AMD Radeon RX 5300M vs NVIDIA P104-100 Comparison

Head-to-Head Benchmarks

The recorded benchmark data contains a single direct comparison between the AMD Radeon RX 5300M and the NVIDIA P104-100. That test is the Geekbench OpenCL workload, which measures raw compute throughput across a variety of GPU tasks. The results are decisive: the NVIDIA P104-100 scores 52,368 points, while the AMD Radeon RX 5300M scores 36,529 points. The performance gap is 30.2 percent in favor of the NVIDIA card. In other words, the P104-100 completes the OpenCL workload roughly a third faster than the RX 5300M, a substantial margin that places them in different performance tiers despite both being classified as end-of-life products.

Looking at the broader database context, the RX 5300M sits at the 80th percentile among all GPUs, which is a strong showing. Its average benchmark score of 36,529 places it in a cluster of professional and high-end consumer cards. The nearest rival in the database is the NVIDIA GeForce GTX TITAN X, which averages 36,530 points, a difference of essentially zero percent. The AMD Radeon PRO W6400 is 1.7 percent ahead with 37,157 points, while the NVIDIA T1000 trails by 0.7 percent at 36,289 points. The AMD Radeon Pro Duo sits 1.9 percent behind at 35,860 points. These narrow margins indicate that the RX 5300M delivers compute performance comparable to a flagship-class desktop GPU from an earlier generation, despite being a mobile part.

The P104-100, by contrast, holds the 77th percentile across all GPUs, slightly lower than the RX 5300M, yet its raw OpenCL score is far higher. Its average benchmark score across all recorded tests is 32,982 points, but that figure includes results from multiple workloads. The Geekbench OpenCL score alone is 52,368, and the card also records 45,165 in Geekbench Vulkan and 1,413 in 3DMark Steel Nomad DX12. The average is pulled down by the 3DMark result, which is a demanding next-generation rendering workload. Still, the OpenCL margin over the RX 5300M is unambiguous.

The P104-100's nearest rivals in the database are all mobile or professional parts with similar average scores. The NVIDIA T600 Mobile averages 32,849 points, just 0.4 percent behind. The NVIDIA T550 Mobile is 0.5 percent ahead at 33,161 points. The NVIDIA GeForce RTX 3050 Mobile is 0.6 percent ahead at 33,170 points, and the AMD Radeon Pro 570 is 0.7 percent ahead at 33,207 points. These figures show that the P104-100's average score, when all its workloads are combined, lands in the middle of a tight pack of modern mobile GPUs. However, its OpenCL performance is exceptional relative to that peer group, suggesting the card excels in compute-heavy tasks while lagging in newer graphics APIs.

Where Each One Wins

The benchmark results split cleanly by workload type. The NVIDIA P104-100 wins the only direct head-to-head test, the Geekbench OpenCL benchmark, by 30.2 percent. That workload is a general-purpose compute test that exercises shader throughput, memory bandwidth, and arithmetic throughput. The P104-100's advantage there is rooted in its larger shader array and wider memory bus, which the architecture section will detail. For users running OpenCL-based compute tasks, such as certain scientific simulations, image processing, or machine learning inference, the data clearly favors the NVIDIA card.

The RX 5300M, however, shows its strength in a different way. Although it loses the direct comparison, its percentile ranking among all GPUs is higher (80th versus 77th). This is because the database records only one benchmark for the RX 5300M, the Geekbench OpenCL score, while the P104-100 has additional results in Vulkan and 3DMark Steel Nomad DX12 that lower its average. The RX 5300M's single score is consistent and strong, placing it alongside desktop-class GPUs. For scenarios where a mobile GPU must deliver reliable compute performance without the overhead of a mining-specific design, the RX 5300M is the more balanced option.

In terms of raw compute density, the RX 5300M also wins on efficiency. Its transistor density is 40.5 million per square millimeter, versus 22.9 million for the P104-100, a direct consequence of the 7 nm manufacturing process versus 16 nm. That efficiency advantage translates into a lower thermal design power of 85 watts for the RX 5300M, while the P104-100 has no listed TDP but recommends a 200-watt power supply. The P104-100 is a dual-slot card with a single 8-pin power connector, whereas the RX 5300M has no power connectors and is intended for portable devices. Therefore, the RX 5300M wins in scenarios where power draw, physical footprint, and system integration flexibility are priorities. The P104-100 wins where maximum compute throughput per dollar of silicon area is the goal, which aligns with its mining-oriented design.

Architecture Differences

The two GPUs come from different architectural generations and manufacturing processes, which explains their divergent performance profiles. The AMD Radeon RX 5300M is built on the Navi 14 chip using the RDNA 1.0 architecture, fabricated on a 7 nm process at TSMC. It contains 6,400 million transistors on a die size of 158 square millimeters, yielding a transistor density of 40.5 million per square millimeter. The NVIDIA P104-100 uses the GP104 chip with the Pascal architecture, also fabricated at TSMC but on a 16 nm process. It packs 7,200 million transistors into a much larger die of 314 square millimeters, resulting in a transistor density of just 22.9 million per square millimeter. The P104-100 has more transistors overall, but the RX 5300M packs them far more densely, reflecting the newer process node.

The compute resources differ substantially. The RX 5300M has 1,408 shading units, 88 texture mapping units, and 32 render output units. The P104-100 has 1,920 shading units, 120 TMUs, and 64 ROPs. These numbers translate into peak rates: the RX 5300M achieves 46.24 GPixel/s and 127.2 GTexel/s, while the P104-100 reaches 110.9 GPixel/s and 208.0 GTexel/s. The P104-100 is roughly 2.4 times faster in pixel fill and 1.6 times faster in texture fill. Floating-point performance follows the same pattern: the RX 5300M delivers 4.069 TFLOPS in FP32, while the P104-100 delivers 6.655 TFLOPS, a 63.5 percent advantage. In FP16, the RX 5300M reaches 8.138 TFLOPS using a 2:1 ratio, while the P104-100 is heavily limited at 104.0 GFLOPS with a 1:64 ratio. That means the AMD card is vastly superior for half-precision workloads, while the NVIDIA card is built for FP32 throughput.

Memory configurations diverge as well. The RX 5300M uses 3 GB of GDDR6 on a 96-bit bus, with a memory clock of 1750 MHz (14 Gbps effective) and bandwidth of 168.0 GB/s. The P104-100 has 4 GB of GDDR5X on a 256-bit bus, with a memory clock of 1251 MHz (10 Gbps effective) and bandwidth of 320.3 GB/s. The P104-100 offers nearly double the memory bandwidth, which is critical for compute workloads that stream large datasets. The RX 5300M's narrower bus limits its bandwidth but reduces power consumption and die footprint.

The interface and output capabilities also differ. The RX 5300M uses PCIe 4.0 x8 and has display outputs described as "Portable Device Dependent," meaning it is designed for laptops where the system integrator decides the video connections. The P104-100 uses PCIe 1.0 x4, a legacy interface that is unusual for a 2017 card, and has no display outputs at all, confirming its mining-only purpose. The P104-100 is a dual-slot card measuring 267 mm (10.5 inches) in length, while the RX 5300M has no listed dimensions, reflecting its mobile BGA form factor.

Both cards support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, so API compatibility is identical. Neither has ray tracing cores or tensor cores. The release dates are far apart: the RX 5300M launched on November 12, 2019, while the P104-100 launched on December 11, 2017. Both are now end-of-life.

FAQ

Q: Which GPU is faster in the Geekbench OpenCL benchmark?

A: The NVIDIA P104-100 scores 52,368 points versus 36,529 for the AMD Radeon RX 5300M, a 30.2 percent advantage for the NVIDIA card.

Q: Does the RX 5300M have any performance advantage over the P104-100?

A: The RX 5300M has a higher percentile ranking among all GPUs (80th versus 77th) and delivers 8.138 TFLOPS in FP16, which is far higher than the P104-100's 104.0 GFLOPS. It also consumes less power with an 85 W TDP and requires no power connectors.

Q: Why does the P104-100 have a lower average benchmark score than its OpenCL score?

A: The database records three benchmarks for the P104-100: 52,368 in Geekbench OpenCL, 45,165 in Geekbench Vulkan, and 1,413 in 3DMark Steel Nomad DX12. The low 3DMark result pulls the average down to 32,982 points.

Q: What are the memory differences between the two cards?

A: The RX 5300M has 3 GB of GDDR6 on a 96-bit bus with 168.0 GB/s bandwidth. The P104-100 has 4 GB of GDDR5X on a 256-bit bus with 320.3 GB/s bandwidth, which is nearly double the bandwidth.

Q: Can the P104-100 be used for display output?

A: No. The P104-100 has no display outputs at all, while the RX 5300M's outputs are dependent on the portable device it is installed in.

Q: Which process node is smaller?

A: The RX 5300M uses a 7 nm process, while the P104-100 uses a 16 nm process. Both are fabricated at TSMC, but the 7 nm node allows the RX 5300M to fit 40.5 million transistors per square millimeter versus 22.9 million for the P104-100.

Specification Differences

The following fields differ between the AMD Radeon RX 5300M and the NVIDIA P104-100:

• Architecture: RDNA 1.0 versus Pascal
• Process Node: 7 nm versus 16 nm
• Transistors: 6,400 million versus 7,200 million
• Die Size: 158 mm² versus 314 mm²
• Transistor Density: 40.5M / mm² versus 22.9M / mm²
• Base Clock: 1000 MHz versus 1607 MHz
• Boost Clock: 1445 MHz versus 1733 MHz
• Game Clock: 1181 MHz versus not listed
• Memory Clock: 1750 MHz (14 Gbps effective) versus 1251 MHz (10 Gbps effective)
• Memory Size: 3 GB versus 4 GB
• Memory Type: GDDR6 versus GDDR5X
• Memory Bus Width: 96 bit versus 256 bit
• Memory Bandwidth: 168.0 GB/s versus 320.3 GB/s
• Shading Units: 1408 versus 1920
• TMUs: 88 versus 120
• ROPs: 32 versus 64
• Pixel Rate: 46.24 GPixel/s versus 110.9 GPixel/s
• Texture Rate: 127.2 GTexel/s versus 208.0 GTexel/s
• FP32 Performance: 4.069 TFLOPS versus 6.655 TFLOPS
• FP16 Performance: 8.138 TFLOPS (2:1) versus 104.0 GFLOPS (1:64)
• TDP: 85 W versus not listed
• Slot Width: not listed versus Dual-slot
• Power Connectors: None versus 1x 8-pin
• Suggested PSU: not listed versus 200 W
• Bus Interface: PCIe 4.0 x8 versus PCIe 1.0 x4
• Display Outputs: Portable Device Dependent versus No outputs
• Dimensions: not listed versus 267 mm (10.5 inches)
• Release Date: 2019-11-12 versus 2017-12-11
• Generation: Navi Mobile (RX 5000M) versus Mining GPUs

The Verdict

The data points to a clear split in intended use cases. The NVIDIA P104-100 is the stronger choice for raw compute throughput. Its OpenCL score of 52,368 is 30.2 percent ahead of the RX 5300M, and it also leads in pixel rate, texture rate, FP32 performance, and memory bandwidth. The card is physically large, uses a legacy PCIe 1.0 x4 interface, has no display outputs, and recommends a 200 W power supply, all of which indicate a mining-focused design. If the workload is OpenCL-based and the system can accommodate a dual-slot card with an 8-pin connector, the P104-100 is measurably faster.

The AMD Radeon RX 5300M is the better choice for portable systems and efficiency-sensitive deployments. It draws 85 W with no external power connectors, uses a modern PCIe 4.0 x8 interface, and is built on a 7 nm process that packs more transistors per square millimeter. Its FP16 performance of 8.138 TFLOPS is dramatically higher than the P104-100's 104.0 GFLOPS, making it suitable for half-precision compute tasks. The RX 5300M also holds a higher percentile ranking among all GPUs (80th versus 77th), and its OpenCL score places it in the company of desktop-class GPUs like the GeForce GTX TITAN X.

For users who prioritize compute performance in a fixed, power-tolerant environment, the P104-100 wins outright. For users who need a capable GPU in a mobile chassis with minimal power draw, the RX 5300M is the appropriate pick. The benchmark data does not support a single universal winner; it supports a workload-dependent decision. The P104-100 is faster, but the RX 5300M is more efficient and more flexible in terms of system integration.