AMD Ryzen 9 PRO 8945HS vs Intel Core Ultra 7 265H Comparison

The AMD Ryzen 9 PRO 8945HS and Intel Core Ultra 7 265H are two high-end mobile processors that take fundamentally different approaches to achieving performance. The AMD part is an 8-core, 16-thread Zen 4 design from the 8000 series, while Intel’s Core Ultra 7 265H is a 16-core, 16-thread Arrow Lake-H part from the Core Ultra Series 2. Benchmark data shows a split decision: the Intel chip wins 10 of the 17 head-to-head tests, while the AMD chip wins 7. However, the margin of victory in each direction tells a more nuanced story, with AMD posting dramatic wins in some Cinebench tests and Intel dominating in math and physics workloads.

Where Each One Wins

The Intel Core Ultra 7 265H asserts its dominance in heavily parallel, throughput-oriented workloads that reward raw core count. It wins the Cinebench R15 multicore test with a score of 2989 against AMD’s 2536, a 15.2% advantage. The trend continues in Cinebench R20 multicore, where Intel scores 12131 versus 10569, a 12.9% lead. The PassMark suite reinforces this pattern: Intel wins multithread (34027 vs 30378, +10.7%), floating point math (109123 vs 63490, +41.8%), physics (2497 vs 1430, +42.7%), and find prime numbers (335 vs 91, +72.8%). Intel also wins data encryption (26005 vs 21820, +16.1%) and single-thread tests (4334 vs 3920, +9.6%).

The AMD Ryzen 9 PRO 8945HS, despite having half the core count, secures its wins in specific Cinebench scenarios and integer-heavy tasks. Its most striking victory is in Cinebench R23 single-core, scoring 3552 against Intel’s 2080 — a massive 70.8% lead. It also wins Cinebench R23 multi-core with 25165 versus 19940, a 26.2% advantage. In PassMark, AMD takes data compression (368884 vs 334711, +10.2%), integer math (103390 vs 85479, +21%), extended instructions (27714 vs 26805, +3.4%), and random string sorting (44668 vs 40742, +9.6%). The AMD chip also wins Cinebench R15 single-core (357 vs 307, +16.3%). Notably, AMD’s wins in R23 are so large that they offset Intel’s broader win count in the average benchmark score comparison.

Architecture Differences

The two processors are built on different process nodes and microarchitectures. AMD’s Ryzen 9 PRO 8945HS uses the Zen 4 architecture, codenamed Hawk Point, fabricated on a 4 nm process at TSMC. Intel’s Core Ultra 7 265H uses the Arrow Lake architecture, codenamed Arrow Lake-H, on a 3 nm process, also from TSMC. The transistor counts and die sizes are only listed for AMD: 25,000 million transistors on a 178 mm² die; Intel’s figures are not provided in the data.

The core configurations differ sharply. AMD has 8 cores and 16 threads, while Intel has 16 cores and 16 threads. This means Intel’s cores are all single-threaded, whereas AMD relies on simultaneous multithreading to reach 16 threads. Base clocks reflect this: AMD runs at 4.00 GHz base and 5.20 GHz boost, while Intel runs at 2.20 GHz base and 5.30 GHz boost. Cache hierarchies are also distinct. AMD provides 64 KB L1 per core, 1 MB L2 per core, and 16 MB shared L3. Intel offers 192 KB L1 per core, 3 MB L2 per core, and 24 MB shared L3. Intel’s larger L3 cache and per-core L2 may explain its advantage in certain memory-sensitive workloads.

Memory and I/O further differentiate the two. Both support DDR5, but Intel also supports LPDDR5X. Memory bandwidth is listed at 89.6 GB/s for AMD and 102.4 GB/s for Intel, both over a dual-channel bus. PCIe support differs significantly: AMD provides Gen 4 with 20 lanes (CPU only), while Intel provides Gen 5 with 8 lanes (CPU only). Both support ECC memory. The integrated graphics also differ: AMD uses Radeon 780M, Intel uses Arc Graphics 140T. The TDP ratings are 45 W for AMD and 28 W for Intel, though the data does not specify whether these are configurable or fixed. AMD uses the AMD Socket FP7, while Intel uses Intel BGA 2049. Both processors are currently active in production, with AMD released on 2024-04-15 and Intel on 2025-01-12.

Head-to-Head Benchmarks

The Cinebench results are the most contradictory. In Cinebench R15 multicore, Intel wins with 2989 against AMD’s 2536, a 15.2% margin. But in Cinebench R23 multicore, AMD flips the result, scoring 25165 versus Intel’s 19940, a 26.2% lead. This reversal suggests that the R23 workload scales differently with AMD’s 8-core/16-thread configuration versus Intel’s 16-core/16-thread design. In single-core tests, the gap is even more extreme. AMD wins R15 single-core by 16.3% (357 vs 307), but then Intel wins R20 single-core by 12.9% (1712 vs 1491). Then in R23 single-core, AMD obliterates Intel with a 70.8% margin (3552 vs 2080). These swings across Cinebench versions indicate that the two architectures respond very differently to the same vendor’s successive benchmark revisions.

The PassMark suite provides a clearer picture of workload-specific strengths. Intel’s largest win is in find prime numbers, where it scores 335 against AMD’s 91, a 72.8% advantage. This is an integer-heavy, branch-intensive workload. Intel also wins floating point math by 41.8% (109123 vs 63490) and physics by 42.7% (2497 vs 1430). AMD’s biggest PassMark win is in integer math, scoring 103390 versus 85479, a 21% lead. AMD also takes data compression by 10.2% (368884 vs 334711) and random string sorting by 9.6% (44668 vs 40742). Intel wins data encryption by 16.1% (26005 vs 21820) and single-thread by 9.6% (4334 vs 3920). The extended instructions test is nearly a tie, with AMD ahead by only 3.4% (27714 vs 26805).

The overall average benchmark scores are very close: AMD’s average is 41963, while Intel’s is 41621. Both sit at the 88th percentile of all CPUs. AMD’s nearest rivals include the Intel Core i7-14700T (delta 0.1%), AMD Ryzen 5 7400 (delta -0.2%), and Intel Core i7-12850HX (delta 0.4%). Intel’s nearest rivals include the Intel Core 7 251TE (delta -0.1%), Intel Core i7-14650HX (delta 0.1%), and AMD Ryzen 9 5900X (delta 0.6%). This near-parity in average scores, despite divergent individual test results, underscores how benchmark methodology can obscure real architectural differences.

FAQ

Q: Which processor has a higher core count?

A: The Intel Core Ultra 7 265H has 16 cores, while the AMD Ryzen 9 PRO 8945HS has 8 cores. Both have 16 threads, but Intel achieves this with all single-threaded cores, whereas AMD uses simultaneous multithreading.

Q: How do the processors compare in Cinebench R23 single-core?

A: AMD wins decisively. The Ryzen 9 PRO 8945HS scores 3552, which is 70.8% higher than Intel’s 2080. This is the largest single-core margin in the head-to-head data.

Q: What is the memory bandwidth difference?

A: Intel’s Core Ultra 7 265H has a higher memory bandwidth of 102.4 GB/s, compared to AMD’s 89.6 GB/s. Both use a dual-channel memory bus and support DDR5, but Intel also supports LPDDR5X.

Q: Which chip wins in floating point math?

A: Intel wins by a wide margin. The Core Ultra 7 265H scores 109123 in PassMark floating point math, versus AMD’s 63490, a 41.8% advantage.

Q: Are both processors currently in production?

A: Yes, both the AMD Ryzen 9 PRO 8945HS and the Intel Core Ultra 7 265H have an active production status. AMD was released on 2024-04-15, and Intel was released on 2025-01-12.

Q: What is the TDP difference?

A: The AMD processor has a TDP of 45 W, while the Intel processor has a TDP of 28 W. This does not necessarily indicate a direct relationship to performance, as benchmark results vary by workload.

Specification Differences

| Specification | AMD Ryzen 9 PRO 8945HS | Intel Core Ultra 7 265H |

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

| Series | 8000 series | Core Ultra Series 2 |

| Architecture | Zen 4 | Arrow Lake |

| Codename | Hawk Point | Arrow Lake-H |

| Generation | Ryzen 9 (Zen 4 (Hawk Point)) | Ultra 7 (Arrow Lake-H) |

| Process Node | 4 nm | 3 nm |

| Transistors | 25,000 million | Not listed |

| Die Size | 178 mm² | Not listed |

| Cores | 8 | 16 |

| Threads | 16 | 16 |

| Base Clock | 4.00 GHz | 2.20 GHz |

| Boost Clock | 5.20 GHz | 5.30 GHz |

| TDP | 45 W | 28 W |

| Socket | AMD Socket FP7 | Intel BGA 2049 |

| L1 Cache | 64 KB (per core) | 192 KB (per core) |

| L2 Cache | 1 MB (per core) | 3 MB (per core) |

| L3 Cache | 16 MB (shared) | 24 MB (shared) |

| Memory Support | DDR5 | DDR5, LPDDR5X |

| Memory Bandwidth | 89.6 GB/s | 102.4 GB/s |

| PCIe | Gen 4, 20 Lanes (CPU only) | Gen 5, 8 Lanes (CPU only) |

| Integrated Graphics | Radeon 780M | Arc Graphics 140T |

| Release Date | 2024-04-15 | 2025-01-12 |

| Part Number | 100-000001314(FP7r2),100-000001386(FP7) | SRQAQ |