AMD Radeon Pro W5700 vs NVIDIA P106-100 Comparison

The AMD Radeon Pro W5700 and the NVIDIA P106-100 occupy very different corners of the GPU market, and the recorded data reflects that split. The W5700 is a professional workstation card built on AMD's RDNA 1.0 architecture, while the P106-100 belongs to NVIDIA's mining-focused generation, shipping without display outputs and targeting compute workloads exclusively. In the two head-to-head benchmarks recorded in the database, the W5700 wins both, doubling the P106-100's scores in each case. Yet the P106-100 holds a percentileVsAllGpus of 68 against the W5700's 71, so the gap in overall standing is narrower than the raw compute results suggest.

Head-to-Head Benchmarks

Only two tests exist for both cards, and both are decisive.

In Geekbench OpenCL, the Radeon Pro W5700 scores 74,613 against the P106-100's 35,951. That is a 107.5 percent margin, meaning the W5700 delivers more than double the OpenCL compute throughput. Geekbench Vulkan tells the same story with an even wider spread: 70,706 versus 32,897, a 114.9 percent lead for the AMD card.

For context on what these scores mean in the broader database: the W5700's average benchmark score is 25,726, placing it in the 71st percentile of all GPUs. Its nearest rivals by average score include the NVIDIA GeForce RTX 3080 Ti Mobile (average 25,740, just 0.1 percent higher), the AMD Radeon RX 6700M (25,633, 0.4 percent lower), the AMD FirePro D700 (25,842, 0.4 percent higher), and the AMD FirePro W7100 (25,856, 0.5 percent higher). In other words, the W5700 performs in the same band as a modern high-end mobile GeForce part and AMD's own workstation predecessors.

The P106-100 averages 23,249 and sits in the 68th percentile. Its closest rivals are the AMD Radeon Pro Vega 16 (23,250, effectively tied), the Radeon RX 6600M (23,273, 0.1 percent higher), the R9 M290X (23,276, 0.1 percent higher), and the Radeon AI PRO R9700 (23,315, 0.3 percent higher). Three percentile points separate the two cards in overall standing, but the head-to-head compute results show the W5700 is far stronger when the same workload runs on both.

The W5700's full benchmark sheet reinforces its compute positioning: Geekbench Metal at 89,557, PassMark G3D at 14,520, PassMark GPU Compute at 6,495, and PassMark 2D at 899. Its PassMark DirectX results run from 54 (DirectX 12) to 225 (DirectX 9). The P106-100 has one additional result, 3DMark Steel Nomad DirectX 12 at 899, which reflects its gaming-adjacent silicon even though it lacks any way to connect a monitor.

Architecture Differences

These cards come from different eras and different design philosophies. The W5700 uses the Navi 10 chip on AMD's RDNA 1.0 architecture, manufactured by TSMC on a 7 nm process. It packs 10,300 million transistors into a 251 mm² die, for a density of 41.0M transistors per mm². The P106-100 uses the GP106 chip on NVIDIA's Pascal architecture, built on TSMC's 16 nm process, with 4,400 million transistors on a 200 mm² die and a density of 22.0M per mm². The W5700 more than doubles the transistor count and roughly doubles the density.

The core counts follow the same pattern. The W5700 has 2,304 shading units, 144 TMUs, and 64 ROPs; the P106-100 has 1,280 shading units, 80 TMUs, and 48 ROPs. Neither card has RT cores or tensor cores.

Clocks favor the P106-100 at base: 1506 MHz versus 1400 MHz. But the W5700 boosts to 1880 MHz, well above the P106-100's 1709 MHz. The memory story is one-sided: the W5700 pairs 8 GB of GDDR6 on a 256-bit bus running at 1750 MHz (14 Gbps effective) for 448.0 GB/s of bandwidth, while the P106-100 has 6 GB of GDDR5 on a 192-bit bus at 2002 MHz (8 Gbps effective) for 192.2 GB/s. That is more than double the bandwidth for the AMD card.

Theoretical throughput numbers line up with the benchmark results. The W5700 delivers 8.663 TFLOPS FP32 and 17.33 TFLOPS FP16 at a 2:1 ratio, with a pixel rate of 120.3 GPixel/s and a texture rate of 270.7 GTexel/s. The P106-100 manages 4.375 TFLOPS FP32 and just 68.36 GFLOPS FP16 at a punishing 1:64 ratio, with 82.03 GPixel/s and 136.7 GTexel/s. The half-float gap is enormous and matters for any workload that leans on FP16.

Connectivity differs sharply. The W5700 uses PCIe 4.0 x16 and offers five mini-DisplayPort 1.4 outputs plus one USB Type-C. The P106-100 runs on PCIe 1.0 x16 and has no display outputs at all, a defining trait of its mining heritage. Both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4.

Power draw also splits them. The W5700 is rated at 205 W TDP, needs one 6-pin plus one 8-pin connector, and carries a suggested PSU of 550 W. The P106-100 sits at 120 W TDP with a single 6-pin connector and a 300 W suggested PSU. Both are dual-slot cards; the W5700 measures 267 mm long and 111 mm tall, while the P106-100 is 250 mm long. Both are end-of-life products. The W5700 launched with a launch MSRP of 799 USD and succeeded the Radeon Pro Vega line; the P106-100's recorded launch date is earlier and it has no listed predecessor or successor.

Where Each One Wins

The W5700 wins every recorded head-to-head contest. Its OpenCL lead of 107.5 percent and Vulkan lead of 114.9 percent make it the clear choice for compute workloads that appear in the database: GPGPU tasks, rendering, and any pipeline built on OpenCL or Vulkan APIs. Its Metal score of 89,557 and PassMark GPU Compute score of 6,495 further support the compute profile, and its superior memory bandwidth of 448.0 GB/s means large datasets move through the card faster.

The P106-100's wins are situational rather than benchmark-based. It draws 120 W versus 205 W, needs only one 6-pin connector, and carries a 300 W suggested PSU against 550 W. For a constrained build where power budget matters more than throughput, that difference is real. It is also physically shorter at 250 mm versus 267 mm. Its percentile placement of 68 shows it still ranks respectably overall, essentially tied with the Radeon Pro Vega 16 on average score.

What the P106-100 cannot do is drive a display: it has no outputs. Any use case requiring monitors goes to the W5700 by default, with its five mini-DisplayPort 1.4 connectors and USB Type-C.

The Verdict

The data makes this a straightforward call for performance. The W5700 doubles the P106-100 in both shared benchmarks, holds a higher overall percentile (71 versus 68), and delivers more than twice the memory bandwidth and FP32 throughput. For workstation compute, multi-display setups, and any OpenCL or Vulkan workload, the W5700 is the only sensible pick from these measurements.

The P106-100 suits a narrow niche: low-power auxiliary compute where the 120 W TDP, single 6-pin connector, and 300 W PSU recommendation fit the build, and where display output is irrelevant. Its average score of 23,249 keeps it competitive with cards like the Radeon Pro Vega 16 and RX 6600M in aggregate terms, but the head-to-head data shows the W5700 is decisively faster when both run the same test. If raw performance is the question, the W5700 is the answer.

FAQ

Q: Which card is faster in the shared benchmarks?

A: The Radeon Pro W5700 wins both, scoring 74,613 versus 35,951 in Geekbench OpenCL (107.5 percent ahead) and 70,706 versus 32,897 in Geekbench Vulkan (114.9 percent ahead).

Q: How do the two compare against the full GPU database?

A: The W5700 sits in the 71st percentile with an average score of 25,726; the P106-100 sits in the 68th percentile averaging 23,249.

Q: Can the P106-100 connect to a monitor?

A: No. It has no display outputs at all. The W5700 offers five mini-DisplayPort 1.4 outputs plus one USB Type-C.

Q: Which card needs more power?

A: The W5700, at 205 W TDP with one 6-pin and one 8-pin connector and a 550 W suggested PSU. The P106-100 runs at 120 W TDP with a single 6-pin connector and a 300 W suggested PSU.

Q: How much memory does each card have, and what type?

A: The W5700 has 8 GB of GDDR6 on a 256-bit bus with 448.0 GB/s of bandwidth. The P106-100 has 6 GB of GDDR5 on a 192-bit bus with 192.2 GB/s.

Q: Do either of these cards support ray tracing or AI acceleration hardware?

A: No. Neither card has RT cores or tensor cores according to the recorded specifications.

Specification Differences

• Manufacturer: AMD vs NVIDIA
• Chip: Navi 10 vs GP106
• Architecture: RDNA 1.0 vs Pascal
• Generation: Radeon Pro Navi (Navi Series) vs Mining GPUs
• Process node: 7 nm vs 16 nm (both TSMC)
• Transistors: 10,300 million vs 4,400 million
• Die size: 251 mm² vs 200 mm²
• Transistor density: 41.0M/mm² vs 22.0M/mm²
• Base clock: 1400 MHz vs 1506 MHz
• Boost clock: 1880 MHz vs 1709 MHz
• Memory: 8 GB GDDR6, 256-bit, 448.0 GB/s vs 6 GB GDDR5, 192-bit, 192.2 GB/s
• Memory speed: 1750 MHz (14 Gbps effective) vs 2002 MHz (8 Gbps effective)
• Shading units: 2304 vs 1280
• TMUs: 144 vs 80
• ROPs: 64 vs 48
• Pixel rate: 120.3 GPixel/s vs 82.03 GPixel/s
• Texture rate: 270.7 GTexel/s vs 136.7 GTexel/s
• FP32: 8.663 TFLOPS vs 4.375 TFLOPS
• FP16: 17.33 TFLOPS (2:1) vs 68.36 GFLOPS (1:64)
• TDP: 205 W vs 120 W
• Power connectors: 1x 6-pin + 1x 8-pin vs 1x 6-pin
• Suggested PSU: 550 W vs 300 W
• Bus interface: PCIe 4.0 x16 vs PCIe 1.0 x16
• Display outputs: 5x mini-DisplayPort 1.4a, 1x USB Type-C vs none
• Length: 267 mm vs 250 mm
• Height: 111 mm vs not recorded
• Release date: 2019-11-18 vs 2017-06-18
• Predecessor: Radeon Pro Vega vs none
• Launch MSRP: 799 USD vs not recorded