AMD Ryzen 9 8940HX vs AMD Ryzen AI Embedded P164 Comparison

FAQ

Q: Which processor has more cores and threads?

A: The AMD Ryzen 9 8940HX has 16 cores and 32 threads, while the AMD Ryzen AI Embedded P164 has 8 cores and 16 threads. The Ryzen 9 doubles the core and thread counts of the embedded part.

Q: What is the difference in average benchmark scores?

A: The Ryzen 9 8940HX records an average benchmark score of 81,103, placing it in the 95th percentile of all CPUs. The Ryzen AI Embedded P164 scores 52,901 on average, which places it in the 91st percentile. This is a 53.3% gap in average score (81,103 versus 52,901).

Q: Which chip wins in single-threaded performance?

A: The Ryzen AI Embedded P164 wins in the PassMark single-thread test with a score of 4,029, compared to 3,874 for the Ryzen 9 8940HX. The delta is 3.8% in favor of the embedded processor.

Q: What are the TDP ratings of these two processors?

A: The Ryzen 9 8940HX has a TDP of 55, while the Ryzen AI Embedded P164 has a TDP of 28. The embedded chip is rated for roughly half the thermal envelope of the Ryzen 9.

Q: Do both processors support ECC memory?

A: No. The Ryzen AI Embedded P164 supports ECC memory, while the Ryzen 9 8940HX does not.

Q: Which processor has a higher boost clock?

A: The Ryzen 9 8940HX boosts to 5.30 GHz, while the Ryzen AI Embedded P164 boosts to 5.00 GHz. The Ryzen 9 holds a 0.30 GHz advantage in maximum boost clock.

Architecture Differences

The Ryzen 9 8940HX uses the Zen 4 architecture on the Dragon Range codename, built on a 5 nm process at TSMC. Its die size is listed as 2x 71 mm², with a transistor count of 13,140 million. The Ryzen AI Embedded P164 uses the Gorgon Point codename, based on a Zen 5 / Zen 5c generation, built on a 4 nm process at TSMC, with a die size of 233 mm². The embedded chip does not have a listed transistor count.

Cache configurations differ substantially. The Ryzen 9 has 64 KB of L1 per core, 1 MB of L2 per core, and 64 MB of L3. The Ryzen AI Embedded P164 has 80 KB of L1 per core, 1 MB of L2 per core, and only 8 MB of L3. The Ryzen 9's L3 cache is eight times larger, which supports its higher core count and multi-threaded workloads.

Memory support differs as well. The Ryzen 9 supports DDR5 only, while the embedded part supports both DDR5 and LPDDR5X. Both use a dual-channel memory bus, but memory bandwidth favors the embedded chip: 89.6 GB/s versus 83.2 GB/s. The Ryzen 9 uses PCIe Gen 5 with 28 lanes, while the embedded part uses PCIe Gen 4 with 16 lanes.

Integrated graphics also differ: the Ryzen 9 includes Radeon 610M, while the embedded part includes Radeon 880M. The Ryzen 9 has an unlocked multiplier, whereas the embedded processor is locked. The embedded part supports ECC memory, matching its embedded market positioning.

Where Each One Wins

The Ryzen 9 8940HX dominates the multi-threaded and compute-heavy segments of this comparison. It wins 9 of the 11 head-to-head benchmark comparisons. In data compression, it scores 650,984 versus 327,891, a 98.5% advantage. In integer math, it scores 189,221 versus 87,940, a 115.2% lead. In floating-point math, it scores 113,921 versus 55,799, a 104.2% lead. The find-prime-numbers test shows the largest gap: 251 versus 71, a 253.5% difference. These results align with the Ryzen 9's 16 cores, 32 threads, and 64 MB of L3 cache, all of which directly benefit parallel workloads.

The Ryzen AI Embedded P164 wins the single-threaded tests. It scores 4,029 in PassMark single-thread versus 3,874 for the Ryzen 9, a 3.8% margin. This advantage likely stems from its newer Zen 5 / Zen 5c architecture and higher per-core L1 cache of 80 KB. For workloads that rely on single-core frequency and per-thread efficiency, the embedded chip delivers slightly better results.

The Ryzen 9 also wins in the multithread PassMark test with 49,731 versus 25,889, a 92.1% lead. It wins in data encryption with 39,318 versus 16,055, a 144.9% margin. It wins in extended instructions with 47,802 versus 24,193, a 97.6% lead. It wins in random string sorting with 75,381 versus 34,801, a 116.6% lead. It wins in physics with 2,104 versus 1,210, a 73.9% lead. The only tests the embedded part wins are the two single-thread entries.

Specification Differences

The two processors differ across nearly every specification field. The Ryzen 9 8940HX uses 16 cores and 32 threads, while the Ryzen AI Embedded P164 uses 8 cores and 16 threads. Base clocks differ: 2.40 GHz for the Ryzen 9 versus 2.00 GHz for the embedded part. Boost clocks differ: 5.30 GHz versus 5.00 GHz. TDP differs: 55 versus 28.

The socket differs: AMD Socket FL1 for the Ryzen 9 versus AMD Socket FP8 for the embedded part. The architecture field shows Zen 4 for the Ryzen 9, while the embedded part lists no architecture field but a generation of Zen 5 / Zen 5c. The codenames differ: Dragon Range versus Gorgon Point. Process node differs: 5 nm versus 4 nm. Die size differs: 2x 71 mm² versus 233 mm². Transistor count is listed only for the Ryzen 9 at 13,140 million.

Cache differs: L1 per core is 64 KB for the Ryzen 9 versus 80 KB for the embedded part. L2 per core is identical at 1 MB. L3 is 64 MB for the Ryzen 9 versus 8 MB for the embedded part. Memory support differs: DDR5 only for the Ryzen 9 versus DDR5 and LPDDR5X for the embedded part. Memory bandwidth differs: 83.2 GB/s versus 89.6 GB/s. ECC support differs: false for the Ryzen 9 versus true for the embedded part. PCIe differs: Gen 5 with 28 lanes for the Ryzen 9 versus Gen 4 with 16 lanes for the embedded part. Integrated graphics differ: Radeon 610M versus Radeon 880M. Multiplier unlock status differs: true for the Ryzen 9 versus false for the embedded part. Release dates differ: 2025-04-22 for the Ryzen 9 versus 2026-03-08 for the embedded part.

Head-to-Head Benchmarks

The largest win for the Ryzen 9 8940HX comes in the find-prime-numbers test. It scores 251 against 71 for the embedded part, a 253.5% delta. This test stresses integer arithmetic and loop execution, and the Ryzen 9's additional cores and larger L3 cache provide a massive advantage.

Data encryption shows a 144.9% lead for the Ryzen 9: 39,318 versus 16,055. This workload benefits from parallel cryptographic operations, where the 16-core Ryzen 9 clearly outperforms the 8-core embedded chip. Random string sorting follows at 116.6%: 75,381 versus 34,801. Integer math shows a 115.2% lead: 189,221 versus 87,940. Floating-point math shows a 104.2% lead: 113,921 versus 55,799.

Data compression shows a 98.5% lead: 650,984 versus 327,891. Extended instructions show a 97.6% lead: 47,802 versus 24,193. The multithread test shows a 92.1% lead: 49,731 versus 25,889. Physics shows a 73.9% lead: 2,104 versus 1,210. In every multi-threaded workload, the Ryzen 9 roughly doubles or more than doubles the embedded part's score.

The single-thread tests are the only wins for the Ryzen AI Embedded P164. It scores 4,029 in PassMark single-thread versus 3,874 for the Ryzen 9, a 3.8% margin. The same result appears in the passmark_singlethread test, which is listed identically. This indicates that the embedded chip's Zen 5 architecture provides better per-thread efficiency, despite having fewer cores and a lower boost clock of 5.00 GHz versus 5.30 GHz. The Ryzen 9's higher boost clock does not translate into a single-thread win, likely due to the architectural generation difference.

Overall, the Ryzen 9 8940HX wins 9 of 11 head-to-head comparisons, with an average benchmark score of 81,103 versus 52,901 for the embedded part. The embedded processor's 91st percentile ranking and 4,029 single-thread score do not offset the Ryzen 9's overwhelming multi-threaded dominance. The data indicates a clear split: the Ryzen 9 for parallel compute, the embedded part for single-thread efficiency and lower power draw.