AMD EPYC 9335 vs Intel Xeon 6740E Comparison

FAQ

Q: Which processor has the higher average benchmark score?

A: The AMD EPYC 9335 posts an average benchmark score of 194,228, which is 3.5% higher than the Intel Xeon 6740E’s 187,718. Both chips sit in the 99th percentile of all CPUs.

Q: Which chip wins the most head-to-head benchmark comparisons?

A: The Intel Xeon 6740E wins 8 of the 11 head-to-head tests, while the AMD EPYC 9335 takes 3. Intel’s wins are concentrated in throughput-heavy workloads, while AMD’s wins are in single-thread and extended-instruction performance.

Q: How large is the single-thread performance gap?

A: The EPYC 9335 scores 2,732 in PassMark single-thread, which is 36.8% ahead of the Xeon 6740E’s 1,997. This is the largest relative advantage for AMD in the head-to-head set.

Q: What about encryption and compression workloads?

A: The Xeon 6740E leads decisively in both. It scores 1,786,845 in data compression versus 1,203,096 for the EPYC (a 32.7% delta), and 137,106 in data encryption versus 63,159 (a 53.9% delta).

Q: Do both processors support ECC memory?

A: Yes, both the AMD EPYC 9335 and the Intel Xeon 6740E include ECC memory support. Both also use DDR5, though they differ in memory channel count and bandwidth.

Q: Are these chips currently in production?

A: Yes, both are marked as Active in production status. The EPYC 9335 was released on October 9, 2024, while the Xeon 6740E launched earlier on June 2, 2024.

Architecture Differences

The AMD EPYC 9335 is built on the Zen 5 architecture, codenamed Turin, and is part of the EPYC 9005 series. It uses a 4 nm process from TSMC and packs 33,260 million transistors across a 4x 70.6 mm² die configuration. The Intel Xeon 6740E uses the Sierra Forest architecture, also its codename, from the Xeon 6 (Sierra Forest-SP) generation. Intel fabricates this chip on a 5 nm process at its own foundry, with a single 578 mm² die.

Core counts diverge sharply. The EPYC 9335 has 32 cores and 64 threads, meaning each physical core supports two threads. The Xeon 6740E has 96 cores but only 96 threads, indicating single-threaded cores. This is a fundamental design difference: AMD relies on simultaneous multithreading (SMT) to boost throughput, while Intel’s Sierra Forest uses a higher core count with no SMT.

Cache hierarchies also differ. The EPYC 9335 provides 80 KB of L1 per core, 1 MB of L2 per core, and 128 MB of shared L3. The Xeon 6740E offers 96 KB of L1 per core, 4 MB of L2 per module, and 96 MB of shared L3. AMD’s larger L3 cache is notable, but Intel’s L2 is organized at the module level, which can be advantageous for certain access patterns.

Memory subsystems are another key differentiator. The EPYC 9335 uses a twelve-channel memory bus with 576.0 GB/s of bandwidth, while the Xeon 6740E uses an eight-channel bus with 409.6 GB/s. Both support DDR5 and ECC. PCIe connectivity also favors AMD: the EPYC 9335 offers Gen 5 with 128 lanes (CPU only), versus Intel’s Gen 5 with 88 lanes.

Clock speeds reflect the design philosophies. The EPYC 9335 runs at 3.00 GHz base and 4.40 GHz boost, while the Xeon 6740E runs at 2.40 GHz base and 3.20 GHz boost. TDPs are 210 W for AMD and 250 W for Intel. Neither chip has an unlocked multiplier, and both lack integrated graphics.

Head-to-Head Benchmarks

The Intel Xeon 6740E dominates the throughput-oriented tests, often by wide margins. In data compression, Intel scores 1,786,845 against AMD’s 1,203,096, a 32.7% advantage. Data encryption shows an even larger gap: Intel’s 137,106 is 53.9% higher than AMD’s 63,159. These are the two biggest wins for Intel in percentage terms.

Intel also leads in several math and sorting workloads. Floating-point math goes to Intel at 309,333 versus 228,123, a 26.3% delta. Integer math is similarly one-sided: 457,614 for Intel against 346,291 for AMD, a 24.3% gap. Random string sorting favors Intel at 220,684 versus 116,608, a 47.2% difference. Prime number finding sees Intel at 526 versus AMD’s 340, a 35.4% advantage. The multithreaded PassMark score also goes to Intel, 76,167 versus 65,811, a 13.6% lead.

The physics test is the most lopsided result of all. Intel’s 7,608 score is 75% higher than AMD’s 1,905. This suggests that the Xeon 6740E’s 96 cores have a decisive advantage in simulation-style workloads that scale well with core count.

AMD’s wins are fewer but significant. The EPYC 9335 takes the extended instructions test with 105,706 versus Intel’s 78,968, a 33.9% advantage. This indicates that AMD’s cores handle advanced instruction sets more efficiently. Single-thread performance is also firmly in AMD’s corner: 2,732 versus 1,997, a 36.8% lead. This is the largest percentage win for either chip in the entire head-to-head set.

The overall benchmark count stands at 8 wins for Intel and 3 for AMD. However, the average score comparison tells a different story on magnitude. The EPYC 9335’s average of 194,228 is 3.5% above the Xeon 6740E’s 187,718. This discrepancy arises because the head-to-head tests are weighted differently than the average score metric, which includes additional benchmarks beyond the 11 compared here.

The Verdict

The data paints a clear picture of two processors optimized for different workloads. The Intel Xeon 6740E is the throughput champion, winning 8 of 11 head-to-head comparisons and posting large margins in compression, encryption, floating-point math, and physics. Its 96 cores and 96 threads deliver raw parallel performance, particularly in tasks that scale linearly with core count.

The AMD EPYC 9335 is the efficiency and single-thread leader. It wins the extended instructions and single-thread tests by 33.9% and 36.8% respectively, and it does so with a lower TDP of 210 W versus Intel’s 250 W. For workloads that rely on per-core performance, such as databases with high query rates or applications with serial bottlenecks, the EPYC 9335 is the better choice.

Memory and I/O also favor AMD. The twelve-channel memory bus with 576.0 GB/s of bandwidth is substantially wider than Intel’s eight-channel 409.6 GB/s configuration. The EPYC 9335 also offers 128 PCIe Gen 5 lanes versus Intel’s 88, making it more suitable for systems with many accelerators or high-speed storage devices.

The average benchmark scores are close, with AMD leading by 3.5%. The nearest rival data confirms this: the EPYC 9335’s closest competitor is the Intel Xeon 6741P at 194,901 (a 0.3% delta), while the Xeon 6740E’s nearest rival is the AMD EPYC 8534P at 185,092 (a 1.4% delta). This suggests that at the aggregate level, the two chips are in the same performance tier, but their strengths are in different domains.

For buyers prioritizing highest core count and raw throughput in multi-threaded server workloads, the Intel Xeon 6740E is the data-backed pick. For those needing strong single-thread performance, higher memory bandwidth, and more PCIe lanes, the AMD EPYC 9335 is the better option. The launch MSRP for the EPYC 9335 is $3178, while the Xeon 6740E is $5265.

Specification Differences

| Field | AMD EPYC 9335 | Intel Xeon 6740E |

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

| Cores | 32 | 96 |

| Threads | 64 | 96 |

| Base Clock | 3.00 GHz | 2.40 GHz |

| Boost Clock | 4.40 GHz | 3.20 GHz |

| TDP | 210 W | 250 W |

| Socket | AMD Socket SP5 | Intel Socket 4710 |

| Architecture | Zen 5 | Sierra Forest |

| Process Node | 4 nm (TSMC) | 5 nm (Intel) |

| Die Size | 4x 70.6 mm² | 578 mm² |

| L1 Cache | 80 KB (per core) | 96 KB (per core) |

| L2 Cache | 1 MB (per core) | 4 MB (per module) |

| L3 Cache | 128 MB (shared) | 96 MB (shared) |

| Memory Bus | Twelve-channel | Eight-channel |

| Memory Bandwidth | 576.0 GB/s | 409.6 GB/s |

| PCIe | Gen 5, 128 Lanes (CPU only) | Gen 5, 88 Lanes (CPU only) |

| Release Date | 2024-10-09 | 2024-06-02 |

| Launch MSRP | $3178 | $5265 |

| Part Number | 100-000001149 | SRPFV |