AMD EPYC 9965 vs Intel Xeon 696X Comparison

The benchmark data presents a clear hierarchy between the AMD EPYC 9965 and the Intel Xeon 696X, with the AMD processor dominating in almost every measurable category. The EPYC 9965 records 12 outright wins across the tested workloads, while the Intel Xeon 696X manages only 2, both in single-threaded tests. Despite the Intel part's higher peak clock speed, the sheer core and thread advantage of the AMD processor establishes an overwhelming lead in throughput-oriented tasks.

Head-to-Head Benchmarks

The most dramatic separation occurs in multi-core rendering workloads. In Cinebench R15, R20, and R23 multicore tests, the AMD EPYC 9965 consistently outperforms the Intel Xeon 696X by a delta of 53.2% in each instance. The scores are 13775 versus 8994 in R15, 57397 versus 37475 in R20, and 136661 versus 89227 in R23. This uniform margin across all three versions of the benchmark indicates a stable architectural advantage rather than a workload-specific anomaly.

The gap widens significantly in integer-heavy PassMark tests. The PassMark integer math score shows the EPYC 9965 at 1926069 against the Xeon's 572072, a delta of 236.7%. Similarly, the random string sorting test favors the AMD chip by 250.9%, with scores of 633030 and 180392 respectively. These results point to a massive advantage in general-purpose compute tasks that rely on parallel integer execution.

Encryption and data compression workloads further amplify the disparity. The EPYC 9965 achieves a PassMark data encryption score of 348449, which is 209.7% higher than the Xeon 696X's 112529. In data compression, the AMD part scores 5679990 versus 2264907, a delta of 150.8%. Floating point math also heavily favors AMD, with the EPYC 9965 reaching 1153453 compared to the Xeon's 450164, a 156.2% advantage.

The physics simulation test produces the single largest delta in the entire comparison. The EPYC 9965 scores 18707, while the Xeon 696X manages only 3382, resulting in a staggering 453.1% difference. Extended instruction set performance shows a 121.6% lead for AMD (383298 versus 172975), and even the prime number finding test, which often favors higher clock speeds, goes to the EPYC 9965 by 41.6% (1208 versus 853).

The only areas where the Intel Xeon 696X takes the lead are the PassMark single-threaded tests. The Xeon scores 3742 in both singlethread and single_thread tests, compared to the EPYC 9965's 3176, a 15.1% advantage for Intel. This confirms that the Xeon's higher boost clock of 4.80 GHz provides a tangible benefit in lightly threaded workloads, but it is a narrow victory in an otherwise one-sided contest.

Architecture Differences

The fundamental divergence lies in the core count and manufacturing process. The AMD EPYC 9965 packs 192 cores and 384 threads, built on a 3 nm process at TSMC. The Intel Xeon 696X offers 64 cores and 128 threads, fabricated on Intel's 5 nm node. This triple core advantage for AMD is the primary driver of its multicore dominance, as the benchmark deltas closely track the thread ratio.

Cache hierarchies also differ substantially. The EPYC 9965 provides 80 KB of L1 cache per core and 1 MB of L2 per core, with a massive 384 MB of shared L3 cache. The Xeon 696X offers larger per-core caches at 112 KB L1 and 2 MB L2, but its shared L3 cache tops out at 336 MB. The AMD processor's larger aggregate L3 capacity benefits workloads with large working sets that require frequent data sharing across cores.

Memory architecture follows a similar pattern. Both processors support DDR5 and ECC memory, but the EPYC 9965 utilizes a twelve-channel memory bus delivering 576.0 GB/s of bandwidth. The Xeon 696X operates with an eight-channel bus and achieves 409.6 GB/s. This 166.4 GB/s difference in memory bandwidth directly contributes to the AMD part's superiority in memory-intensive tasks like data compression and encryption.

The physical implementation differs as well. The Xeon 696X uses a dual-die design with a die size of 2x 598 mm², while the EPYC 9965's die dimensions are not recorded in the database. Both processors feature PCIe Gen 5 with 128 lanes, so expansion capabilities are equal. The Xeon 696X has an unlocked multiplier, whereas the EPYC 9965 does not, though this has no bearing on the recorded benchmark performance.

Clock speeds favor Intel on paper. The Xeon 696X runs at a 2.40 GHz base clock and boosts to 4.80 GHz, while the EPYC 9965 operates at 2.25 GHz base and 3.70 GHz boost. The 1.10 GHz boost advantage for Intel explains its single-threaded wins, but the EPYC 9965's 3.70 GHz boost is evidently sufficient to keep pace in mixed workloads. The TDP figures also differ, with the EPYC 9965 rated at 500 W and the Xeon 696X at 350 W.

Where Each One Wins

The AMD EPYC 9965 is the clear choice for any workload that scales with core count and memory bandwidth. The data shows decisive wins in rendering, as evidenced by the 53.2% lead across all Cinebench multicore tests. Server virtualization, database workloads, and scientific computing that can utilize 384 threads will see proportional gains. The 236.7% advantage in integer math and 156.2% lead in floating point math make it suitable for financial modeling, simulations, and general number crunching.

The encryption and compression results are particularly telling. With a 209.7% lead in data encryption and a 150.8% advantage in compression, the EPYC 9965 is better suited for security gateways, data storage servers, and any environment involving heavy cryptographic operations or data archival. The 453.1% physics score gap suggests substantial superiority in physics simulation and similar computational fluid dynamics workloads.

The Intel Xeon 696X wins only in single-threaded performance. Its 15.1% advantage in PassMark single-thread tests makes it preferable for legacy applications, database transactions that cannot parallelize, or lightly threaded server workloads where a single fast core matters more than aggregate throughput. The Xeon's higher boost clock of 4.80 GHz also means faster response times for interactive tasks or low-latency operations that depend on a single thread's speed.

For mixed environments with both heavily parallel and latency-sensitive workloads, the EPYC 9965's 192 cores still provide strong single-core performance at 3176, even if it trails the Xeon. The Xeon's 64 cores may struggle in parallel phases, as the 52.9% deficit in multithread performance demonstrates. The Intel part is best reserved for niche deployments where single-thread speed is the absolute priority and parallel workloads are minimal.

The Verdict

The data unequivocally favors the AMD EPYC 9965 for nearly all server and workstation use cases. Its 12 benchmark wins against 2 for Intel, combined with an average benchmark score of 620487 versus 286102, establishes it as the superior processor in this comparison. The EPYC 9965 sits at the 100th percentile of all CPUs in the database, while the Xeon 696X ranks at the 99th percentile, but the raw performance gap is substantial.

The EPYC 9965 outperforms its nearest rival, the AMD EPYC 9845, by 18.5% in average score, and leads the EPYC 9755 by 22.7%. This positions it as a top-tier part even within AMD's own lineup. The Xeon 696X, by contrast, trades nearly evenly with the AMD EPYC 9565 (0.2% ahead) and the EPYC 9555P (0.3% behind), indicating it competes with mid-range AMD parts rather than the flagship.

For buyers prioritizing raw compute density, parallel throughput, and memory bandwidth, the EPYC 9965 is the definitive choice. Its 384 threads, 384 MB of L3 cache, and 576.0 GB/s memory bandwidth deliver unmatched multicore performance. The Xeon 696X should only be considered for single-threaded applications where its 4.80 GHz boost clock provides a tangible benefit, and even then, the 15.1% single-thread advantage may not justify the massive multicore sacrifice.

FAQ

Q: Which processor has more cores and threads?

A: The AMD EPYC 9965 has 192 cores and 384 threads, while the Intel Xeon 696X has 64 cores and 128 threads.

Q: What is the difference in single-threaded performance?

A: The Intel Xeon 696X wins in PassMark single-thread tests with a score of 3742 versus 3176 for the AMD EPYC 9965, a 15.1% advantage for Intel.

Q: How do the processors compare in multi-core rendering?

A: The AMD EPYC 9965 leads by 53.2% across Cinebench R15, R20, and R23 multicore tests, with scores of 13775, 57397, and 136661 respectively.

Q: Which processor has higher memory bandwidth?

A: The AMD EPYC 9965 provides 576.0 GB/s via a twelve-channel memory bus, while the Intel Xeon 696X offers 409.6 GB/s through an eight-channel bus.

Q: Are both processors on the same manufacturing process?

A: No, the AMD EPYC 9965 uses a 3 nm process at TSMC, while the Intel Xeon 696X uses a 5 nm process at Intel.

Q: What is the largest performance gap between the two?

A: The PassMark physics test shows the largest delta, with the AMD EPYC 9965 scoring 18707 versus 3382 for the Intel Xeon 696X, a 453.1% difference.

Specification Differences

| Specification | AMD EPYC 9965 | Intel Xeon 696X |

| Core Count | 192 | 64 |

| Thread Count | 384 | 128 |

| Base Clock | 2.25 GHz | 2.40 GHz |

| Boost Clock | 3.70 GHz | 4.80 GHz |

| TDP | 500 W | 350 W |

| Process Node | 3 nm | 5 nm |

| L1 Cache | 80 KB (per core) | 112 KB (per core) |

| L2 Cache | 1 MB (per core) | 2 MB (per core) |

| L3 Cache | 384 MB (shared) | 336 MB (shared) |

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

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

| Socket | AMD Socket SP5 | Intel Socket 4710 |

| Die Size | Not recorded | 2x 598 mm² |

| Multiplier Unlocked | No | Yes |

| Launch MSRP | $14813 | $5599 |