AMD Radeon RX 6450M vs Intel Arc B390 Comparison

Head-to-Head Benchmarks

The database contains no direct head-to-head benchmark comparisons between the AMD Radeon RX 6450M and the Intel Arc B390. The only recorded benchmark score belongs to the Intel Arc B390, which posts a result of 1482 points in the 3DMark Steel Nomad DX12 test. This single data point places the Arc B390 at the 9th percentile among all GPUs in the database, indicating that its performance sits well below the majority of recorded graphics hardware.

The AMD Radeon RX 6450M shows no benchmark entries in the database, meaning its average benchmark score is zero and its percentile ranking stands at 50. Without a recorded score for the Radeon RX 6450M, the database cannot establish a direct performance delta between the two parts. However, the Intel Arc B390's nearest rivals in the database provide useful context for its standing. The Arc B390 leads the NVIDIA GeForce GT 520MX by 1.3 percent, the NVIDIA GeForce 800M by 1.5 percent, the NVIDIA GeForce GT 625 OEM by 2.5 percent, and the NVIDIA GeForce GT 710 by 2.7 percent. These margins are narrow, which suggests the Arc B390 operates in the same performance tier as those older, low-end discrete GPUs.

The lack of a benchmark score for the Radeon RX 6450M is notable because its theoretical compute capabilities are substantially higher than what the Arc B390 delivers in practice. The RX 6450M is rated for 3.779 TFLOPS of FP32 throughput, while the Arc B390 is rated for 7.680 TFLOPS. The Arc B390's recorded score of 1482, however, does not reflect that raw advantage, likely due to driver maturity, memory bottlenecks, or thermal constraints in its integrated form factor. The database indicates zero wins for either GPU in head-to-head testing, which aligns with the absence of comparative benchmark data.

Architecture Differences

The two GPUs come from entirely different architectural lineages. The AMD Radeon RX 6450M uses the Navi 24 chip built on RDNA 2.0 architecture, manufactured on a 6 nm process at TSMC. It belongs to the Navi Mobile generation within the RX 6000M series. The Intel Arc B390, by contrast, uses the Panther Lake chip built on Xe3-LPG architecture, manufactured on a 3 nm process at Intel, and belongs to the Arc Graphics-M generation for Panther Lake.

The transistor counts differ dramatically. The RX 6450M packs 5,400 million transistors on a 107 mm² die, yielding a transistor density of 50.5 million per square millimeter. The Arc B390's transistor count and die size are listed as unknown in the database, so no direct comparison can be made on those metrics. The process node advantage clearly favors Intel, with 3 nm versus 6 nm, but the absence of transistor data for the Arc B390 prevents a full structural comparison.

Memory architecture presents another major divergence. The RX 6450M uses dedicated 4 GB of GDDR6 memory on a 64-bit bus, delivering 128.0 GB/s of bandwidth. Its memory clock runs at 2000 MHz with 16 Gbps effective data rate. The Arc B390 uses system-shared memory, meaning it has no dedicated VRAM; its memory size, type, bus width, and bandwidth are all listed as system dependent. This is a fundamental difference: the RX 6450M has guaranteed local memory bandwidth, while the Arc B390's performance depends entirely on the host system's memory configuration.

Compute unit counts also differ significantly. The RX 6450M has 768 shading units, 48 texture mapping units, and 32 ROPs. The Arc B390 has 1536 shading units, 48 texture mapping units, and 24 ROPs. Both GPUs feature 12 ray tracing cores. The RX 6450M's pixel rate is 78.72 GPixel/s and its texture rate is 118.1 GTexel/s. The Arc B390 posts a lower pixel rate of 60.00 GPixel/s but a slightly higher texture rate of 120.0 GTexel/s. The RX 6450M's higher ROP count gives it an edge in fill-rate-bound scenarios, while the Arc B390's additional shading units provide more raw compute parallelism.

Clock behavior differs as well. The RX 6450M runs at a base clock of 2000 MHz, a boost clock of 2460 MHz, and a game clock of 2220 MHz. The Arc B390 has a much lower base clock of 300 MHz but a boost clock of 2500 MHz. This wide clock range on the Intel part suggests aggressive power management, with the GPU spending most of its time at lower clocks unless demand spikes. The RX 6450M's higher base clock indicates more consistent sustained performance at its modest 50 W TDP, whereas the Arc B390 carries an 80 W TDP and still relies on system-shared resources.

Both GPUs support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Both are listed as integrated-class parts with "IGP" slot width and no dedicated power connectors. Display outputs for both are portable device dependent. The RX 6450M uses a PCIe 4.0 x4 bus interface, while the Arc B390's bus interface is simply listed as IGP, reflecting its fully integrated nature.

FAQ

Q: Which GPU has a higher FP32 compute throughput?

A: The Intel Arc B390 is rated at 7.680 TFLOPS of FP32 performance, which is roughly double the AMD Radeon RX 6450M's 3.779 TFLOPS.

Q: How does the Intel Arc B390 compare to its nearest rivals in the database?

A: The Arc B390 leads the NVIDIA GeForce GT 520MX by 1.3 percent, the NVIDIA GeForce 800M by 1.5 percent, the NVIDIA GeForce GT 625 OEM by 2.5 percent, and the NVIDIA GeForce GT 710 by 2.7 percent in the Steel Nomad DX12 test.

Q: What memory configuration does each GPU use?

A: The RX 6450M uses 4 GB of dedicated GDDR6 memory on a 64-bit bus with 128.0 GB/s bandwidth. The Arc B390 uses system-shared memory, with its size, type, bus width, and bandwidth all dependent on the host system.

Q: What are the process node and foundry differences?

A: The RX 6450M is built on a 6 nm process at TSMC, while the Arc B390 is built on a 3 nm process at Intel.

Q: Do both GPUs support the same graphics APIs?

A: Yes, both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: What is the TDP difference between the two GPUs?

A: The RX 6450M has a TDP of 50 W, while the Arc B390 has a TDP of 80 W.

The Verdict

The data points to two very different products with different intended roles. The AMD Radeon RX 6450M offers a complete, self-contained graphics solution with dedicated memory, a 64-bit bus, and a stable clock profile. Its 50 W TDP and PCIe 4.0 x4 interface make it suitable for compact mobile platforms where consistent local bandwidth matters. The Intel Arc B390, while boasting higher raw FP32 throughput and a more advanced 3 nm manufacturing process, relies entirely on system-shared memory and posts a modest benchmark score of 1482, placing it at the 9th percentile of all GPUs.

The benchmark evidence indicates that the Arc B390's real-world performance, at least in the Steel Nomad DX12 test, sits alongside very old low-end discrete GPUs such as the GeForce GT 710 and GeForce GT 520MX. Its theoretical compute advantage over the RX 6450M does not translate into a superior recorded result, likely because shared memory bandwidth and the absence of dedicated VRAM create severe constraints. The RX 6450M, with its dedicated 128.0 GB/s of bandwidth and 32 ROPs, is better positioned for consistent rendering output, even though its raw FP32 count is lower.

For users who need predictable local memory performance and a lower power envelope, the RX 6450M's specifications are more compelling on paper. For users who prioritize raw shading throughput and are willing to depend on system memory, the Arc B390 offers more compute horsepower, but the database's only recorded benchmark provides no evidence that this translates into superior performance. The Arc B390's 80 W TDP is also higher, which may be a consideration in thermally constrained mobile designs.

Specification Differences

The following specifications differ between the two GPUs:

• Process node: 6 nm (RX 6450M) versus 3 nm (Arc B390)
• Foundry: TSMC (RX 6450M) versus Intel (Arc B390)
• Transistors: 5,400 million (RX 6450M) versus unknown (Arc B390)
• Die size: 107 mm² (RX 6450M) versus unknown (Arc B390)
• Transistor density: 50.5M / mm² (RX 6450M) versus not listed (Arc B390)
• Base clock: 2000 MHz (RX 6450M) versus 300 MHz (Arc B390)
• Boost clock: 2460 MHz (RX 6450M) versus 2500 MHz (Arc B390)
• Game clock: 2220 MHz (RX 6450M) versus not listed (Arc B390)
• Memory clock: 16 Gbps effective (RX 6450M) versus system shared (Arc B390)
• Memory size: 4 GB (RX 6450M) versus system shared (Arc B390)
• Memory type: GDDR6 (RX 6450M) versus system shared (Arc B390)
• Memory bus width: 64 bit (RX 6450M) versus system shared (Arc B390)
• Memory bandwidth: 128.0 GB/s (RX 6450M) versus system dependent (Arc B390)
• Shading units: 768 (RX 6450M) versus 1536 (Arc B390)
• ROPs: 32 (RX 6450M) versus 24 (Arc B390)
• Pixel rate: 78.72 GPixel/s (RX 6450M) versus 60.00 GPixel/s (Arc B390)
• Texture rate: 118.1 GTexel/s (RX 6450M) versus 120.0 GTexel/s (Arc B390)
• FP32 throughput: 3.779 TFLOPS (RX 6450M) versus 7.680 TFLOPS (Arc B390)
• FP16 throughput: 7.557 TFLOPS (RX 6450M) versus 15.36 TFLOPS (Arc B390)
• TDP: 50 W (RX 6450M) versus 80 W (Arc B390)
• Bus interface: PCIe 4.0 x4 (RX 6450M) versus IGP (Arc B390)
• Release date: 2023-01-03 (RX 6450M) versus 2026-01-26 (Arc B390)
• Architecture: RDNA 2.0 (RX 6450M) versus Xe3-LPG (Arc B390)
• Chip: Navi 24 (RX 6450M) versus Panther Lake (Arc B390)
• Generation: Navi Mobile RX 6000M (RX 6450M) versus Arc Graphics-M Panther Lake (Arc B390)

Where Each One Wins

The AMD Radeon RX 6450M wins in scenarios that demand consistent memory bandwidth and lower power consumption. Its dedicated 128.0 GB/s of bandwidth, 4 GB of GDDR6, and 32 ROPs give it a clear structural advantage for rasterization-bound tasks where fill rate matters. The higher pixel rate of 78.72 GPixel/s compared to the Arc B390's 60.00 GPixel/s reinforces this positioning. The RX 6450M's 50 W TDP also makes it the more power-efficient choice on paper, which is relevant for thin-and-light mobile devices where thermal headroom is limited.

The Intel Arc B390 wins on raw compute specifications. Its 1536 shading units deliver 7.680 TFLOPS of FP32 throughput, double the RX 6450M's 3.779 TFLOPS. Its FP16 output of 15.36 TFLOPS similarly doubles the RX 6450M's 7.557 TFLOPS. The Arc B390 also has a higher texture rate at 120.0 GTexel/s versus 118.1 GTexel/s. The 3 nm manufacturing process at Intel gives it a fabrication advantage, and its 2500 MHz boost clock is slightly higher than the RX 6450M's 2460 MHz. For compute-heavy workloads that can use the additional shading units and system memory, the Arc B390 has the theoretical edge.

The benchmark record tells a different story. The Arc B390's only recorded score of 1482 in Steel Nomad DX12 places it near the bottom of the database, with a 9th percentile ranking. The RX 6450M has no recorded benchmarks, so its actual performance cannot be confirmed. The Arc B390's narrow wins over decades-old NVIDIA parts suggest that its architectural advantages do not translate into meaningful real-world performance gains in the tested workload. The RX 6450M's dedicated memory and higher pixel rate likely make it the safer choice for typical gaming and rendering tasks, while the Arc B390 remains an unproven quantity whose theoretical compute capabilities exceed its demonstrated results.