AMD EPYC 9275F vs Intel Xeon 6730P Comparison

The AMD EPYC 9275F and Intel Xeon 6730P are both high-end server processors, but the benchmark data reveals a clear performance hierarchy. The EPYC 9275F secures a decisive victory across the majority of workloads, winning 15 of the 17 head-to-head comparisons, while the Xeon 6730P manages only two wins. This dominance is reflected in the overall average benchmark scores, where the EPYC 9275F scores 133,174 compared to the Xeon 6730P's 124,756, a margin of 6.7%. Both processors sit in the 97th percentile of all CPUs, placing them firmly in the top tier of available hardware.

The Verdict

The data overwhelmingly favors the AMD EPYC 9275F for most server and workstation applications. Its lead is most pronounced in single-threaded and latency-sensitive tasks. In PassMark single-thread testing, the EPYC 9275F scores 3810 against the Xeon 6730P's 2995, a substantial 27.2% advantage. This trend continues across all Cinebench iterations, where the AMD part consistently leads by 14.2% in both single and multi-core tests. For workloads that depend on high-frequency, per-core performance—such as database transactions, real-time analytics, or legacy single-threaded applications—the EPYC 9275F is the clear choice based on the data.

The Xeon 6730P, however, is not without merit. Its two wins come in PassMark floating-point math (226,838 vs 201,888, an 11% advantage) and extended instructions (96,204 vs 94,889, a 1.4% edge). This suggests that for scientific computing, financial modeling, or other workloads that heavily utilize floating-point arithmetic and vectorized instruction sets, the Intel part may offer a tangible performance benefit. The Xeon also offers a higher core count (32 vs 24) and thread count (64 vs 48), which could be relevant for highly parallel throughput workloads that are not captured in the specific benchmarks provided.

However, the data suggests that the EPYC 9275F's architectural efficiency overcomes the Xeon's raw core count. The AMD processor leads in PassMark multi-thread (84,620 vs 74,113, +14.2%) and integer math (317,777 vs 290,740, +9.3%). The most striking difference is in prime number finding (991 vs 686, +44.5%) and physics calculations (12,089 vs 8,606, +40.5%), where the EPYC 9275F nearly doubles the performance of the Xeon. The verdict is clear: for general-purpose compute, the EPYC 9275F is the superior processor. The Xeon 6730P is only preferable in specific, float-heavy scientific workloads where its 11% advantage in floating-point math is the primary requirement.

FAQ

Q: Which processor is faster in single-threaded workloads?

A: The AMD EPYC 9275F is significantly faster. In PassMark single-thread testing, it scores 3810 versus the Intel Xeon 6730P's 2995, a 27.2% advantage. The same 14.2% lead is observed in all Cinebench single-core tests.

Q: Does the Intel Xeon 6730P win any benchmark tests?

A: Yes, it wins two tests. It leads in PassMark floating-point math (226,838 vs 201,888, +11%) and PassMark extended instructions (96,204 vs 94,889, +1.4%).

Q: How do the two processors compare on core and thread counts?

A: The Intel Xeon 6730P has more cores and threads, with 32 cores and 64 threads. The AMD EPYC 9275F has 24 cores and 48 threads. Despite this, the AMD part still leads in multi-threaded benchmark scores.

Q: What is the performance difference in Cinebench R23 multi-core?

A: The AMD EPYC 9275F scores 71,927, while the Intel Xeon 6730P scores 62,996. This gives the AMD processor a 14.2% lead in this multi-threaded rendering workload.

Q: Are there any significant differences in memory bandwidth?

A: Yes. The AMD EPYC 9275F has a twelve-channel memory bus with a bandwidth of 576.0 GB/s. The Intel Xeon 6730P has an eight-channel bus with 409.6 GB/s of bandwidth.

Q: Which processor has a higher average benchmark score?

A: The AMD EPYC 9275F has an average benchmark score of 133,174, which is 6.7% higher than the Intel Xeon 6730P's score of 124,756.

Architecture Differences

The AMD EPYC 9275F is built on the Zen 5 architecture, codenamed Turin, and is part of the EPYC 9005 series. It is manufactured on a 4 nm process at TSMC, with a die size of 8x 70.6 mm² and 66,520 million transistors. In contrast, the Intel Xeon 6730P uses the Granite Rapids architecture, part of the Xeon 6 (Granite Rapids-SP) generation. It is fabricated on Intel's 5 nm process with a significantly larger die size of 2x 598 mm²; the transistor count is not listed.

The cache hierarchies also differ considerably. The EPYC 9275F has an L1 cache of 80 KB per core and an L2 cache of 1 MB per core, with a large shared L3 cache of 256 MB. The Xeon 6730P has a larger per-core L1 (112 KB) and L2 (2 MB) cache, and a bigger shared L3 cache of 288 MB. This larger last-level cache on the Intel part could benefit workloads with large working sets, though the benchmark data does not show an overall advantage.

Memory architecture is another major differentiator. The AMD processor supports twelve-channel DDR5 memory with a peak bandwidth of 576.0 GB/s. The Intel processor supports eight-channel DDR5 with a peak bandwidth of 409.6 GB/s. This difference in memory bandwidth is substantial and likely contributes to the AMD part's lead in memory-intensive tasks like data compression and sorting.

PCIe connectivity also differs. The AMD EPYC 9275F provides 128 PCIe Gen 5 lanes (CPU only), while the Intel Xeon 6730P provides 88 lanes. This makes the AMD processor more suitable for high-density GPU or NVMe storage configurations that require extensive I/O. Both processors require external graphics, as they have no integrated graphics. The AMD processor uses AMD Socket SP5, while the Intel uses Intel Socket 4710.

Specification Differences

The most obvious specification difference is core count: the Intel Xeon 6730P has 32 cores and 64 threads, while the AMD EPYC 9275F has 24 cores and 48 threads. Clock speeds are also notably different. The AMD part has a base clock of 4.10 GHz and a boost clock of 4.80 GHz. The Intel part has a base clock of 2.50 GHz and a boost clock of 3.80 GHz, which explains the AMD's significant lead in single-threaded performance.

Thermal design power (TDP) differs, with the AMD EPYC 9275F rated at 320 W and the Intel Xeon 6730P at 250 W. This indicates the AMD part draws more power to achieve its higher clock speeds. Memory channels and bandwidth are also different, as detailed earlier, with the AMD part supporting twelve channels and the Intel part supporting eight. The PCIe lane count is another key difference: 128 lanes on the AMD part versus 88 on the Intel part.

The process node and foundry are different: the AMD uses a 4 nm process at TSMC, while the Intel uses a 5 nm process at Intel. The release dates differ, with the AMD EPYC 9275F released on 2024-10-09 and the Intel Xeon 6730P released on 2025-02-23. The launch MSRP for the AMD part is $3439, and for the Intel part it is $3726. Both processors have locked multipliers and are actively in production.

Head-to-Head Benchmarks

The AMD EPYC 9275F dominates the Cinebench suite, winning all six tests by a uniform margin of 14.2%. In Cinebench R23 multi-core, the AMD scores 71,927 versus the Intel's 62,996. The single-core R23 test shows a similar story: 10,154 for the AMD versus 8,893 for the Intel. This consistent lead across all Cinebench iterations, from R15 to R23, indicates a fundamental per-core performance advantage for the AMD architecture.

In PassMark tests, the AMD EPYC 9275F wins most categories. The largest win is in find prime numbers, where the AMD scores 991 compared to the Intel's 686, a massive 44.5% difference. Physics calculations also heavily favor the AMD part, with scores of 12,089 versus 8,606, a 40.5% lead. Random string sorting shows a 26.4% advantage for the AMD (144,037 vs 113,919). Data encryption is 12% faster on the AMD (62,664 vs 55,964), and data compression is 6.5% faster (1,212,560 vs 1,138,470). Integer math also favors the AMD part, with 317,777 versus 290,740, a 9.3% difference.

The Intel Xeon 6730P's two wins are notable. In floating-point math, the Intel scores 226,838 versus the AMD's 201,888, an 11% lead. This is a significant margin for a single workload and suggests the Intel architecture has a stronger floating-point unit. The other win is in extended instructions, where the Intel scores 96,204 versus 94,889, a marginal 1.4% lead. These two wins indicate that the Intel part is not without specialized strengths, but they are isolated to specific instruction-heavy tasks and do not offset the AMD's breadth of victories.