AMD Ryzen AI Embedded P164 vs Intel Core Ultra 7 366H Comparison

FAQ

Q: Which processor has the higher overall benchmark average?

A: The AMD Ryzen AI Embedded P164 records an average benchmark score of 52901, while the Intel Core Ultra 7 366H records 41263. The AMD part also sits in the 91st percentile of all CPUs, compared to the Intel part's 87th percentile.

Q: How does the AMD chip compare to its nearest rivals?

A: The AMD Ryzen AI Embedded P164 is nearly identical to the AMD Ryzen 5 9500F, with a delta of 0.1%. It trails the Intel Xeon 634 by 0.1%, the AMD EPYC 7313P by 0.6%, and the AMD Ryzen 9 7900X by 0.7%.

Q: What are the closest competitors to the Intel Core Ultra 7 366H?

A: The Intel Core Ultra 7 356H matches it within 0.1%, the AMD Ryzen AI 5 PRO 440 also matches within 0.1%, the AMD Ryzen 9 5900X is 0.3% ahead, and the Intel Core Ultra X7 358H trails by 0.7%.

Q: Which chip wins in single-threaded performance?

A: The Intel Core Ultra 7 366H wins the PassMark single-thread test with a score of 4043 versus 4029 for the AMD Ryzen AI Embedded P164, a margin of just 0.3%.

Q: How many benchmark wins does each processor claim?

A: The Intel Core Ultra 7 366H wins 9 of the 11 head-to-head benchmark comparisons. The AMD Ryzen AI Embedded P164 wins only 2: data compression and integer math.

Q: Do both processors support the same memory types?

A: Yes, both support DDR5 and LPDDR5X memory in a dual-channel configuration. However, the Intel chip has a higher recorded memory bandwidth of 115.2 GB/s versus 89.6 GB/s for the AMD chip.

Architecture Differences

The AMD Ryzen AI Embedded P164 is built on a 4 nm process at TSMC with a die size of 233 mm². It uses the Gorgon Point codename and belongs to the Ryzen AI Embedded generation based on Zen 5 and Zen 5c cores. The Intel Core Ultra 7 366H uses a 3 nm process at Intel with the Panther Lake codename and architecture, belonging to the Ultra 7 Panther Lake-H generation.

Core counts differ significantly. The AMD chip has 8 cores and 16 threads, while the Intel chip has 16 cores and 16 threads. Both have a base clock of 2.00 GHz, but the AMD part boosts to 5.00 GHz versus 4.80 GHz for Intel. Cache hierarchies also differ substantially. AMD uses 80 KB L1 per core, 1 MB L2 per core, and 8 MB L3. Intel uses 192 KB L1 per core, 2.5 MB L2 per core, and 18 MB of shared L3.

The integrated graphics differ: AMD provides Radeon 880M while Intel provides Xe3 Graphics. AMD supports ECC memory, Intel does not. PCIe connectivity also differs, with AMD offering Gen 4 with 16 CPU lanes and Intel offering Gen 5 with 12 CPU lanes. Sockets are distinct: AMD Socket FP8 versus Intel BGA 2540.

Process node and foundry differences are notable. The 3 nm Intel process is smaller than the 4 nm TSMC process used by AMD. The Intel part has a lower TDP of 25 watts compared to 28 watts for AMD. Neither processor has an unlocked multiplier. The Intel part carries part number SA4R9Q9EL, while the AMD part's part number is listed as unknown.

The release dates differ, with Intel launching earlier. The Intel Core Ultra 7 366H has a release date in early January 2026, while the AMD Ryzen AI Embedded P164 followed in early March 2026. Both are classified as mobile market segments and both are marked as active in production.

Head-to-Head Benchmarks

The Intel Core Ultra 7 366H dominates the head-to-head comparison, winning 9 of 11 tests. The largest Intel victory comes in the find prime numbers test, where Intel scores 326 versus AMD's 71, a delta of 78.2% in Intel's favor. This is a dramatic gap that suggests fundamental differences in how the two architectures handle this workload.

Floating point math also heavily favors Intel, with a score of 103615 versus 55799, a 46.1% difference. Physics benchmarks show Intel at 2880 versus AMD's 1210, a 58% advantage. Multithread performance favors Intel at 33429 versus 25889, a 22.6% gap. Data encryption shows Intel at 25845 versus 16055, a 37.9% difference. Extended instructions favor Intel by 10.1%, random string sorting favors Intel by 12.6%, and single-thread performance favors Intel by just 0.3%.

The AMD Ryzen AI Embedded P164 wins two tests. Integer math shows AMD at 87940 versus Intel's 83695, a 5.1% advantage. Data compression is nearly identical, with AMD at 327891 versus Intel at 327455, a razor-thin 0.1% margin.

The single-thread result is remarkably close at 4029 versus 4043, suggesting that both chips have similar per-core performance ceilings despite their architectural differences. The multithread gap of 22.6% is more substantial and likely reflects the Intel chip's 16 cores versus AMD's 8 cores.

The data compression result deserves attention. Both processors land within 436 points of each other, which is less than 0.2% difference. This indicates that for compression workloads, the architectural differences largely cancel out. The integer math win for AMD shows that the Zen 5 design retains an edge in certain integer-heavy operations, even with fewer cores.

Specification Differences

| Specification | AMD Ryzen AI Embedded P164 | Intel Core Ultra 7 366H |

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

| Cores | 8 | 16 |

| Threads | 16 | 16 |

| Boost Clock | 5.00 GHz | 4.80 GHz |

| TDP | 28 W | 25 W |

| Socket | AMD Socket FP8 | Intel BGA 2540 |

| Process Node | 4 nm | 3 nm |

| Foundry | TSMC | Intel |

| L1 Cache | 80 KB per core | 192 KB per core |

| L2 Cache | 1 MB per core | 2.5 MB per core |

| L3 Cache | 8 MB | 18 MB shared |

| Memory Bandwidth | 89.6 GB/s | 115.2 GB/s |

| ECC Memory | Yes | No |

| PCIe | Gen 4, 16 lanes | Gen 5, 12 lanes |

| Integrated Graphics | Radeon 880M | Intel Xe3 Graphics |

| Release Date | 2026-03-08 | 2026-01-04 |

| Part Number | unknown | SA4R9Q9EL |

Both processors share the same base clock of 2.00 GHz, dual-channel memory bus, DDR5 and LPDDR5X memory support, and are both locked (multiplier not unlocked). Neither has an unlocked multiplier. The Intel chip has a smaller process node and larger caches across the board. The AMD chip has a higher boost clock and ECC support.

The memory bandwidth difference is notable: Intel's 115.2 GB/s is about 28.7% higher than AMD's 89.6 GB/s. This could influence memory-sensitive workloads. PCIe generation also differs, with Intel on Gen 5 but fewer lanes, while AMD offers Gen 4 with more lanes.

The Verdict

The recorded data presents a clear separation between these two processors. The Intel Core Ultra 7 366H wins the vast majority of benchmark comparisons, including the heavily weighted multithread and floating point tests. Its 16 cores deliver a 22.6% lead in multithread performance, and its physics score is more than double the AMD chip's result.

The AMD Ryzen AI Embedded P164 holds advantages in integer math and data compression, but these are narrow wins. The 5.1% integer math lead and the 0.1% data compression edge do not offset the substantial Intel leads in encryption, prime finding, physics, and floating point operations.

The average benchmark scores reinforce this hierarchy. AMD's 52901 average sits in the 91st percentile, while Intel's 41263 sits in the 87th percentile. However, the average score includes data beyond the head-to-head tests, and the Intel chip's own nearest rivals cluster around the 41000 to 41376 range, indicating consistent positioning.

Users seeking maximum throughput in multithreaded integer or floating point workloads should favor the Intel part based on the data. Users who require ECC memory support or who have workloads dominated by integer math should consider the AMD part. The AMD chip also offers a higher boost clock, which may matter in short burst workloads.

The single-thread results are essentially tied, meaning both chips will feel similar in lightly threaded applications. The decision rests on workload type rather than raw clock speed.

Where Each One Wins

AMD Ryzen AI Embedded P164 wins in:

• Integer math operations, with a 5.1% lead over Intel
• Data compression, with a marginal 0.1% edge
• ECC memory support, which Intel does not offer
• Higher boost clock at 5.00 GHz versus 4.80 GHz
• Overall average benchmark score and percentile ranking

Intel Core Ultra 7 366H wins in:

• Multithread performance, leading by 22.6%
• Floating point math, leading by 46.1%
• Physics calculations, leading by 58%
• Data encryption, leading by 37.9%
• Prime number finding, leading by 78.2%
• Extended instructions, leading by 10.1%
• Random string sorting, leading by 12.6%
• Single-thread performance, leading by 0.3%
• Memory bandwidth at 115.2 GB/s versus 89.6 GB/s
• Larger L3 cache at 18 MB versus 8 MB
• Smaller process node at 3 nm versus 4 nm

The data indicates that the Intel chip is the stronger all-around performer for compute-heavy tasks, while the AMD chip holds specific niches in integer workloads and memory integrity features. The Intel part's 16 cores provide a structural advantage in parallel workloads, while the AMD part's higher boost clock and ECC support serve specialized use cases.