AMD Radeon Pro Vega 16 vs NVIDIA P106-100 Comparison

Head-to-Head Benchmarks

The data is unambiguous: the NVIDIA P106-100 dominates the AMD Radeon Pro Vega 16 in every benchmark where both were tested. The P106-100 wins both head-to-head comparisons, and the margins are substantial. In Geekbench OpenCL, the P106-100 scores 35,951 against the Radeon Pro Vega 16’s 18,268 — a 49.2% advantage for NVIDIA. That is not a close contest; it is a near-doubling of raw compute throughput in a compute-heavy API. The Vulkan result tells a similar story, with the P106-100 posting 32,897 versus 21,832 for AMD, a 33.6% lead. Both deltas are decisive, and neither benchmark offers the Radeon Pro Vega 16 a single point of redemption.

Looking at the broader benchmark averages, the two cards are virtually tied. The AMD Radeon Pro Vega 16 has an average benchmark score of 23,250, while the NVIDIA P106-100 sits at 23,249 — a difference of exactly one point, or 0.0%. This apparent contradiction resolves when you account for the fact that the AMD card has additional benchmark results (Geekbench Metal at 29,650) that are not part of the head-to-head comparison. The Metal score is the Radeon Pro Vega 16’s strongest showing, and it lifts its average to parity despite losing both shared tests. The NVIDIA card’s own extra result, a 3DMark Steel Nomad DX12 score of 899, is comparatively weak and does not offset its OpenCL and Vulkan wins.

The percentile ranking reinforces this split personality. Both cards sit at the 68th percentile among all GPUs, meaning they outperform roughly two-thirds of the database. But percentile rank is an aggregate measure; it masks the fact that the two cards achieve that standing through entirely different benchmark profiles. The Radeon Pro Vega 16 is a specialist that excels in Apple’s Metal API, while the P106-100 is a generalist that wins in the cross-platform OpenCL and Vulkan workloads. The deltaPct values in the nearest-rival data confirm the tight grouping: the AMD Radeon RX 6600M is 0.1% behind the Radeon Pro Vega 16’s average, and the AMD Radeon R9 M290X is also 0.1% behind. The AMD Radeon AI PRO R9700 trails by 0.3%. The P106-100 sits in the same cluster, with the RX 6600M 0.1% above it and the R9 M290X 0.1% above it.

The Verdict

The NVIDIA P106-100 is the better card for anyone prioritizing raw compute performance in OpenCL or Vulkan. The data shows a 49.2% lead in OpenCL and a 33.6% lead in Vulkan, and those are the two APIs that matter most for cross-platform workloads. The Radeon Pro Vega 16’s only benchmark win is in Metal, where it scores 29,650 — but that result is not part of the head-to-head comparison, and Metal is a platform-specific API with limited applicability outside Apple’s ecosystem. If your workload runs on Metal, the Radeon Pro Vega 16 is the only choice between these two, because the P106-100 has no Metal result in the data.

For everyone else, the P106-100 is the straightforward pick. It wins both shared tests, it has double the memory capacity (6 GB versus 4 GB), and it offers higher raw throughput in every measured category. The Radeon Pro Vega 16’s average score is one point higher, but that is a statistical artifact of its Metal benchmark, not evidence of real-world superiority. The verdict is not close: pick the P106-100 for OpenCL and Vulkan workloads, pick the Radeon Pro Vega 16 only if Metal compatibility is a hard requirement.

Where Each One Wins

The AMD Radeon Pro Vega 16 wins in exactly one scenario: Metal-based applications. Its Geekbench Metal score of 29,650 is the single highest benchmark result for either card in the full dataset. This suggests the card is optimized for Apple’s graphics stack, which aligns with its generation name "Radeon Pro Mac (Vega Series)" and its integrated form factor (IGP slot width). The card’s HBM2 memory with a 1024-bit bus and 307.2 GB/s bandwidth is a niche configuration that appears designed for specific compute patterns rather than broad compatibility.

The NVIDIA P106-100 wins everywhere else. Its OpenCL score of 35,951 is nearly double the Radeon Pro Vega 16’s 18,268, and its Vulkan score of 32,897 is 11,065 points higher. The P106-100 also has a 3DMark Steel Nomad DX12 score of 899, a benchmark the AMD card does not have a result for. The card’s 6 GB GDDR5 memory on a 192-bit bus delivers 192.2 GB/s of bandwidth, and its 4.375 TFLOPS FP32 performance is substantially higher than the Radeon Pro Vega 16’s 2.437 TFLOPS. For compute-heavy tasks, rendering, or any workload that leverages OpenCL or Vulkan, the P106-100 is the clear winner. Its "Mining GPUs" generation name and lack of display outputs indicate it is purpose-built for headless compute, which explains its performance profile.

FAQ

Q: Which card has the higher average benchmark score?

A: The AMD Radeon Pro Vega 16 has an average benchmark score of 23,250, while the NVIDIA P106-100 has 23,249 — a difference of one point, or 0.0%.

Q: Does the AMD Radeon Pro Vega 16 win any benchmark against the NVIDIA P106-100?

A: In the head-to-head comparison, no. The P106-100 wins both Geekbench OpenCL (35,951 vs 18,268) and Geekbench Vulkan (32,897 vs 21,832). However, the Radeon Pro Vega 16 has a Geekbench Metal score of 29,650, which the P106-100 has no result for.

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

A: The largest delta is in Geekbench OpenCL, where the NVIDIA P106-100 leads by 49.2%. The Vulkan gap is smaller but still substantial at 33.6%.

Q: How does the memory configuration differ?

A: The AMD Radeon Pro Vega 16 uses 4 GB of HBM2 with a 1024-bit bus and 307.2 GB/s bandwidth. The NVIDIA P106-100 uses 6 GB of GDDR5 with a 192-bit bus and 192.2 GB/s bandwidth.

Q: Which card has higher FP32 performance?

A: The NVIDIA P106-100 delivers 4.375 TFLOPS FP32, while the AMD Radeon Pro Vega 16 delivers 2.437 TFLOPS. The P106-100 is 79.5% higher in this metric.

Q: What percentile do both cards rank at?

A: Both the AMD Radeon Pro Vega 16 and the NVIDIA P106-100 are at the 68th percentile among all GPUs in the database.

Architecture Differences

The two cards come from fundamentally different architectural lineages. The AMD Radeon Pro Vega 16 uses the Vega 12 chip built on GCN 5.0 architecture, fabricated on a 14 nm process at GlobalFoundries. The NVIDIA P106-100 uses the GP106 chip on the older Pascal architecture, built on a 16 nm process at TSMC. The process node difference is small, but the architectural philosophies diverge sharply. AMD’s GCN 5.0 is a compute-oriented design with 1,024 shading units, 64 texture mapping units, and 32 raster operations units. NVIDIA’s Pascal has 1,280 shading units, 80 TMUs, and 48 ROPs — higher in every count, which explains its 4.375 TFLOPS FP32 versus the Radeon Pro Vega 16’s 2.437 TFLOPS.

Memory architecture is another major divergence. The Radeon Pro Vega 16 uses HBM2 with a 1024-bit bus, achieving 307.2 GB/s of bandwidth despite only 4 GB of capacity. The P106-100 uses GDDR5 on a 192-bit bus, delivering 192.2 GB/s from 6 GB of memory. The AMD card has a bandwidth advantage of 115 GB/s, but the NVIDIA card has 50% more memory capacity. The Radeon Pro Vega 16’s FP16 throughput is 4.874 TFLOPS at a 2:1 ratio, while the P106-100’s FP16 is a paltry 68.36 GFLOPS at a 1:64 ratio — a massive difference in half-precision compute that favors AMD by a factor of over 70.

The API support also differs. Both cards support DirectX 12 (12_1) and OpenGL 4.6, but the Radeon Pro Vega 16 supports Vulkan 1.3 while the P106-100 supports Vulkan 1.4. The newer Vulkan version on the NVIDIA card is consistent with its higher Vulkan benchmark score. The P106-100 also has a larger die at 200 mm² with 4,400 million transistors, giving a transistor density of 22.0M per mm²; the Radeon Pro Vega 16’s die size and transistor count are not listed in the data.

Specification Differences

The most consequential difference is memory capacity: the NVIDIA P106-100 has 6 GB of GDDR5, while the AMD Radeon Pro Vega 16 has 4 GB of HBM2. The bus widths differ correspondingly, with the AMD card using a 1024-bit bus and the NVIDIA card using a 192-bit bus. Bandwidth favors AMD at 307.2 GB/s versus 192.2 GB/s for NVIDIA. Clock speeds are also markedly different. The P106-100 has a base clock of 1506 MHz and a boost clock of 1709 MHz, while the Radeon Pro Vega 16 runs at 815 MHz base and 1190 MHz boost. The NVIDIA card’s memory runs at 2002 MHz (8 Gbps effective), while the AMD card’s memory runs at 1200 MHz (2.4 Gbps effective).

Compute unit counts favor NVIDIA across the board: 1,280 shading units versus 1,024, 80 TMUs versus 64, and 48 ROPs versus 32. Pixel rate is 82.03 GPixel/s for the P106-100 versus 38.08 GPixel/s for the Radeon Pro Vega 16, and texture rate is 136.7 GTexel/s versus 76.16 GTexel/s. The TDP difference is substantial: the P106-100 draws 120 W and requires a single 6-pin power connector with a suggested 300 W PSU, while the Radeon Pro Vega 16 draws 75 W and uses no power connectors. The P106-100 is a dual-slot card measuring 250 mm (9.8 inches) in length, while the Radeon Pro Vega 16 is an IGP (integrated graphics processor) with no listed dimensions.

The bus interface differs as well: the Radeon Pro Vega 16 uses PCIe 3.0 x16, while the P106-100 uses PCIe 1.0 x16. Display outputs are another key split — the Radeon Pro Vega 16 has "Portable Device Dependent" outputs, while the P106-100 has no outputs at all. The P106-100 was released on 2018-06-18, five months before the Radeon Pro Vega 16’s 2018-11-13 release date. Both cards are end-of-life. Neither has a listed launch MSRP. The transistor density for the P106-100 is listed at 22.0M / mm²; the Radeon Pro Vega 16’s transistor count and die size are not provided in the data.