AMD Ryzen Z2 Go GPU vs Intel Arc Pro B390 Comparison

Head-to-Head Benchmarks

The recorded data shows no direct head-to-head benchmark results between the AMD Ryzen Z2 Go GPU and the Intel Arc Pro B390. The database contains zero wins for either part in this pairing, and the average benchmark score for both is zero. This absence of measured performance data means a direct numerical comparison of application speed, gaming frame rates, or compute workloads cannot be constructed from the available facts.

Both graphics processors occupy the 50th percentile among all GPUs tracked in the database. This percentile placement indicates they sit at the median of the recorded performance distribution, though the lack of specific scores prevents any statement about how far apart they are in real-world execution. The AMD part reaches a boost clock of 2700 MHz, while the Intel part boosts to 2500 MHz. Clock frequency alone does not determine performance, but the higher AMD boost clock suggests a potential advantage in lightly threaded or clock-bound scenarios, provided the architecture can sustain it within its power envelope.

The Intel Arc Pro B390 delivers 7.680 TFLOPS of FP32 compute, which is 85.2% higher than the AMD Ryzen Z2 Go GPU's 4.147 TFLOPS. In FP16 throughput, the Intel part reaches 15.36 TFLOPS, compared to 8.294 TFLOPS for the AMD part, a 85.2% advantage as well. These figures indicate that for raw shader math, the Intel solution has a substantial theoretical lead, though the AMD part's lower power draw may allow it to sustain its clocks differently in constrained environments.

Texture rate favors the AMD part: 129.6 GTexel/s versus 120.0 GTexel/s for Intel, a 8.0% margin. Pixel rate also goes to AMD: 86.40 GPixel/s versus 60.00 GPixel/s, a 44.0% advantage. These rates suggest the AMD architecture is more efficient at filling pixels and sampling textures per clock, which could matter in resolution-bound or texture-heavy workloads. The Intel part compensates with more than double the shading units, 1536 versus 768, and a higher FP32 ceiling.

Both parts feature 12 ray tracing cores, so hardware-accelerated ray tracing capacity is nominally equal, though the actual ray tracing performance depends on the underlying architecture's efficiency, which the database does not quantify. The AMD part uses 48 texture mapping units, matching the Intel part's 48 TMUs. The AMD part has 32 raster output units, while the Intel part has 24, a 33.3% higher RO count for AMD.

Memory configuration differs fundamentally. The AMD Ryzen Z2 Go GPU uses 16 GB of LPDDR5 memory on a 128-bit bus, delivering 102.4 GB/s of bandwidth. The Intel Arc Pro B390 uses system-shared memory with system-dependent bandwidth, so no fixed bandwidth figure exists for comparison. The AMD part's dedicated memory pool provides a predictable 102.4 GB/s, while the Intel part's shared memory performance depends entirely on the host platform's memory subsystem.

Power consumption shows a 52 W gap between the two parts. The AMD part has a 28 W TDP, while the Intel part has an 80 W TDP. The AMD part requires no power connectors, and the Intel part also lists no power connectors, consistent with the Intel part's IGP slot width and bus interface. The AMD part's lower TDP suggests it can operate in more thermally constrained designs, while the Intel part's higher TDP indicates a need for more robust cooling and power delivery.

FAQ

Q: Which GPU has higher FP32 compute throughput?

A: The Intel Arc Pro B390 delivers 7.680 TFLOPS FP32, which is 85.2% higher than the AMD Ryzen Z2 Go GPU's 4.147 TFLOPS. The Intel part also leads in FP16 with 15.36 TFLOPS versus 8.294 TFLOPS.

Q: How do the pixel and texture rates compare?

A: The AMD Ryzen Z2 Go GPU leads in both. It produces 86.40 GPixel/s versus 60.00 GPixel/s for Intel, a 44.0% advantage, and 129.6 GTexel/s versus 120.0 GTexel/s, an 8.0% advantage.

Q: What are the memory subsystem differences?

A: The AMD part uses 16 GB of LPDDR5 on a 128-bit bus with 102.4 GB/s bandwidth. The Intel part uses system-shared memory with system-dependent bandwidth, so its memory size, bus width, and bandwidth are not fixed specifications.

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. Neither has a listed advantage in API feature support.

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

A: The AMD Ryzen Z2 Go GPU has a 28 W TDP, while the Intel Arc Pro B390 has an 80 W TDP. The AMD part draws 52 W less power, and both use no external power connectors.

Q: How many ray tracing cores does each GPU have?

A: Both GPUs have 12 ray tracing cores. The AMD part also has 768 shading units, 48 TMUs, and 32 ROPs, while the Intel part has 1536 shading units, 48 TMUs, and 24 ROPs.

Architecture Differences

The AMD Ryzen Z2 Go GPU uses the Rembrandt+ chip with RDNA 2.0 architecture, fabricated on a 6 nm process at TSMC. The chip contains 13,100 million transistors on a 208 mm² die, yielding a transistor density of 63.0M per mm². This is a console GPU generation part, and its production status is active. The release date recorded is December 31, 2024.

The Intel Arc Pro B390 uses the Panther Lake chip with Xe3-LPG architecture, fabricated on a 3 nm process at Intel. The transistor count and die size are listed as unknown, so no density figure exists. This is part of the Arc Graphics-WM (Panther Lake) generation, with a predecessor listed as HD Graphics-WM. Its production status is active, and its release date is January 26, 2026.

The process node difference is significant: 6 nm for AMD versus 3 nm for Intel. The Intel part's smaller node suggests higher transistor density potential, though the unknown transistor count prevents confirmation. The AMD part's known 13,100 million transistors on 208 mm² indicates a mature, well-characterized design, while the Intel part's metrics remain opaque.

The AMD part has a base clock of 800 MHz and a boost clock of 2700 MHz. The Intel part has a base clock of 300 MHz and a boost clock of 2500 MHz. The AMD part's base clock is 2.67 times higher, and its boost clock is 8.0% higher. The Intel part's lower base clock suggests it relies more heavily on boost behavior, potentially ramping up only under load.

Shading unit counts differ substantially: the Intel part has 1536 shading units, exactly double the AMD part's 768. This doubling explains the Intel part's higher FP32 throughput, as each shading unit can execute more operations per clock. The AMD part compensates with higher clocks, but the raw unit count advantage for Intel is clear.

Both parts have 48 TMUs and 12 ray tracing cores. The AMD part has 32 ROPs versus 24 for Intel, which explains its pixel rate advantage. The AMD part's memory is dedicated LPDDR5 with a fixed 128-bit bus, while the Intel part uses system-shared memory, meaning its memory architecture is entirely dependent on the host platform.

The AMD part's display output is 1x USB Type-C, a specific configuration. The Intel part's display outputs are portable device dependent, meaning they vary by the host device. This reflects the Intel part's IGP nature, integrated into a processor package, while the AMD part appears to be a discrete GPU solution.

The Intel part has a slot width of IGP and a bus interface of IGP, confirming its integrated design. The AMD part has no listed slot width, bus interface, or power connectors beyond "None." The TDP difference, 28 W versus 80 W, aligns with the discrete versus integrated positioning, though the Intel part's higher TDP for an IGP is notable.

The Verdict

The data shows two GPUs with complementary strengths. The AMD Ryzen Z2 Go GPU offers higher pixel and texture rates, a dedicated 16 GB memory pool with 102.4 GB/s bandwidth, and a much lower 28 W TDP. The Intel Arc Pro B390 offers more than double the FP32 compute, 7.680 TFLOPS, and double the shading units, 1536, but with a higher 80 W TDP and system-shared memory.

For workloads that stress fill rate and memory bandwidth, such as high-resolution rendering or texture-heavy scenes, the database indicates the AMD part has the advantage. Its 86.40 GPixel/s pixel rate and 129.6 GTexel/s texture rate exceed the Intel part's figures by 44.0% and 8.0%, respectively. The fixed 102.4 GB/s bandwidth provides predictable performance, whereas the Intel part's bandwidth is system dependent.

For compute-heavy tasks, such as FP32 shader math or FP16 machine learning inference, the Intel part's theoretical throughput is superior. Its 7.680 TFLOPS FP32 and 15.36 TFLOPS FP16 lead the AMD part by 85.2% in both metrics. The doubling of shading units gives it a structural advantage in parallel workloads that scale with unit count rather than clock speed.

The 52 W TDP gap is a decisive differentiator. The AMD part's 28 W TDP enables deployment in power-constrained designs, while the Intel part's 80 W TDP requires more substantial cooling and power delivery. Both parts avoid external power connectors, but the Intel part's IGP nature means it shares power with the host processor.

The release dates differ by roughly one year, with the AMD part recorded at December 31, 2024, and the Intel part at January 26, 2026. This temporal gap means the Intel part is a newer design, though the database does not provide benchmark scores to confirm whether newer translates to faster in practice.

The lack of head-to-head benchmarks and average scores means the verdict rests on architectural and specification differences rather than measured performance. The AMD part suits scenarios prioritizing fill rate, memory bandwidth, and power efficiency. The Intel part suits scenarios prioritizing raw compute throughput and shading unit count, provided the host system can supply adequate memory bandwidth and cooling.

Specification Differences

| Specification | AMD Ryzen Z2 Go GPU | Intel Arc Pro B390 |

|---|---|---|

| Chip | Rembrandt+ | Panther Lake |

| Architecture | RDNA 2.0 | Xe3-LPG |

| Process Node | 6 nm, TSMC | 3 nm, Intel |

| Transistors | 13,100 million | unknown |

| Die Size | 208 mm² | unknown |

| Transistor Density | 63.0M / mm² | not listed |

| Base Clock | 800 MHz | 300 MHz |

| Boost Clock | 2700 MHz | 2500 MHz |

| Memory Size | 16 GB | System Shared |

| Memory Type | LPDDR5 | System Shared |

| Memory Bus Width | 128 bit | System Shared |

| Memory Bandwidth | 102.4 GB/s | System Dependent |

| Shading Units | 768 | 1536 |

| ROPs | 32 | 24 |

| Pixel Rate | 86.40 GPixel/s | 60.00 GPixel/s |

| Texture Rate | 129.6 GTexel/s | 120.0 GTexel/s |

| FP32 | 4.147 TFLOPS | 7.680 TFLOPS |

| FP16 | 8.294 TFLOPS | 15.36 TFLOPS |

| TDP | 28 W | 80 W |

| Slot Width | not listed | IGP |

| Bus Interface | not listed | IGP |

| Display Outputs | 1x USB Type-C | Portable Device Dependent |

| Release Date | 2024-12-31 | 2026-01-26 |

| Predecessor | not listed | HD Graphics-WM |

Shared specifications include 48 TMUs, 12 ray tracing cores, DirectX 12 Ultimate (12_2), OpenGL 4.6, Vulkan 1.4, no power connectors, no suggested PSU, no launch MSRP, and no successor. Both parts hold the 50th percentile among all GPUs, and both have zero average benchmark scores and zero head-to-head wins in the database.