AMD EPYC 7663 vs Intel Xeon 6741P Comparison

# Head-to-Head Benchmarks

The Intel Xeon 6741P dominates this comparison in raw compute throughput. Across seventeen recorded benchmark comparisons, the Intel part wins fourteen, with the AMD EPYC 7663 taking three. The data shows a recurring pattern: the Xeon 6741P leads by roughly 22-23% in nearly every Cinebench test, while the EPYC 7663 manages narrow wins in specific workloads.

Starting with rendering and CPU stress tests, the Cinebench results are consistent and decisive. In Cinebench R15 multi-core, the Xeon 6741P scores 8,624 against 7,032 for the EPYC 7663, a 22.6% advantage. The single-core R15 result is similar: 1,217 versus 992, again 22.7% ahead. Moving to R20, multi-core shows 35,935 versus 29,304, and single-core 5,073 versus 4,137, both 22.6% gaps. R23 follows the same trajectory: multi-core 85,561 versus 69,773, single-core 12,079 versus 9,850, each with a 22.6% delta.

The PassMark suite tells a more varied story. The Xeon 6741P wins data compression decisively, scoring 1,816,408 versus 1,447,020, a 25.5% margin. Extended instructions show the largest gap of any test: the Intel part scores 142,682 versus 73,719, a 93.5% advantage. Prime number finding also heavily favors Intel, 1,242 versus 676, an 83.7% lead. Floating point math is another strong Intel showing at 358,423 versus 251,535, a 42.5% difference. Physics simulation favors Intel by 73.2%, with 13,890 versus 8,020. PassMark multi-threaded results give Intel a 22.6% edge (100,660 versus 82,087), and single-threaded scores mirror that at 3,195 versus 2,606.

The EPYC 7663 wins are narrower but real. Data encryption is its biggest victory: 111,725 versus 89,746, a 19.7% margin for AMD. Integer math is close, 468,096 versus 458,058, a 2.1% edge. Random string sorting also goes AMD's way, 184,367 versus 177,322, a 3.8% difference. These wins cluster around memory-access-sensitive or integer-heavy operations, but they do not offset the broad Intel lead in most other categories.

Architecture Differences

The two processors come from fundamentally different design generations. The Intel Xeon 6741P uses the Granite Rapids architecture on a 5 nm process from Intel Foundry, while the AMD EPYC 7663 uses the older Zen 3 (Milan) architecture on a 7 nm process from TSMC. The Intel part is built from two dies totaling 2x 598 mm², whereas AMD uses eight smaller dies of 81 mm² each, with 33,200 million transistors in the package.

Core counts favor AMD: the EPYC 7663 has 56 cores and 112 threads, versus 48 cores and 96 threads for the Xeon 6741P. Yet Intel compensates with much higher clock speeds. The Xeon 6741P runs at a 2.50 GHz base and 3.80 GHz boost, while the EPYC 7663 sits at 2.00 GHz base and 3.50 GHz boost. The database shows that per-core performance from the Intel part is strong enough to overcome the eight-core deficit.

Cache layouts differ sharply. Intel allocates 112 KB of L1 per core, 2 MB of L2 per core, and 288 MB of shared L3. AMD provides 64 KB L1 per core, 512 KB L2 per core, and 256 MB shared L3. Intel's larger per-core L2 and larger total L3 likely contribute to its wins in data compression and floating point tests.

Memory technology is a major differentiator. The Xeon 6741P supports DDR5 over an eight-channel bus with 409.6 GB/s of bandwidth. The EPYC 7663 uses DDR4, also eight-channel, but tops out at 204.8 GB/s. That is exactly double the bandwidth for Intel, which helps explain the 25.5% data compression win and the general throughput advantage. PCIe also favors Intel: Gen 5 with 136 lanes versus Gen 4 with 128 lanes.

Power and thermal envelopes differ as well. The Xeon 6741P is rated at 300 W TDP, the EPYC 7663 at 240 W. Both are server/workstation parts, both support ECC memory, and neither has integrated graphics. The Intel part was released in 2025, the AMD part in 2021.

FAQ

Q: Which processor has more cores?

A: The AMD EPYC 7663 has 56 cores and 112 threads, while the Intel Xeon 6741P has 48 cores and 96 threads.

Q: Why does the Xeon 6741P win most benchmarks despite fewer cores?

A: The Xeon 6741P has higher clocks (2.50 GHz base, 3.80 GHz boost versus 2.00 GHz base, 3.50 GHz boost) and double the memory bandwidth (409.6 GB/s versus 204.8 GB/s), which the recorded data indicates more than compensates for the core deficit.

Q: In which benchmark does the EPYC 7663 have its largest win?

A: Data encryption is the EPYC 7663's biggest victory, scoring 111,725 versus 89,746, a 19.7% margin.

Q: What is the largest single benchmark gap between the two?

A: PassMark extended instructions show the largest delta: the Xeon 6741P scores 142,682 versus 73,719, a 93.5% advantage.

Q: Do both processors support ECC memory?

A: Yes, both the Intel Xeon 6741P and the AMD EPYC 7663 support ECC memory.

Q: How do their release dates compare?

A: The AMD EPYC 7663 was released in March 2021, while the Intel Xeon 6741P came out in February 2025.

The Verdict

The data points to a clear performance hierarchy. The Intel Xeon 6741P is the faster processor in fourteen of seventeen benchmark comparisons, with an average benchmark score of 194,901 versus 161,973 for the EPYC 7663. That places Intel in the 99th percentile of all CPUs, while AMD sits in the 98th percentile. The Intel part also carries a higher launch MSRP of $4,421, compared to $6,366 for the AMD part at its launch, though pricing context is limited to those recorded figures.

Workloads that are rendering-heavy, floating-point-intensive, or memory-bandwidth-sensitive should favor the Xeon 6741P. The 22.6% lead across all Cinebench versions, the 42.5% floating-point win, and the 73.2% physics advantage make this the obvious choice for compute-dense data center tasks. The 93.5% extended instructions lead suggests the Intel part is substantially better for workloads using advanced instruction sets.

The EPYC 7663 remains viable for specific niches. Its 19.7% encryption win and small margins in integer math and string sorting indicate that some single-threaded or memory-latency-sensitive operations still benefit from the AMD design. However, these are narrow victories in a head-to-head that otherwise goes Intel's way.

Specification Differences

| Field | Intel Xeon 6741P | AMD EPYC 7663 |

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

| Cores | 48 | 56 |

| Threads | 96 | 112 |

| Base clock | 2.50 GHz | 2.00 GHz |

| Boost clock | 3.80 GHz | 3.50 GHz |

| TDP | 300 W | 240 W |

| Socket | Intel Socket 4710 | AMD Socket SP3 |

| Architecture | Granite Rapids | Zen 3 |

| Codename | Granite Rapids | Milan |

| Process node | 5 nm | 7 nm |

| Foundry | Intel | TSMC |

| Die size | 2x 598 mm² | 8x 81 mm² |

| Transistors | Not recorded | 33,200 million |

| L1 cache | 112 KB (per core) | 64 KB (per core) |

| L2 cache | 2 MB (per core) | 512 KB (per core) |

| L3 cache | 288 MB (shared) | 256 MB (shared) |

| Memory support | DDR5 | DDR4 |

| Memory bandwidth | 409.6 GB/s | 204.8 GB/s |

| PCIe | Gen 5, 136 lanes | Gen 4, 128 lanes |

| Release date | 2025-02-23 | 2021-03-14 |

| Launch MSRP | $4,421 | $6,366 |

| Part number | SRVEY | 100-000000318100-100000318WOF |

Where Each One Wins

The Intel Xeon 6741P is the clear choice for multi-threaded compute workloads. Every Cinebench test, both multi-core and single-core, shows a consistent 22.6-22.7% advantage. Data compression, extended instructions, prime number finding, floating point math, physics simulation, and PassMark multi-threaded all favor Intel by margins ranging from 22.6% to 93.5%. The doubled memory bandwidth and higher clocks make this the stronger processor for rendering, scientific simulation, and any workload that stresses the FPU or relies on large data movement.

The AMD EPYC 7663 wins in three specific areas. Data encryption is its standout performance, leading by 19.7%. Integer math shows a slim 2.1% edge, and random string sorting a 3.8% edge. These results suggest AMD retains an advantage in certain integer-heavy or latency-sensitive operations, possibly due to its larger core count and lower TDP. For workloads that are encryption-heavy or involve heavy integer arithmetic with moderate memory demands, the EPYC 7663 remains competitive.

For general server consolidation, virtualization, or mixed workloads, the benchmark data favors Intel overwhelmingly. The 48-core Xeon 6741P outperforms the 56-core EPYC 7663 in most aggregate tests, and its 99th percentile ranking versus 98th confirms the overall superiority. The AMD part's advantage is confined to a narrow set of operations, while Intel wins broadly and often by large margins.