its generation label states, with no display outputs and a PCIe 1.0 x16 bus interface. The RX 590 is a standard consumer graphics card with PCIe 3.0 x16 and full display connectivity. The P106-100's FP16 ratio of 1:64 indicates a design optimized for FP32 compute, whereas the RX 590's 1:1 FP16 ratio shows a more balanced compute architecture.

Specification Differences

The two cards differ in every major specification category. The RX 590's base clock is 1,469 MHz versus 1,506 MHz for the P106-100, but the NVIDIA card boosts higher at 1,709 MHz versus 1,545 MHz. Memory capacity is 8 GB versus 6 GB, bus width is 256-bit versus 192-bit, and bandwidth is 256.0 GB/s versus 192.2 GB/s.

Shading units: 2,304 versus 1,280. TMUs: 144 versus 80. ROPs: 32 versus 48. Pixel rate: 49.44 GPixel/s versus 82.03 GPixel/s. Texture rate: 222.5 GTexel/s versus 136.7 GTexel/s. FP32: 7.119 TFLOPS versus 4.375 TFLOPS. FP16: 7.119 TFLOPS versus 68.36 GFLOPS.

TDP: 175 W versus 120 W. Power connectors: 1x 8-pin versus 1x 6-pin. Suggested PSU: 450 W versus 300 W. Bus interface: PCIe 3.0 x16 versus PCIe 1.0 x16. Display outputs: 1x HDMI 2.0b + 3x DisplayPort 1.4a versus none.

Process node: 12 nm versus 16 nm. Transistors: 5,700 million versus 4,400 million. Die size: 232 mm² versus 200 mm². Transistor density: 24.6M per mm² versus 22.0M per mm². DirectX support: 12 (12_0) versus 12 (12_1). Vulkan support: 1.3 versus 1.4.

Physical dimensions are similar, with the RX 590 at 241 mm (9.5 inches) and the P106-100 at 250 mm (9.8 inches), both dual-slot. The RX 590 has a launch MSRP of 279 USD; no launch MSRP is available for the P106-100. Release dates differ by roughly 17 months, with the RX 590 launching on 2018-11-14 and the P106-100 on 2017-06-18.