AMD Ryzen AI Max+ 392 vs Intel Core Ultra 5 235H Comparison

Head-to-Head Benchmarks

The recorded data presents a clear split between the AMD Ryzen AI Max+ 392 and the Intel Core Ultra 5 235H. Across the eleven head-to-head PassMark tests, the AMD part claims nine wins, while the Intel part takes two. The margins, however, are not uniform. The AMD Ryzen AI Max+ 392 shows its dominance in multi-threaded and data-heavy workloads, while the Intel Core Ultra 5 235H demonstrates a notable advantage in single-thread performance.

The largest delta in the entire comparison is in integer math. The AMD Ryzen AI Max+ 392 scores 152414 against the Intel Core Ultra 5 235H's 74247, a 105.3% advantage. This is more than double the output, indicating a substantial difference in raw integer processing capability. Extended instructions follow a similar pattern, with the AMD part scoring 45666 versus 23354, a 95.5% delta. Data compression also heavily favors the AMD chip, with a score of 554760 against 301979, an 83.7% gap. These three tests alone establish the AMD processor as the clear leader for compute-intensive tasks that scale with core count and cache.

The lead continues, albeit with smaller margins, in other multi-threaded metrics. Random string sorting shows the AMD part at 59487 versus 36208, a 64.3% advantage. Multithread performance, a general indicator of parallel workload capability, shows 45231 for AMD against 30091 for Intel, a 50.3% delta. Physics calculations also favor the AMD chip, with 2887 versus 1985, a 45.4% lead. Prime number finding, another test that scales with parallel execution, shows a 33.9% delta, with scores of 320 and 239 respectively. Data encryption completes the AMD sweep of the data-processing suite, with 27784 versus 23121, a 20.2% advantage.

The narrowest margin for the AMD processor is in floating point math. The AMD Ryzen AI Max+ 392 scores 99548, while the Intel Core Ultra 5 235H scores 93509. The delta here is just 6.5%, which is a much closer contest than the integer math results. This suggests that while the AMD architecture is superior for integer operations, the gap narrows significantly when dealing with fractional calculations.

The Intel Core Ultra 5 235H secures its two wins in the single-threaded tests. Both the passmark_single_thread and passmark_singlethread entries record identical scores of 4359 for Intel and 3927 for AMD. This results in a 9.9% delta in favor of the Intel part. This is the only category where the Intel chip demonstrates a clear performance edge, and it is a consistent one, as both entries for the same test confirm the same outcome.

The average benchmark score for the AMD Ryzen AI Max+ 392 is 90541, placing it in the 96th percentile of all CPUs in the database. In contrast, the Intel Core Ultra 5 235H has an average score of 37522, which places it in the 85th percentile. The gap in average scores is substantial, reflecting the AMD chip's placement among higher-performing desktop and workstation processors. The AMD part's nearest rivals, such as the Intel Xeon 654 and AMD Ryzen 9 9955HX, have average scores of 90717 and 91199 respectively, showing a delta of -0.2% and -0.7% against the AMD chip. This places the AMD Ryzen AI Max+ 392 in a performance tier well above the Intel Core Ultra 5 235H, whose nearest rivals, like the Intel Core i9-13900HK and Intel Core i5-13600K, have average scores of 37425 and 37685, with deltas of 0.3% and -0.4% respectively.

Where Each One Wins

The benchmark data suggests distinct use-case strengths for each processor. The AMD Ryzen AI Max+ 392 is the clear choice for workloads that demand high throughput in parallel processing. Its wins in data compression, encryption, integer math, and extended instructions indicate that it is well-suited for tasks like file archiving, cryptographic operations, code compilation, and scientific simulations that can leverage multiple cores. The 105.3% lead in integer math is particularly relevant for database operations, financial modeling, and other integer-heavy applications. The 83.7% lead in data compression also points to strong performance in data storage and transfer scenarios.

The Intel Core Ultra 5 235H, with its 9.9% lead in single-thread performance, is better positioned for tasks that are not easily parallelized. This includes legacy software, certain types of scripting, and applications with a primary single-threaded execution path. The higher single-thread score of 4359 suggests a snappier response in daily tasks like web browsing, document editing, and general desktop navigation where a single core often dictates perceived speed.

The multithread score of 45231 for the AMD part against 30091 for Intel reinforces the AMD processor's suitability for content creation, 3D rendering, and video encoding. These tasks typically scale well with core count and benefit from the larger cache and memory bandwidth. The Intel part, with a TDP of 28 watts, might be more appropriate for thin-and-light laptops where power efficiency is a priority, but the benchmark data does not quantify that efficiency directly, only the performance outcomes.

The physics test, where AMD leads by 45.4%, is another indicator of strength in simulation and gaming physics calculations. While the integrated graphics differ, the CPU-side physics processing is a separate metric, and the AMD part's lead here suggests a stronger foundation for gaming laptops that handle complex physics engines. The Intel part's wins are isolated to single-thread, making it a niche performer in that specific area. For any workload that can utilize more than a few cores, the data overwhelmingly favors the AMD Ryzen AI Max+ 392.

Architecture Differences

The two processors are built on fundamentally different architectures, which explains the observed benchmark results. The AMD Ryzen AI Max+ 392 uses the Zen 5 architecture, codenamed Strix Halo, and is manufactured on a 4 nm process node by TSMC. The Intel Core Ultra 5 235H uses the Arrow Lake architecture, codenamed Arrow Lake-H, and is built on a 3 nm process node, also by TSMC. The smaller process node for Intel does not translate into a performance advantage in the recorded data, as the AMD part's architectural design proves more effective in multi-threaded tests.

Core configuration differs significantly. The AMD part has 12 cores and 24 threads, supporting simultaneous multithreading. The Intel part has 14 cores but only 14 threads, meaning it lacks hyper-threading. This is a critical difference. The AMD chip can process two threads per core, effectively doubling its thread count to 24. The Intel chip, despite having two more physical cores, is limited to 14 threads. This explains the substantial lead in multithreaded benchmarks for the AMD part, as it can schedule and execute more concurrent workloads.

Cache hierarchies also diverge. The AMD Ryzen AI Max+ 392 has an L1 cache of 80 KB per core, an L2 cache of 1 MB per core, and a shared L3 cache of 64 MB. The Intel Core Ultra 5 235H has a larger L1 cache at 192 KB per core and a larger L2 cache at 3 MB per core, but its shared L3 cache is only 18 MB. The AMD part's 64 MB L3 cache is over three times larger, which is a significant factor in its data compression and integer math wins. A larger shared cache reduces the need to access slower system memory, which is a distinct advantage for data-intensive workloads.

Memory support further separates the two. The AMD part uses LPDDR5X memory with a quad-channel bus, offering a memory bandwidth of 256.0 GB/s. The Intel part supports both DDR5 and LPDDR5X, but on a dual-channel bus, with a memory bandwidth of 102.4 GB/s. The AMD part's memory bandwidth is exactly double that of the Intel part. This is a massive advantage for the AMD processor in any workload that is memory-bandwidth bound, such as data compression, encryption, and large dataset manipulations. The AMD part also supports ECC memory, while the Intel part does not.

The integrated graphics differ as well. The AMD part features the Radeon 8060S, while the Intel part features the Arc Graphics 140T. The PCIe support also varies, with the AMD part offering Gen 4 with 16 lanes, and the Intel part offering Gen 5 with 8 lanes. The AMD part is built on the AMD Socket FP11, while the Intel part uses Intel BGA 2049. The base clock for the AMD part is 3.20 GHz, and for the Intel part it is 2.40 GHz. Both parts have a boost clock of 5.00 GHz. The TDP also differs, with the AMD part rated at 55 watts and the Intel part at 28 watts, which is a notable difference in power envelope. Neither processor has an unlocked multiplier.

The Verdict

The benchmark data provides a definitive answer based on performance metrics. The AMD Ryzen AI Max+ 392 is the superior processor for almost all computational tasks measured. Its 9 out of 11 head-to-head wins, including a 105.3% lead in integer math and an 83.7% lead in data compression, make it the clear choice for demanding, parallelized workloads. The processor's 96th percentile ranking and average score of 90541 place it among the top-tier CPUs in the database, on par with workstation-class Xeon processors. The Intel Core Ultra 5 235H, while ranking in the 85th percentile with an average score of 37522, is in a completely different performance class.

The Intel part's sole advantage lies in single-thread performance, where it leads by 9.9%. For a user whose primary applications are strictly single-threaded and where higher multi-core performance is not needed, the Intel Core Ultra 5 235H offers a specific benefit. However, the data shows that this advantage is limited to a single metric. For any workload that can utilize multiple cores or threads, the AMD Ryzen AI Max+ 392 delivers significantly higher performance. The multithread score of 45231 versus 30091, a 50.3% delta, is a clear indicator of the AMD part's dominance in modern, multi-threaded software environments.

The architecture confirms this split. The AMD part's 64 MB L3 cache, quad-channel memory with 256.0 GB/s bandwidth, and 24 threads are designed for heavy lifting. The Intel part, with its 18 MB L3 cache, dual-channel memory at 102.4 GB/s, and 14 threads, is more constrained in parallel environments. The data suggests that the AMD Ryzen AI Max+ 392 is the better choice for content creation, software development, data analysis, and any other task that demands sustained multi-core performance. The Intel Core Ultra 5 235H is a competent processor, but its performance profile is better suited for lighter, single-thread-dominant tasks. The recorded benchmarks indicate a clear hierarchy: the AMD Ryzen AI Max+ 392 is the higher-performing processor, and the Intel Core Ultra 5 235H is its lower-performing counterpart in multi-threaded work.

FAQ

Q: What is the biggest performance difference between the AMD Ryzen AI Max+ 392 and the Intel Core Ultra 5 235H?

A: The largest delta is in integer math, where the AMD Ryzen AI Max+ 392 scores 152414 against 74247 for the Intel Core Ultra 5 235H, a 105.3% advantage.

Q: Does the Intel Core Ultra 5 235H win in any benchmark?

A: Yes, the Intel Core Ultra 5 235H wins in the single-thread test, scoring 4359 against 3927 for the AMD Ryzen AI Max+ 392, a 9.9% delta in favor of Intel.

Q: How does the core and thread configuration differ between the two processors?

A: The AMD Ryzen AI Max+ 392 has 12 cores and 24 threads, while the Intel Core Ultra 5 235H has 14 cores and 14 threads. The AMD part supports simultaneous multithreading, while the Intel part does not.

Q: What is the difference in L3 cache size?

A: The AMD Ryzen AI Max+ 392 has 64 MB of shared L3 cache, while the Intel Core Ultra 5 235H has 18 MB of shared L3 cache.

Q: How does memory bandwidth compare?

A: The AMD Ryzen AI Max+ 392 uses a quad-channel LPDDR5X memory bus with 256.0 GB/s bandwidth, while the Intel Core Ultra 5 235H uses a dual-channel DDR5 or LPDDR5X bus with 102.4 GB/s bandwidth.

Q: What are the average benchmark scores and percentiles for each processor?

A: The AMD Ryzen AI Max+ 392 has an average benchmark score of 90541 and is in the 96th percentile of all CPUs. The Intel Core Ultra 5 235H has an average score of 37522 and is in the 85th percentile.