AMD Ryzen 9 9955HX vs AMD Ryzen AI Embedded P164 Comparison

Architecture Differences

The AMD Ryzen 9 9955HX and the AMD Ryzen AI Embedded P164 are both 4 nm TSMC parts, but they belong to different families with distinct design goals. The 9955HX is a 16-core, 32-thread processor built on the Zen 5 architecture under the Fire Range codename, part of the 9000 series. The P164 is an 8-core, 16-thread chip using a hybrid Zen 5 / Zen 5c configuration under the Gorgon Point codename, part of the Ryzen AI Embedded line.

The core count disparity is the single largest architectural divide. The 9955HX doubles the P164 in both cores and threads, which directly explains its dominance in every multithreaded workload. The P164's smaller 8 MB of shared L3 cache stands in stark contrast to the 9955HX's 64 MB shared L3, a 8x difference that affects how much data each processor can keep on-die for repeated access. Both chips use 80 KB of L1 per core and 1 MB of L2 per core, so the per-core cache hierarchy is identical; the difference lies entirely in the shared pool.

Clock speeds also favor the larger chip. The 9955HX has a 2.50 GHz base clock and a 5.40 GHz boost clock, while the P164 runs at 2.00 GHz base and 5.00 GHz boost. That 400 MHz base-clock gap and 400 MHz boost-clock gap contribute to the 9955HX's single-thread advantage, though the gap is far narrower than the multicore deltas.

The process node is the same 4 nm TSMC fabrication, and both chips use dual-channel memory with 89.6 GB/s bandwidth. Memory support differs: the 9955HX supports DDR5 only, while the P164 supports both DDR5 and LPDDR5X. Both support ECC memory. The 9955HX uses AMD Socket FL1, the P164 uses AMD Socket FP8. The 9955HX offers PCIe Gen 5 with 28 CPU lanes, whereas the P164 is limited to PCIe Gen 4 with 16 CPU lanes. The 9955HX integrates a Radeon 610M graphics unit; the P164 steps up to a Radeon 880M.

Power envelopes diverge sharply. The 9955HX has a 55 W TDP, the P164 a 28 W TDP. That 27 W difference reflects the embedded segment's focus on efficiency and thermal constraints versus the high-performance mobile segment. The 9955HX has an unlocked multiplier; the P164 does not. The 9955HX's die is described as 2x 70.6 mm² with 16,630 million transistors, while the P164 is a single 233 mm² die with no transistor count listed. The 9955HX was released in January 2025; the P164 is dated March 2026 in the database. Both are listed as active production parts.

The Verdict

The benchmark data points to a clear hierarchy: the 9955HX wins all 11 head-to-head tests, with a 117% to 304.2% margin across the workload suite. The P164 does not win a single recorded comparison. The 9955HX's average benchmark score is 91199 against the P164's 52901, a gap of roughly 72%. In percentile terms, the 9955HX sits at the 96th percentile of all CPUs, while the P164 ranks at the 91st percentile.

The 9955HX is the choice for workloads that demand maximum throughput: rendering, compilation, encryption, data compression, and physics simulation. Its 16 cores and 64 MB L3 give it an insurmountable lead in every multithreaded test. The P164, with 8 cores and 8 MB L3, is the part for environments where the 28 W TDP and smaller footprint matter more than raw performance. It also uses less power, which suits embedded deployments with tight thermal budgets.

The single-thread comparison is the only area where the two are close. The 9955HX scores 4393 in PassMark single-thread tests, the P164 scores 4029, a 9% delta. That is a meaningful but modest gap, meaning the P164 is not far behind in lightly threaded tasks. For mixed workloads where single-thread responsiveness matters alongside low power draw, the P164 remains viable. For anything that scales with cores, the 9955HX is the clear pick.

Neither part has a recorded launch MSRP in the database, so pricing cannot be factored into this analysis.

Head-to-Head Benchmarks

The largest single delta appears in PassMark's find prime numbers test. The 9955HX scores 287, the P164 scores 71, giving the 9955HX a 304.2% advantage. This test is highly sensitive to core count and memory bandwidth, and the 9955HX's 16 cores with 64 MB L3 simply overwhelm the P164's 8 cores with 8 MB L3.

Floating point math shows a 145% lead for the 9955HX: 136682 versus 55799. Integer math follows at 141.8%: 212598 versus 87940. Extended instructions come in at 139.5%: 57946 versus 24193. Data encryption shows a 132.5% delta: 37330 versus 16055. All of these are compute-heavy, multithreaded workloads where the core and thread advantage translates directly into higher scores.

Physics simulation delivers a 124.8% delta: 2720 versus 1210. Random string sorting shows 123.8%: 77890 versus 34801. Data compression is at 123.2%: 731998 versus 327891. The multithread aggregate score gives 117%: 56171 versus 25889. These deltas are all in the same ballpark, reflecting the consistent 2.2x to 2.4x throughput advantage of the 16-core part across diverse tasks.

The single-thread test is the outlier. The 9955HX scores 4393, the P164 scores 4029, a 9% delta. That is the only head-to-head result under double digits. The boost clock difference of 5.40 GHz versus 5.00 GHz, combined with the larger L3 cache, explains this edge. The P164's Zen 5c cores, if they are the smaller efficiency-oriented variant, may also contribute to the single-thread deficit, though the database does not specify which cores handle which clock rates.

FAQ

Q: Which processor has more cores?

A: The AMD Ryzen 9 9955HX has 16 cores and 32 threads. The AMD Ryzen AI Embedded P164 has 8 cores and 16 threads.

Q: How much L3 cache does each chip have?

A: The 9955HX has 64 MB of shared L3 cache. The P164 has 8 MB of L3 cache.

Q: What is the single-thread performance difference?

A: In PassMark single-thread tests, the 9955HX scores 4393 and the P164 scores 4029. The 9955HX leads by 9%.

Q: Which chip supports faster PCIe?

A: The 9955HX supports PCIe Gen 5 with 28 CPU lanes. The P164 supports PCIe Gen 4 with 16 CPU lanes.

Q: Do both processors support ECC memory?

A: Yes, both the 9955HX and the P164 support ECC memory.

Q: What is the TDP difference?

A: The 9955HX has a 55 W TDP. The P164 has a 28 W TDP.

Where Each One Wins

The 9955HX wins every benchmark category in the database. Its 304.2% lead in prime number finding and 145% lead in floating point math make it the obvious choice for number-crunching, scientific computing, and financial modeling. Data encryption at 132.5% ahead and integer math at 141.8% ahead position it well for security workloads, hashing, and general-purpose coding. The 123.2% advantage in data compression and 123.8% in random string sorting suit file archiving, database operations, and text processing. Physics simulation at 124.8% ahead covers game physics, structural analysis, and particle simulation.

The 9955HX's 64 MB L3 cache and 16 cores give it a decisive edge in any workload that can use more than 8 threads. Its 5.40 GHz boost clock also gives it the single-thread crown, though the 9% margin is modest compared to the multicore gaps. The unlocked multiplier means overclocking headroom exists, which could widen those deltas further in practice. The 55 W TDP is higher, but the performance per watt in multithreaded tasks is still superior based on the score ratios.

The P164 wins no benchmark tests in the database. Its strengths lie outside the recorded metrics: 28 W TDP, support for LPDDR5X memory, and a Radeon 880M integrated GPU. The 8 MB L3 cache is small, but the chip's 5.00 GHz boost clock keeps single-thread performance within 9% of the 9955HX. For embedded systems where power draw is the primary constraint, the P164 offers 80% of the single-thread performance at roughly half the TDP. The 16 PCIe Gen 4 lanes are sufficient for many embedded I/O needs, and the FP8 socket may fit smaller board designs.

The P164's 91st percentile ranking still places it above most CPUs, and its nearest rivals include the AMD Ryzen 5 9500F (0.1% delta), Intel Xeon 634 (-0.1%), AMD EPYC 7313P (-0.6%), and AMD Ryzen 9 7900X (-0.7%). The 9955HX's nearest rivals are the Intel Xeon 654 (0.5%), AMD Ryzen AI Max+ 392 (0.7%), Intel Core Ultra 7 270K Plus (-2.8%), and Intel Xeon 6520P (-2.8%). The 9955HX competes with workstation-class parts, while the P164 sits closer to desktop and embedded processors.

Specification Differences

| Specification | AMD Ryzen 9 9955HX | AMD Ryzen AI Embedded P164 |

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

| Cores | 16 | 8 |

| Threads | 32 | 16 |

| Base Clock | 2.50 GHz | 2.00 GHz |

| Boost Clock | 5.40 GHz | 5.00 GHz |

| TDP | 55 W | 28 W |

| Socket | AMD Socket FL1 | AMD Socket FP8 |

| Architecture | Zen 5 | Not specified |

| Codename | Fire Range | Gorgon Point |

| Generation | Ryzen 9 (Zen 5 (Fire Range)) | Ryzen AI Embedded (Zen 5 / Zen 5c) |

| Process Node | 4 nm | 4 nm |

| Die Size | 2x 70.6 mm² | 233 mm² |

| Transistors | 16,630 million | Not specified |

| L3 Cache | 64 MB (shared) | 8 MB |

| Memory Support | DDR5 | DDR5, LPDDR5X |

| Memory Bus | Dual-channel | Dual-channel |

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

| ECC Memory | Yes | Yes |

| PCIe | Gen 5, 28 Lanes (CPU only) | Gen 4, 16 Lanes (CPU only) |

| Integrated Graphics | Radeon 610M | Radeon 880M |

| Multiplier Unlocked | Yes | No |

| Release Date | 2025-01-05 | 2026-03-08 |

| Part Number | 100-000001028 | Unknown |

The L1 and L2 cache configurations are identical at 80 KB per core and 1 MB per core, respectively. Both use dual-channel memory with 89.6 GB/s bandwidth. The process node and foundry (TSMC 4 nm) are shared. The 9955HX has a significantly larger L3 pool and twice the core count, while the P164 offers a more power-efficient package with a newer integrated GPU and broader memory type support.