AMD Ryzen AI Embedded P164 vs Intel Core Ultra 5 235HX Comparison

AMD
AMD

AMD Ryzen AI Embedded P164

CORE STATE Gorgon Point
CORE SPECS 8 Cores / 16 Threads
CLOCK SPEED 2 Base / 5 GHz Turbo
CACHE 8 MB
MAX TDP 28W
ARCHITECTURE Gorgon Point
nm
PROCESS 4 nm
LAUNCH DATE 2026
VS
Intel
INTEL

Core Ultra 5 235HX

CORE STATE Arrow Lake-HX
CORE SPECS 14 Cores / 14 Threads
CLOCK SPEED 2.9 Base / 5.1 GHz Turbo
CACHE 24 MB (shared)
MAX TDP 55W
ARCHITECTURE Arrow Lake
nm
PROCESS 3 nm
LAUNCH DATE 2025

PERFORMANCE BENCHMARKS

passmark_data_compression
327,891
426,417
passmark_data_encryption
16,055
32,697
passmark_extended_instructions
24,193
34,808
passmark_find_prime_numbers
71
361
passmark_floating_point_math
55,799
129,819
passmark_integer_math
87,940
97,177
passmark_multithread
25,889
40,849
passmark_physics
1,210
2,550
passmark_random_string_sorting
34,801
50,929
passmark_single_thread
4,029
4,683
passmark_singlethread
4,029
4,683
cinebench_cinebench_r15_multicore
N/A
3,500
cinebench_cinebench_r15_singlecore
N/A
494
cinebench_cinebench_r20_multicore
N/A
14,587
cinebench_cinebench_r20_singlecore
N/A
2,059
cinebench_cinebench_r23_multicore
N/A
34,731
cinebench_cinebench_r23_singlecore
N/A
4,903

Analysis: AMD Ryzen AI Embedded P164 vs Intel Core Ultra 5 235HX

The AMD Ryzen AI Embedded P164 and the Intel Core Ultra 5 235HX are both mobile processors aimed at high-performance computing, but the benchmark data reveals a starkly one-sided contest. Across the eleven head-to-head PassMark tests, the Intel Core Ultra 5 235HX secures a clean sweep, winning every single discipline. The data shows the AMD part’s average benchmark score of 52,901 is nearly identical to the Intel part’s 52,073, yet the individual workload results tell a story of divergent architectural priorities. The Core Ultra 5 235HX is decisively faster in almost every measurable category, while the Ryzen AI Embedded P164 offers a significantly lower power envelope and a unique feature set, making the choice between them a matter of workload and system design rather than raw performance parity.

Head-to-Head Benchmarks

The most dramatic disparity emerges in the `passmark_find_prime_numbers` test, where the Intel Core Ultra 5 235HX scores 361 against the AMD Ryzen AI Embedded P164’s 71. This represents an 80.3% deficit for the AMD part, indicating a massive advantage for Intel in this particular integer-heavy workload. Similarly, the `passmark_floating_point_math` test shows Intel at 129,819 versus AMD’s 55,799, a 57% lead. These results suggest the Intel chip’s higher core count and newer microarchitecture are particularly well-suited to mathematical and scientific computations.

The gap narrows somewhat in more general-purpose tasks, but Intel still maintains a commanding lead. In `passmark_data_encryption`, Intel scores 32,697 against AMD’s 16,055, a 50.9% advantage. The `passmark_physics` test shows Intel at 2,550 versus 1,210, a 52.5% margin. Even in `passmark_integer_math`, where the difference is smallest, Intel’s 97,177 beats AMD’s 87,940 by 9.5%. The `passmark_multithread` score reinforces the trend: Intel’s 40,849 is 36.6% higher than AMD’s 25,889, which aligns with Intel’s 14 cores versus AMD’s 8.

Single-thread performance is closer, but Intel still holds a 14% edge. The `passmark_single_thread` test shows Intel at 4,683 versus AMD’s 4,029. This is reflected in the `passmark_extended_instructions` (34,808 vs 24,193, a 30.5% gap) and `passmark_random_string_sorting` (50,929 vs 34,801, a 31.7% gap) tests. The only area where the AMD part comes within single digits of its rival is `passmark_data_compression`, where Intel’s 426,417 beats AMD’s 327,891 by 23.1%. Across all eleven tests, the average delta is roughly 36% in Intel’s favor, with zero wins for the AMD processor.

Where Each One Wins

Based strictly on the benchmark results, the Intel Core Ultra 5 235HX is the clear winner for compute-intensive applications. Its dominance in floating-point math, prime number finding, and encryption makes it the superior choice for scientific simulation, financial modeling, cryptography, and any workload that relies heavily on parallel mathematical operations. The 36.6% lead in multithreaded performance and the 52.5% advantage in physics suggest it is also better suited for 3D rendering, video encoding, and complex physics simulations in engineering or content creation. The 14% single-thread edge further indicates it will feel snappier in everyday responsiveness and lightly-threaded applications like office suites or web browsing.

The AMD Ryzen AI Embedded P164, despite losing every benchmark, still has a role. Its 28-watt TDP is roughly half of the Intel part’s 55 watts, making it the more power-efficient option for thermally constrained or battery-powered systems. The data does not show a single test where the AMD chip wins, so any advantage must be inferred from its specifications rather than performance numbers. It supports ECC memory, which the Intel chip does not, making it the only option for systems requiring error-correcting memory for data integrity. The AMD part also features a Radeon 880M integrated GPU, while the Intel chip has an Arc Xe-LPG Graphics 48EU, but neither GPU’s performance is quantified in the provided benchmarks, so no direct comparison can be made.

Architecture Differences

The two processors represent fundamentally different design philosophies. The AMD Ryzen AI Embedded P164 is built on a 4 nm TSMC process with a die size of 233 mm², while the Intel Core Ultra 5 235HX uses a smaller 3 nm TSMC node with a slightly larger 243 mm² die. Intel’s die packs 17,800 million transistors, whereas AMD’s transistor count is not listed. The AMD chip is based on the Gorgon Point codename with a Zen 5 / Zen 5c generation, while Intel’s Arrow Lake-HX architecture is part of the Core Ultra Series 2.

Core configurations differ markedly. The AMD part features 8 cores and 16 threads, while the Intel chip has 14 cores but only 14 threads, indicating Intel is relying on a mix of performance and efficiency cores without hyperthreading. Cache hierarchies also diverge: AMD provides 80 KB of L1 and 1 MB of L2 per core, with 8 MB of shared L3, whereas Intel offers 192 KB of L1 and 3 MB of L2 per core, with a much larger 24 MB shared L3 cache. This larger L3 cache likely contributes to Intel’s performance advantage in data-heavy tasks.

Clock speeds are comparable on paper: the AMD part has a 2.00 GHz base and 5.00 GHz boost, while the Intel chip has a 2.90 GHz base and 5.10 GHz boost. However, Intel’s higher base clock and additional cores give it a significant throughput advantage. Memory support also differs: AMD supports both DDR5 and LPDDR5X with a bandwidth of 89.6 GB/s, while Intel only lists DDR5 with a higher 102.4 GB/s bandwidth. Both use dual-channel memory buses, but Intel’s PCIe implementation is newer (Gen 5 with 20 lanes) compared to AMD’s Gen 4 with 16 lanes. AMD’s socket is FP8, while Intel uses BGA 2114. The Intel chip is multiplier unlocked, allowing overclocking, while the AMD part is locked. Neither chip has a listed launch MSRP.

The Verdict

From the benchmark data alone, the Intel Core Ultra 5 235HX is the superior processor for virtually all performance-oriented tasks. It wins 100% of the head-to-head tests, with margins ranging from 9.5% in integer math to 80.3% in prime number finding. Its 14 cores, larger cache, and higher memory bandwidth make it the obvious pick for users who prioritize raw speed in multithreaded applications, content creation, or heavy computation. The 14% single-thread advantage also means it excels in mixed workloads where both single and multi-core performance matter.

The AMD Ryzen AI Embedded P164, however, should be selected for systems where power consumption and data integrity are paramount. Its 28-watt TDP is less than half of Intel’s 55 watts, which is critical for fanless designs, embedded systems, or portable devices where battery life and thermal output are more important than benchmark scores. The inclusion of ECC memory support is a decisive factor for servers, network appliances, or any application where a memory error could cause data corruption. If the workload is light or the system is power-constrained, the AMD part is the logical choice despite its performance deficit.

FAQ

Q: Which processor has a higher multithreaded benchmark score?

A: The Intel Core Ultra 5 235HX scores 40,849 in the PassMark multithread test, which is 36.6% higher than the AMD Ryzen AI Embedded P164’s 25,889.

Q: Are there any benchmark tests where the AMD Ryzen AI Embedded P164 wins?

A: No. Across all eleven head-to-head PassMark tests, the Intel Core Ultra 5 235HX wins every single one, with the AMD part recording zero victories.

Q: What is the largest performance margin between the two processors?

A: The largest gap is in the `passmark_find_prime_numbers` test, where the Intel chip scores 361 versus the AMD chip’s 71, an 80.3% difference.

Q: Does either processor support ECC memory?

A: Yes, the AMD Ryzen AI Embedded P164 supports ECC memory. The Intel Core Ultra 5 235HX does not list ECC support.

Q: How do the thermal design power (TDP) ratings compare?

A: The AMD Ryzen AI Embedded P164 has a TDP of 28 watts, while the Intel Core Ultra 5 235HX has a TDP of 55 watts.

Q: Which processor has more cores and threads?

A: The Intel Core Ultra 5 235HX has 14 cores and 14 threads. The AMD Ryzen AI Embedded P164 has 8 cores and 16 threads.

DETAILED SPECIFICATIONS

SPECIFICATION
AI Embedded P164
Ultra 5 235HX
Core Specs
Cores
8
14 +75.0%
Threads
16
14 -12.5%
Base Clock (GHz)
2
2.9 +45.0%
Boost Clock (GHz)
5
5.1 +2.0%
Frequency (GHz)
2
2.9 +45.0%
Turbo Clock (GHz)
5
5.1 +2.0%
Multiplier
20
29 +45.0%
SMP CPUs
1
1 0.0%
Cache
L1 Cache
80 KB (per core)
192 KB (per core)
L2 Cache
1 MB (per core)
3 MB (per core)
L3 Cache
8 MB
24 MB (shared)
Power
TDP (W)
28
55 +96.4%
PL1
—
55 W
PL2
—
160 W
Configurable TDP
15-54 W
—
Architecture
Architecture
—
Arrow Lake
Codename
Gorgon Point
Arrow Lake-HX
Generation
Ryzen AI Embedded (Zen 5 / Zen 5c)
Ultra 5 (Arrow Lake-HX)
Process Size
4 nm
3 nm
Transistors
—
17,800 million
Die Size
233 mm²
243 mm²
Foundry
TSMC
TSMC
Memory
Memory Support
DDR5, LPDDR5X
DDR5
Memory Bus
Dual-channel
Dual-channel
Memory Bandwidth
89.6 GB/s
102.4 GB/s
ECC Memory
Yes
No
Platform
Socket
AMD Socket FP8
Intel BGA 2114
Chipsets
—
WM880, HM870
PCIe
Gen 4, 16 Lanes(CPU only)
Gen 5, 20 Lanes(CPU only)
Intel Hybrid
Hybrid Cores
3 + 5
P-Cores: 6 E-Cores: 8
E-Core Frequency
2000 MHz up to 3.3 GHz
2.6 GHz up to 4.5 GHz
AI/NPU
NPU
Yes / 50 TOPS
Yes / 13 TOPS
Graphics
Integrated Graphics
Radeon 880M
Arc Xe-LPG Graphics 48EU
Other
Market
Mobile
Mobile
Production Status
Active
Active
Part Number
unknown
SRVFL
Package
FP8
FC-BGA
Tj Max
105°C
105°C
View Ryzen AI Embedded P164 Details View Core Ultra 5 235HX Details