AMD Ryzen AI Max PRO 380

The AMD Ryzen AI Max PRO 380 is a 6-core, 12-thread mobile processor built on the Zen 5 architecture (Strix Halo codename) at TSMC's 4nm node, carrying a 55W TDP, a 3.60 GHz base clock, a 4.90 GHz boost clock, and 16 MB of shared L3 cache. The database places it at the 50th percentile among all tracked CPUs, though no measured benchmark scores are currently recorded for this part.

Benchmark Performance

The FACT PACK lists no benchmark entries for this processor — the benchmarks array is empty and the average benchmark score is 0. The only quantitative performance indicator available is the percentileVsAllCpus value of 50, which places the Ryzen AI Max PRO 380 at the exact median of the database's CPU ranking. That position means half of all tracked CPUs sit above it and half below — a genuinely middle-of-the-pack placement, not a top-tier or entry-level designation.

Without nearestRivals data, direct percentage deltas against specific competitors cannot be computed from the available facts. However, the clock architecture provides a meaningful performance profile. The base clock of 3.60 GHz represents the sustained all-core frequency under load, while the boost clock of 4.90 GHz is the maximum single-core or light-thread ceiling. For a mobile part at 55W, a boost clock of 4.90 GHz is a strong figure, indicating that lightly threaded workloads can reach high instantaneous speeds.

The core configuration — 6 cores and 12 threads — is modest relative to the broader CPU market, but the Zen 5 architecture and the 4nm TSMC process node contribute to per-core efficiency. The 16 MB of shared L3 cache provides a common pool for all cores, while the per-core allocations of 80 KB of L1 and 1 MB of L2 support individual core performance. The 128.0 GB/s memory bandwidth from dual-channel LPDDR5X is a notable asset, feeding both the CPU cores and the integrated Radeon 8040S GPU.

The 50th percentile ranking, combined with the absence of measured scores, suggests the database has not yet collected runtime data for this part. The release date of 2025-01-05 is recent, and the production status is Active, so the processor is new to the market. The percentile figure is therefore likely derived from the specification profile rather than from empirical benchmark runs. This is a common situation for newly released hardware — the specification-based ranking provides a preliminary position, and measured scores will refine it as data accumulates.

Who Should Consider It

The market segment is Mobile, so the Ryzen AI Max PRO 380 targets laptops and compact portable systems. The integrated Radeon 8040S GPU is a significant inclusion — it means the processor can drive displays and handle graphics workloads without requiring a discrete GPU. For gaming, the combination of a 6-core Zen 5 CPU and the Radeon 8040S iGPU suggests the ability to run lighter titles and esports at playable framerates, though the absence of benchmark scores means no specific performance claims can be substantiated from the data.

For content creation, the 12 threads provide parallel throughput for video encoding, 3D rendering, and photo editing. The 128.0 GB/s memory bandwidth from dual-channel LPDDR5X is a strong asset for memory-bound creative workloads — this is a high figure for a mobile part, and it directly feeds both the CPU and the integrated GPU. The 16 MB shared L3 cache helps reduce memory traffic, which benefits workloads that repeatedly access the same data.

For office productivity, the 6 cores and 12 threads are more than sufficient for document work, spreadsheet analysis, web browsing, and video conferencing. The 50th percentile ranking suggests the processor will handle typical office loads without strain, and the 55W TDP makes it suitable for thin-and-light laptops where sustained performance matters more than raw peak output.

The ECC memory support is a notable differentiator. The FACT PACK lists eccMemory as true, which is unusual for a consumer mobile processor and points toward workstation-class reliability use cases. Users who need data integrity — financial analysis, scientific computing, long-running simulations — would find this appealing. The processor is not unlocked (multiplierUnlocked is false), so overclocking is off the table; users who want to tune their system should look elsewhere.

Power and Thermals

The TDP is rated at 55W. This is the thermal design power — the amount of heat the cooling solution must dissipate under sustained load. For a mobile processor, 55W places it in the upper-mid range of laptop CPUs. It is not an ultra-low-power part, nor is it a desktop-class component with a much higher thermal envelope. The 4nm TSMC process helps here; smaller process nodes typically improve efficiency, allowing more performance per watt.

The 55W envelope, combined with the 4.90 GHz boost clock, indicates the processor can reach high clock speeds but will need a capable cooling solution to sustain them. A laptop with a decent heat pipe and fan arrangement should handle it; ultra-thin chassis designs might throttle under sustained all-core loads. The integrated Radeon 8040S GPU shares the same thermal budget — when both CPU and GPU are active, as in gaming, the 55W envelope is divided between them. Sustained gaming performance will therefore depend on how the system balances power between the two components.

The ECC memory support adds a small power overhead, but the dual-channel LPDDR5X at 128.0 GB/s is efficient memory. The 16 MB L3 cache helps reduce memory traffic, which in turn reduces power consumption. The base clock of 3.60 GHz is the sustained all-core frequency, and at that level the 55W TDP is a reasonable envelope for a 6-core Zen 5 part.

Platform and Compatibility

The processor uses AMD Socket FP11, a mobile socket. This is a BGA-style package, typically soldered to the motherboard. The upgrade path is therefore limited — the processor is not meant to be swapped out. Users should plan to keep this processor for the life of the system.

Memory support is LPDDR5X, which is soldered on most mobile boards. The dual-channel configuration delivers 128.0 GB/s of memory bandwidth, and ECC is supported — a differentiator in the mobile space. PCIe support is Gen 4 with 16 lanes available from the CPU. This is enough for a discrete GPU (typically 8-16 lanes) and one or two NVMe SSDs. The Gen 4 standard provides high bandwidth for storage and graphics.

The processor was released on 2025-01-05 and is in active production (productionStatus is Active). The part number is 100-000001425. The integrated graphics is the Radeon 8040S, which is part of the Radeon 8000 series. This means the platform does not require a discrete GPU for basic display output, but the 16 PCIe lanes allow for one if needed. The memory bus is dual-channel, which is standard for mobile parts, and the 128.0 GB/s bandwidth figure is the aggregate across both channels.

FAQ

Q: What socket does the AMD Ryzen AI Max PRO 380 use?

A: It uses AMD Socket FP11, a mobile socket. The processor is soldered to the motherboard and is not user-swappable.

Q: Does the processor support ECC memory?

A: Yes. The FACT PACK lists eccMemory as true, meaning the memory controller supports error-correcting code memory.

Q: What integrated graphics does it have?

A: It has the Radeon 8040S integrated GPU, part of the Radeon 8000 series.

Q: How much cache does it have?

A: It has 16 MB of shared L3 cache, plus 80 KB of L1 cache per core and 1 MB of L2 cache per core.

Q: What is the release date and production status?

A: The release date is 2025-01-05, and the production status is Active.

Q: What memory type and bandwidth does it support?

A: It supports LPDDR5X memory in a dual-channel configuration with 128.0 GB/s of memory bandwidth.

Q: Is the multiplier unlocked?

A: No. The multiplier is locked (multiplierUnlocked is false), so the processor cannot be overclocked by adjusting the multiplier.

Single-Thread vs Multi-Thread Behavior

The Ryzen AI Max PRO 380 has 6 cores and 12 threads. This is a modest core count by 2025 standards — many mobile and desktop processors offer 8 or more cores. But the Zen 5 architecture is designed for strong single-thread performance. The 4.90 GHz boost clock is a key indicator: boost clocks are typically achieved on one or two cores, and 4.90 GHz is a high figure for a 55W mobile part.

The base clock of 3.60 GHz represents the sustained all-core frequency under load. The range from 3.60 GHz to 4.90 GHz shows that single-thread workloads can push much higher than sustained multi-thread loads. This is typical of modern processors: the boost clock is for lightly threaded tasks, while the base clock reflects the all-core sustained capability.

For real workloads, this means single-thread-heavy applications — web browsers, office suites, legacy software, and many games — will benefit from the 4.90 GHz boost. The processor can ramp a single core to its maximum and deliver snappy responsiveness. Multi-threaded applications — video rendering, 3D modeling, data analysis — will use all 6 cores and 12 threads at the 3.60 GHz base clock (or slightly above, depending on thermal headroom). The 12 threads provide parallel execution, but the 6-core count limits the total throughput compared to higher-core-count parts.

The 16 MB shared L3 cache is important for both scenarios. In single-thread workloads, the cache holds frequently accessed data close to the core, reducing memory latency. In multi-thread workloads, the shared cache allows all cores to access a common pool of data, which is efficient for workloads that share data between threads.

The 128.0 GB/s memory bandwidth is particularly relevant for multi-threaded workloads. When all 6 cores are active, they collectively demand more memory bandwidth. The dual-channel LPDDR5X configuration delivers 128.0 GB/s, which is a high figure for a mobile part and helps prevent memory bottlenecks. The ECC memory support and the Radeon 8040S iGPU round out the picture — the iGPU shares the memory bandwidth with the CPU, so in graphics-heavy workloads, the effective CPU bandwidth is reduced. Users running both CPU and GPU-intensive tasks simultaneously should be aware of this shared resource.

The 50th percentile ranking in the database suggests the processor's overall performance — a blend of single-thread and multi-thread — sits at the median of all tracked CPUs. This is a reasonable expectation for a 6-core, 12-thread mobile part with a high boost clock: it will feel fast in everyday use, which is largely single-thread bound, but it will not compete with high-core-count workstation parts in rendering or encoding, which are multi-thread bound.