AMD Ryzen AI 9 HX 375 vs Intel Core Ultra 7 268V Comparison

Head-to-Head Benchmarks

The benchmark data delivers a decisive verdict: the AMD Ryzen AI 9 HX 375 wins 13 of 17 head-to-head comparisons, while the Intel Core Ultra 7 268V takes only 4. The scale of AMD's victories is substantial, not marginal. In Cinebench R23 multi-core, the Ryzen scores 21812 against Intel's 10653, a 104.7% advantage. Cinebench R15 multi-core shows a similar story: 3334 versus 1616, a 106.3% lead. Geekbench multi-core falls in line with a 40.1% gap, 13957 to 9963.

The largest single delta appears in PassMark integer math, where AMD delivers 121754 against Intel's 42669, a 185.3% difference. PassMark data compression also heavily favors AMD: 404918 versus 181443, a 123.2% margin. Random string sorting shows a 98.8% advantage, and extended instructions favor AMD by 91%. PassMark multithread scores 32916 for AMD versus 19297 for Intel, a 70.6% difference. Data encryption shows a 51% lead, floating point math a 30.4% lead, and physics a 12.5% lead.

Intel's four wins are concentrated in single-threaded and specific workloads. Geekbench single-core goes to Intel: 2270 versus 2084, a -8.2% delta for AMD. PassMark single-thread (and the duplicate singlethread entry) shows Intel at 4051 against AMD's 3867, a -4.5% delta. The most interesting Intel victory is PassMark find prime numbers: Intel scores 192, AMD scores 122, a -36.5% delta for AMD. This is Intel's largest margin of victory by far, indicating a specific algorithmic advantage rather than a general throughput edge.

The single-core Cinebench results are much closer. R23 single-core: AMD 1988, Intel 1921, a 3.5% AMD lead. R15 single-core: AMD 301, Intel 293, a 2.7% lead. These narrow margins contrast sharply with the multi-core results. The data confirms AMD's dominance scales with thread count, while Intel holds a narrower edge in lightly threaded workloads.

The aggregate statistics reinforce this picture. AMD's average benchmark score is 46030, placing it in the 89th percentile of all CPUs. Intel's average is 20897, at the 74th percentile. AMD's nearest rivals, by average score, are the Intel Core Ultra 5 235 (46062, -0.1%), AMD EPYC 4364P (45970, 0.1%), and Intel Core i9-13900HX (46098, -0.1%). Intel's nearest rivals are the AMD Ryzen 5 PRO 4655GE (20900, 0%), Intel Core i5-12600T (20917, -0.1%), and AMD EPYC 7J13 (20845, 0.2%). This places AMD's average score roughly 2.2 times higher than Intel's, a gap consistent with the multi-core benchmark deltas.

Architecture Differences

The two processors employ fundamentally different design philosophies. AMD uses a 12-core, 24-thread configuration based on the Zen 5 architecture, codenamed Strix Point. It belongs to the Ryzen AI 300 generation, which combines Zen 5 and Zen 5c cores. The process node is 4 nm, fabricated by TSMC, with a die size of 233 mm². Intel's Core Ultra 7 268V uses 8 cores and 8 threads, meaning no hyperthreading. It is based on Lunar Lake architecture, part of Core Ultra Series 2, on a 3 nm TSMC process. Intel's die size is not recorded in the database.

Cache layouts differ significantly. AMD provides 80 KB of L1 per core, 1 MB of L2 per core, and 16 MB of shared L3. Intel offers 192 KB of L1 per core, 2.5 MB of L2 per core, and 12 MB of shared L3. Intel's larger per-core L1 and L2 caches likely contribute to its single-thread wins in specific tests, while AMD's larger L3 and higher core count drive multi-threaded performance.

Clock speeds are close: AMD base clock is 2.00 GHz with a 5.10 GHz boost, while Intel starts at 2.20 GHz and boosts to 5.00 GHz. The thermal envelope differs substantially: AMD has a 28 W TDP, Intel a 17 W TDP. This 11 W difference explains part of AMD's performance advantage but also indicates Intel's efficiency focus. Memory support also differs: AMD lists DDR5 and LPDDR5X with dual-channel bus and 89.6 GB/s bandwidth. Intel's memory support is listed as unknown, depending on motherboard, with dual-channel bus and no recorded bandwidth figure.

PCIe connectivity varies: AMD provides Gen 4 with 16 CPU lanes, Intel provides Gen 5 with 4 CPU lanes. Integrated graphics differ: AMD uses Radeon 890M, Intel uses Arc 140V. The database does not include graphics benchmarks, so no direct comparison is possible. Socket types are incompatible: AMD uses Socket FP8, Intel uses BGA 2833. Both are mobile processors with locked multipliers, and neither supports ECC memory. Release dates show AMD launched on 2024-06-30, Intel on 2024-09-23, roughly three months apart.

Where Each One Wins

AMD's wins dominate every multi-threaded category. Cinebench R15 and R23 multi-core, Geekbench multi-core, PassMark multithread, integer math, floating point math, data compression, data encryption, extended instructions, random string sorting, and physics all go to AMD. The Ryzen AI 9 HX 375 is the clear choice for rendering, video encoding, compilation, database work, and any workload that scales across cores. Its 12 cores and 24 threads provide double the thread count of Intel's 8-core, 8-thread design.

Intel's wins are limited to three distinct areas. Geekbench single-core shows a 8.2% advantage, PassMark single-thread shows a 4.5% advantage. These indicate Intel's Lunar Lake cores have slightly higher peak single-thread throughput in certain integer-heavy workloads. The find prime numbers result is more dramatic: Intel beats AMD by 36.5%. This test appears to favor Intel's architecture disproportionately, possibly due to its larger per-core cache or specific instruction handling. It is the only test where Intel's margin exceeds single digits.

The Cinebench single-core results complicate the picture slightly. AMD actually wins both: R15 single-core by 2.7% and R23 single-core by 3.5%. This means Intel's single-thread wins are benchmark-specific rather than universal. In rendering workloads, AMD's single-core performance is slightly superior. In Geekbench and PassMark single-thread tests, Intel takes the lead. The data does not support a blanket statement that Intel has better single-core performance; rather, Intel wins in certain synthetic integer tests while AMD wins in rendering-oriented single-thread tests.

For workloads involving prime number calculations or similar algorithmic patterns, Intel shows a clear advantage. For everything else, AMD leads, often by wide margins. The 4 wins for Intel versus 13 for AMD, with AMD's average score more than double Intel's, indicates a lopsided overall performance relationship.

FAQ

Q: Which processor has higher multi-core performance?

A: The AMD Ryzen AI 9 HX 375 wins all multi-core benchmarks. Cinebench R23 multi-core shows 21812 versus 10653, a 104.7% lead. Geekbench multi-core shows 13957 versus 9963, a 40.1% lead. PassMark multithread shows 32916 versus 19297, a 70.6% lead.

Q: Does Intel win any benchmarks?

A: Yes, Intel wins 4 of 17 head-to-head tests. These are Geekbench single-core (2270 versus 2084, -8.2%), PassMark single-thread (4051 versus 3867, -4.5%), the duplicate PassMark singlethread entry, and PassMark find prime numbers (192 versus 122, -36.5%).

Q: How do the core counts compare?

A: AMD has 12 cores and 24 threads. Intel has 8 cores and 8 threads. AMD's thread count is triple Intel's, which explains its multi-core dominance.

Q: What is the TDP difference?

A: AMD has a 28 W TDP, Intel has a 17 W TDP. The 11 W difference indicates Intel targets lower power consumption, while AMD uses more power for higher performance.

Q: Which processor has higher single-core Cinebench performance?

A: AMD wins both Cinebench single-core tests. R15 single-core: AMD 301, Intel 293, a 2.7% lead. R23 single-core: AMD 1988, Intel 1921, a 3.5% lead.

Q: What are the average benchmark scores?

A: AMD's average benchmark score is 46030, placing it in the 89th percentile. Intel's average is 20897, placing it in the 74th percentile. AMD's nearest rival by average score is the Intel Core Ultra 5 235 at 46062, a -0.1% delta.

The Verdict

The data points decisively to the AMD Ryzen AI 9 HX 375 for any workload requiring multi-threaded throughput. Its 104.7% lead in Cinebench R23 multi-core, 123.2% lead in data compression, and 185.3% lead in integer math are not marginal differences; they represent a fundamentally different performance class. The 24-thread configuration versus Intel's 8-thread design ensures AMD wins every scalable workload. The 89th percentile ranking versus Intel's 74th confirms this at the aggregate level.

The Intel Core Ultra 7 268V wins only in narrow single-threaded scenarios and the specific prime number test. Its 8.2% Geekbench single-core advantage and 36.5% find prime numbers advantage are real but limited in scope. For users running Geekbench-style integer workloads or prime number calculations, Intel offers a measurable benefit. The 17 W TDP also suggests lower power draw, which the database records as a 11 W difference versus AMD.

For rendering, compilation, data processing, encryption, and general multi-core productivity, the AMD Ryzen AI 9 HX 375 is the superior choice. For single-threaded integer workloads and prime number finding, the Intel Core Ultra 7 268V holds a specific edge. The overall average benchmark score difference, 46030 versus 20897, leaves little ambiguity: AMD delivers roughly 2.2 times the average performance. The only reasons to select Intel would be the specific benchmark wins it holds, its lower TDP, or its smaller physical footprint implied by the BGA socket versus AMD's Socket FP8.

Specification Differences

| Specification | AMD Ryzen AI 9 HX 375 | Intel Core Ultra 7 268V |

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

| Cores | 12 | 8 |

| Threads | 24 | 8 |

| Base Clock | 2.00 GHz | 2.20 GHz |

| Boost Clock | 5.10 GHz | 5.00 GHz |

| TDP | 28 W | 17 W |

| Socket | AMD Socket FP8 | Intel BGA 2833 |

| Architecture | Zen 5 | Lunar Lake |

| Codename | Strix Point | Lunar Lake |

| Generation | Ryzen AI 300 (Zen 5 / Zen 5c) | Ultra 7 (Lunar Lake) |

| Process Node | 4 nm | 3 nm |

| Foundry | TSMC | TSMC |

| Die Size | 233 mm² | Not recorded |

| L1 Cache | 80 KB (per core) | 192 KB (per core) |

| L2 Cache | 1 MB (per core) | 2.5 MB (per core) |

| L3 Cache | 16 MB | 12 MB (shared) |

| Memory Support | DDR5, LPDDR5X | Unknown, depends on motherboard |

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

| Memory Bandwidth | 89.6 GB/s | Not recorded |

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

| Integrated Graphics | Radeon 890M | Arc 140V |

| Release Date | 2024-06-30 | 2024-09-23 |

| Part Number | 100-000001682 | SRPMLSRPMX |

| Multiplier Unlocked | No | No |