AMD Ryzen AI 9 365 vs Intel Core Ultra 9 285H Comparison
AMD Ryzen AI 9 365
Core Ultra 9 285H
PERFORMANCE BENCHMARKS
Analysis: AMD Ryzen AI 9 365 vs Intel Core Ultra 9 285H
FAQ
Q: Which processor has more cores and threads?
A: The Intel Core Ultra 9 285H has 16 cores and 16 threads, while the AMD Ryzen AI 9 365 has 10 cores and 20 threads. The Intel part uses a 1:1 core-to-thread ratio, whereas the AMD part supports simultaneous multithreading, giving it more threads than cores.
Q: Which chip wins in multi-threaded Cinebench workloads?
A: The Intel Core Ultra 9 285H leads in both Cinebench R23 multi-core (20781.5 vs 18698, a 10% advantage) and Cinebench R15 multi-core (3177.5 vs 2842, a 10.6% advantage). The AMD Ryzen AI 9 365 trails in both recorded multi-core Cinebench tests.
Q: Is there any benchmark where the AMD Ryzen AI 9 365 beats the Intel chip?
A: Yes. The AMD part wins in Geekbench single-core (2253 vs 2178, a 3.4% lead), PassMark data compression (354510 vs 335859, a 5.6% lead), and PassMark integer math (101831 vs 85922, an 18.5% lead). These are the only three head-to-head wins for AMD.
Q: How large is the Intel chip's advantage in floating-point math?
A: The Intel Core Ultra 9 285H scores 109190 in PassMark floating-point math versus 62802 for the AMD Ryzen AI 9 365, a 42.5% difference. This is the largest single benchmark gap between the two processors in either direction.
Q: What are the process nodes and foundries for each processor?
A: The AMD Ryzen AI 9 365 uses a 4 nm process from TSMC with a die size of 233 mm². The Intel Core Ultra 9 285H uses a 3 nm process, also from TSMC; its die size is not recorded in the database.
Q: Do these processors support ECC memory?
A: The Intel Core Ultra 9 285H supports ECC memory. The AMD Ryzen AI 9 365 does not. Both support DDR5 and LPDDR5X memory over a dual-channel bus.
Architecture Differences
The AMD Ryzen AI 9 365 belongs to the Ryzen AI 300 generation, built on the Zen 5 architecture with the Strix Point codename. It uses a hybrid arrangement of Zen 5 and Zen 5c cores, which the database lists as part of its generation string. The processor has 10 cores and 20 threads, with base clock at 2.00 GHz and boost clock at 5.00 GHz. Its thermal design power is 28 W, and it uses the AMD Socket FP8.
The Intel Core Ultra 9 285H is part of the Core Ultra Series 2, based on the Arrow Lake architecture with the Arrow Lake-H codename. Its generation string is "Ultra 9 (Arrow Lake-H)". It has 16 cores and 16 threads, with base clock at 2.90 GHz and boost clock at 5.40 GHz. Its TDP is 45 W, and it uses the Intel BGA 2049 socket.
Cache configurations differ markedly. AMD's chip has 80 KB of L1 per core, 1 MB of L2 per core, and 16 MB of L3 cache. Intel's chip has 192 KB of L1 per core, 3 MB of L2 per core, and 24 MB of shared L3 cache. The per-core L2 figures give the Intel part a substantial capacity advantage on a per-core basis, while the shared L3 allocation favors Intel by 8 MB overall.
The process nodes diverge by one generation step: AMD uses TSMC's 4 nm process, Intel uses TSMC's 3 nm process. The AMD die measures 233 mm², while the Intel die size is not recorded. Both processors are fabricated by TSMC.
Memory pathways are similar in bus width (dual-channel) and supported types (DDR5, LPDDR5X), but bandwidth differs: Intel records 102.4 GB/s versus AMD's 89.6 GB/s. ECC support exists only on the Intel part. PCIe capabilities also differ: AMD provides Gen 4 with 16 lanes (CPU only), while Intel provides Gen 5 with 8 lanes (CPU only).
Integrated graphics differ as well: AMD ships the Radeon 880M, Intel ships the Arc Graphics 140T. Both are mobile-segment parts, both are production-active, and both have locked multipliers. AMD released on 2024-06-30; Intel released on 2025-01-12. The Intel part has a recorded launch MSRP of $651; AMD's launch MSRP is not recorded.
The Verdict
Benchmark results indicate a clear overall winner in the Intel Core Ultra 9 285H, which takes 14 of 17 head-to-head comparisons. The AMD Ryzen AI 9 365 wins only 3. The Intel chip's victories span multi-core rendering, encryption, extended instructions, prime number finding, floating-point math, multithreaded workloads, physics, random string sorting, and single-thread PassMark. The AMD chip wins in Geekbench single-core, data compression, and integer math.
The data suggests the Intel part is the stronger choice for users whose workloads lean on floating-point math, physics simulations, encryption, and multithreaded rendering. Its 42.5% lead in floating-point math and 32.2% lead in physics are decisive margins. The AMD part is preferable when integer-heavy workloads dominate, given its 18.5% lead in integer math and 5.6% lead in data compression. For general single-core performance, the two chips trade blows: AMD leads Geekbench single-core by 3.4%, but Intel leads Cinebench R23 single-core by 6.5% and PassMark single-thread by 13%.
The percentile rankings are close: AMD sits at the 87th percentile of all CPUs, Intel at the 86th. Average benchmark scores favor AMD slightly (40048 vs 38312), though Intel's nearest rivals list shows comparable averages (38335 for the Intel Core 9 270H, 38261 for the Intel Core i5-13600HX). The AMD part's nearest rival, the AMD Ryzen 7 7700, posts an average score of 40081, nearly identical to AMD's own 40048.
For mobile users prioritizing sustained multi-threaded performance, the Intel Core Ultra 9 285H's 10% lead in Cinebench R23 multi-core and 13.8% lead in PassMark multithread are the strongest signals. For users who value compression throughput and integer workloads, the AMD Ryzen AI 9 365 delivers those wins but falls behind across most other measured categories.
Specification Differences
| Specification | AMD Ryzen AI 9 365 | Intel Core Ultra 9 285H |
| --- | --- | --- |
| Cores | 10 | 16 |
| Threads | 20 | 16 |
| Base Clock | 2.00 GHz | 2.90 GHz |
| Boost Clock | 5.00 GHz | 5.40 GHz |
| TDP | 28 W | 45 W |
| Socket | AMD Socket FP8 | Intel BGA 2049 |
| Architecture | Zen 5 | Arrow Lake |
| Codename | Strix Point | Arrow Lake-H |
| Generation | Ryzen AI 300 (Zen 5 / Zen 5c) | Ultra 9 (Arrow Lake-H) |
| Process Node | 4 nm | 3 nm |
| Die Size | 233 mm² | Not recorded |
| L1 Cache | 80 KB (per core) | 192 KB (per core) |
| L2 Cache | 1 MB (per core) | 3 MB (per core) |
| L3 Cache | 16 MB | 24 MB (shared) |
| Memory Bandwidth | 89.6 GB/s | 102.4 GB/s |
| ECC Memory | No | Yes |
| PCIe | Gen 4, 16 Lanes (CPU only) | Gen 5, 8 Lanes (CPU only) |
| Integrated Graphics | Radeon 880M | Arc Graphics 140T |
| Release Date | 2024-06-30 | 2025-01-12 |
| Part Number | 100-000001530 | SRQAL |
The two processors share DDR5 and LPDDR5X memory support, dual-channel memory bus, TSMC as foundry, mobile market segment, active production status, and locked multipliers. Their only common memory-related field is the supported types and bus width.
Head-to-Head Benchmarks
The Intel Core Ultra 9 285H dominates the Cinebench suite. In Cinebench R15 multi-core, it scores 3177.5 against AMD's 2842, a 10.6% lead. In Cinebench R23 multi-core, Intel scores 20781.5 versus 18698, a 10% advantage. Single-core Cinebench results also favor Intel: R15 single-core shows 313 versus 303 (3.2% lead), and R23 single-core shows 2129.5 versus 1992 (6.5% lead).
Geekbench splits the pair. Intel wins multi-core with 14743 against 13760, a 6.7% margin. AMD wins single-core with 2253 against 2178, a 3.4% margin. This is AMD's only single-core victory in the recorded head-to-head set.
PassMark results show the widest gaps. Intel takes floating-point math by a massive 42.5% (109190 vs 62802), physics by 32.2% (2513 vs 1704), data encryption by 30% (26140 vs 18297), and multithread by 13.8% (34171 vs 29467). Intel also leads in extended instructions by 6.3% (26794 vs 25113), random string sorting by 3.6% (40931 vs 39447), and single-thread by 13% (4415 vs 3841). The prime number finding test shows Intel at 330 versus AMD's 117, a 64.5% difference, the largest proportional gap in any recorded benchmark.
AMD's PassMark wins are concentrated in two areas. Data compression favors AMD at 354510 versus 335859, a 5.6% lead. Integer math favors AMD at 101831 versus 85922, an 18.5% lead. These two wins, combined with the Geekbench single-core result, give AMD 3 victories against Intel's 14.
The overall pattern shows Intel winning across rendering, encryption, physics, and floating-point workloads, while AMD holds narrower advantages in integer math, compression, and one single-core test. The average benchmark scores tell a slightly different story: AMD's 40048 average edges Intel's 38312, despite Intel winning the majority of individual tests. This discrepancy reflects the influence of AMD's 354510 data compression score, which is the single highest recorded value in either processor's benchmark list, and its 101831 integer math score, both of which lift the AMD average.