AMD EPYC 4545P vs Intel Core Ultra 9 275HX Comparison

The AMD EPYC 4545P and Intel Core Ultra 9 275HX both sit at the 97th percentile of all CPUs, yet they achieve that status through radically different designs and benchmark profiles. The AMD part, a 16-core/32-thread Zen 5 server/workstation chip on Socket AM5, and the Intel part, a 24-core/24-thread Arrow Lake-HX mobile processor on BGA 2114, are separated by an average benchmark score of just 0.5% in the AMD’s favor (76,433 vs. 76,024). That razor-thin aggregate margin, however, masks a benchmark suite where the two trade blows with deltas ranging from a 0.4% AMD win to a 39.9% AMD blowout and a 35.2% Intel blowout. The data shows a clear specialization: Intel dominates the Cinebench family and most PassMark compute tasks, while AMD secures decisive victories in specific integer and compression workloads.

Where Each One Wins

The Intel Core Ultra 9 275HX is the broader winner, taking 13 of the 17 head-to-head benchmarks. Its most significant advantages come in floating-point math (193,324 vs. 127,400, a 34.1% lead), prime number finding (457 vs. 296, a 35.2% lead), and data encryption (48,016 vs. 39,033, an 18.7% lead). The Intel chip also sweeps every Cinebench test, from R15 single-core (683 vs. 669) through R23 multi-core (48,008 vs. 47,079), with margins consistently around 1.9-2.0%. In PassMark’s multithreaded suite, Intel wins 56,415 to 55,388, a modest 1.8% edge, and it also leads in single-thread performance (4,727 vs. 4,568, a 3.4% advantage). For workloads dominated by encryption, extended instruction sets (47,687 vs. 45,111, a 5.4% lead), or heavy floating-point arithmetic, the Intel part is the clear pick.

The AMD EPYC 4545P wins only four benchmarks, but its victories are lopsided. The standout is PassMark integer math, where it scores 218,623 against Intel’s 156,314 — a massive 39.9% advantage. It also takes data compression (643,120 vs. 618,338, a 4% lead), physics (3,525 vs. 3,377, a 4.4% lead), and random string sorting (76,031 vs. 75,710, a narrow 0.4% margin). This pattern suggests AMD’s Zen 5 architecture with 32 threads excels at integer-heavy, sorting, and compression tasks, while Intel’s higher physical core count (24 vs. 16) and different cache hierarchy favor floating-point and encryption workloads. For database-style integer operations or compression pipelines, the AMD chip is disproportionately stronger than its average score suggests.

FAQ

Q: Which CPU has the higher average benchmark score?

A: The AMD EPYC 4545P has an average benchmark score of 76,433, while the Intel Core Ultra 9 275HX scores 76,024, giving AMD a 0.5% aggregate lead.

Q: How much faster is Intel in Cinebench R23 multi-core?

A: Intel wins Cinebench R23 multi-core with a score of 48,008 against AMD’s 47,079, a 1.9% advantage.

Q: Where does AMD have its largest win over Intel?

A: AMD’s biggest win is in PassMark integer math, where it scores 218,623 to Intel’s 156,314, a 39.9% lead.

Q: What is Intel’s largest margin of victory?

A: Intel’s largest win is in PassMark find prime numbers, scoring 457 against AMD’s 296, a 35.2% advantage.

Q: Do both CPUs have the same boost clock?

A: Yes, both the AMD EPYC 4545P and Intel Core Ultra 9 275HX have a boost clock of 5.40 GHz.

Q: Which processor supports ECC memory?

A: The AMD EPYC 4545P supports ECC memory; the Intel Core Ultra 9 275HX does not.

Head-to-Head Benchmarks

The Cinebench suite provides the most consistent picture: Intel wins all six tests, but by nearly identical margins. In Cinebench R15 multi-core, Intel scores 4,839 to AMD’s 4,745 (a 1.9% delta), and the single-core test shows 683 vs. 669 (a 2% delta). The pattern holds through R20 (20,163 vs. 19,773 multi-core; 2,846 vs. 2,791 single-core) and R23 (48,008 vs. 47,079 multi-core; 6,777 vs. 6,646 single-core), with every delta sitting at 1.9% or 2.0%. This uniformity indicates a consistent per-thread efficiency advantage for Intel’s Arrow Lake design in rendering workloads.

The PassMark suite tells a more polarized story. Intel’s floating-point math win is the second-largest margin on the board: 193,324 vs. 127,400, a 34.1% gap. Its prime number result is even more extreme at 457 vs. 296, a 35.2% lead. Encryption also favors Intel strongly (48,016 vs. 39,033, an 18.7% delta), as do extended instructions (47,687 vs. 45,111, a 5.4% delta) and single-thread tests (4,727 vs. 4,568, a 3.4% delta). Intel’s multithread score in PassMark is 56,415 vs. 55,388, a 1.8% win that is smaller than its Cinebench margins but still a victory.

AMD’s counterattacks are concentrated and severe. The integer math result of 218,623 vs. 156,314 is the single largest delta in the entire comparison at 39.9%. Data compression shows AMD ahead 643,120 to 618,338, a 4% delta, and physics puts AMD up 3,525 to 3,377, a 4.4% delta. Random string sorting is the closest race, with AMD winning 76,031 to 75,710, a mere 0.4% delta. Notably, AMD’s PassMark multithread score is lower than Intel’s (55,388 vs. 56,415), yet AMD still wins integer math by nearly 40%, showing that its 32 threads are far more efficient at integer operations than Intel’s 24 threads.

Specification Differences

The two processors diverge sharply on core and thread counts. AMD offers 16 cores and 32 threads, while Intel provides 24 cores and 24 threads — Intel has 50% more physical cores, but AMD has 33% more threads. Base clocks differ slightly: AMD runs at 3.00 GHz, Intel at 2.70 GHz, though both boost to 5.40 GHz. Thermal design power also differs, with AMD rated at 65W and Intel at 55W. The platforms are incompatible: AMD uses Socket AM5, while Intel uses BGA 2114. Memory bandwidth favors Intel at 102.4 GB/s vs. AMD’s 89.6 GB/s, though both support dual-channel DDR5. PCIe lane counts differ, with AMD providing 24 Gen 5 lanes (CPU only) versus Intel’s 20 Gen 5 lanes (CPU only). ECC memory is supported by AMD only. The AMD chip has integrated Radeon Graphics, while Intel integrates Arc Xe-LPG Graphics 64EU. AMD’s launch MSRP is $549; Intel has no listed launch MSRP. AMD’s multiplier is locked; Intel’s is unlocked.

Architecture Differences

The underlying architectures are from different generations and design philosophies. AMD’s EPYC 4545P uses Zen 5 architecture on a 4nm TSMC process, with a codename of Grado. Intel’s Core Ultra 9 275HX uses Arrow Lake architecture on a 3nm TSMC process, with a codename of Arrow Lake-HX. Transistor counts are close but not identical: AMD packs 16,630 million transistors across a die size of 2x 70.6 mm², while Intel uses 17,800 million transistors on a single 243 mm² die. Cache configurations differ substantially. AMD provides 80 KB of L1 per core, 1 MB of L2 per core, and 64 MB of shared L3. Intel offers 192 KB of L1 per core, 3 MB of L2 per core, but only 36 MB of shared L3. This means AMD’s total L3 cache is nearly double Intel’s (64 MB vs. 36 MB), while Intel’s per-core L1 and L2 are larger. The AMD chip is marked as a Server/Workstation part, while Intel’s is a Mobile part, and both are in active production, with AMD releasing on 2025-05-12 and Intel on 2025-01-12.