AMD EPYC 9455P vs AMD Ryzen Threadripper PRO 9975WX Comparison

The AMD EPYC 9455P and AMD Ryzen Threadripper PRO 9975WX both leverage the Zen 5 architecture, yet benchmark results show they are tuned for very different workloads. The data indicates a clear split between raw multi-threaded throughput and single-thread responsiveness, with the EPYC 9455P dominating the former and the Threadripper taking the latter. This analysis breaks down the numbers to see what each processor does best.

Head-to-Head Benchmarks

The EPYC 9455P establishes a decisive lead in nearly every multi-threaded test. In Cinebench R23 multi-core, the EPYC scores 99,206 against the Threadripper’s 93,553, a 6% advantage. That pattern holds across the entire Cinebench suite, with the EPYC winning R15 multi-core (9,999 vs 9,430), R20 multi-core (41,666 vs 39,292), and R15 single-core (1,411 vs 1,331) by the same 6% margin. The single-core Cinebench results are particularly interesting: the EPYC wins R20 single-core by 6.1% (5,882 vs 5,546) despite the Threadripper having a higher boost clock. This suggests the EPYC’s architecture extracts more performance per clock in these specific workloads.

The gap widens dramatically in PassMark’s specialized tests. The EPYC 9455P crushes the Threadripper in data encryption (115,403 vs 85,035, a 35.7% lead) and integer math (606,239 vs 461,724, a 31.3% lead). The most extreme divergence appears in the physics test, where the EPYC scores 17,315 versus the Threadripper’s 7,288 — an enormous 137.6% advantage. Random string sorting also favors the EPYC by 29.1% (242,701 vs 188,014). These results point to the EPYC’s 48 cores providing massive parallel throughput that a 32-core processor cannot match, even when the 32-core chip runs at higher clocks.

The Threadripper PRO 9975WX fights back in two key areas. PassMark single-thread performance shows the Threadripper at 4,408 versus the EPYC’s 3,745, a 15% victory. This aligns with its higher base clock (4.00 GHz vs 3.15 GHz) and boost clock (5.40 GHz vs 4.40 GHz). The Threadripper also wins extended instructions by a hair: 137,733 vs 137,485, a 0.2% margin that is essentially negligible. However, these two wins do little to offset the EPYC’s dominance elsewhere. The final tally shows the EPYC winning 14 of 17 head-to-head benchmarks, with the Threadripper taking only 3.

Architecture Differences

Both processors are built on TSMC’s 4 nm process and share the Zen 5 architecture, but their physical layouts differ substantially. The EPYC 9455P, codenamed Turin, uses an 8-chiplet design with each die at 70.6 mm², totaling 66,520 million transistors. The Threadripper PRO 9975WX, codenamed Shimada Peak, uses a 4-chiplet design with the same 70.6 mm² per die, but only 33,260 million transistors — roughly half the transistor count. This directly explains the core count disparity: the EPYC packs 48 cores and 96 threads, while the Threadripper offers 32 cores and 64 threads.

Cache allocation follows the core count. The EPYC provides 80 KB of L1 cache per core, 1 MB of L2 per core, and a massive 256 MB of shared L3 cache. The Threadripper offers less L1 (64 KB per core), the same 1 MB L2 per core, but only 128 MB of L3. The EPYC’s larger L3 cache likely contributes to its strong performance in data compression (1,928,897 vs 1,644,573, a 17.3% lead) and encryption tasks, where larger working sets can stay resident on-chip.

Memory subsystems also diverge. The EPYC 9455P supports twelve-channel DDR5 memory with 576.0 GB/s bandwidth, while the Threadripper uses eight-channel DDR5 with 409.6 GB/s. Both support ECC memory and offer 128 PCIe Gen 5 lanes. The EPYC’s higher memory bandwidth aligns with its server positioning, where memory-hungry workloads benefit from the extra channels. The Threadripper compensates with an unlocked multiplier, allowing overclocking, whereas the EPYC is locked. The EPYC uses AMD Socket SP5, while the Threadripper uses Socket sTR5.

FAQ

Q: Which processor wins more head-to-head benchmarks?

A: The AMD EPYC 9455P wins 14 out of 17 head-to-head benchmarks. The AMD Ryzen Threadripper PRO 9975WX wins only 3, which are PassMark single-thread, PassMark singlethread, and PassMark extended instructions.

Q: How large is the performance gap in the physics test?

A: The EPYC 9455P scores 17,315 in PassMark physics, while the Threadripper PRO 9975WX scores 7,288. This represents a 137.6% advantage for the EPYC, the largest delta in any benchmark between the two.

Q: Does the Threadripper have any meaningful single-thread advantage?

A: Yes. In PassMark single-thread, the Threadripper scores 4,408 versus the EPYC’s 3,745, a 15% lead. However, in Cinebench R15 single-core, the EPYC actually wins (1,411 vs 1,331), showing the Threadripper’s advantage is workload-specific.

Q: What causes the EPYC’s large lead in encryption and integer math?

A: The EPYC has 48 cores versus the Threadripper’s 32. In data encryption, the EPYC scores 115,403 versus 85,035 (35.7% higher), and in integer math it scores 606,239 versus 461,724 (31.3% higher). The extra cores and 256 MB L3 cache allow more parallel execution and better data residency.

Q: How do the memory bandwidth specifications compare?

A: The EPYC 9455P supports twelve-channel DDR5 with 576.0 GB/s bandwidth. The Threadripper PRO 9975WX supports eight-channel DDR5 with 409.6 GB/s. The EPYC’s bandwidth is roughly 40% higher, which helps feed its additional cores.

Q: Are both processors on the same manufacturing process?

A: Yes, both are built on TSMC’s 4 nm process and use the Zen 5 architecture. The key difference is the number of chiplets: the EPYC uses 8 dies, while the Threadripper uses 4.

Specification Differences

The table below highlights only the fields where the two processors differ.

| Specification | AMD EPYC 9455P | AMD Ryzen Threadripper PRO 9975WX |

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

| Cores | 48 | 32 |

| Threads | 96 | 64 |

| Base Clock | 3.15 GHz | 4.00 GHz |

| Boost Clock | 4.40 GHz | 5.40 GHz |

| TDP | 300 W | 350 W |

| Socket | AMD Socket SP5 | AMD Socket sTR5 |

| Codename | Turin | Shimada Peak |

| Transistors | 66,520 million | 33,260 million |

| Die Size | 8x 70.6 mm² | 4x 70.6 mm² |

| L1 Cache | 80 KB (per core) | 64 KB (per core) |

| L3 Cache | 256 MB (shared) | 128 MB |

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

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

| Multiplier Unlocked | No | Yes |

| Release Date | 2024-10-09 | 2025-07-22 |

| Launch MSRP | $4819 | $4099 |

| Part Number | 100-000001563 | 100-000000723 |

The Verdict

The data paints a clear picture: the EPYC 9455P is the superior processor for multi-threaded throughput, winning 14 of 17 benchmarks and delivering massive leads in physics (137.6%), encryption (35.7%), and integer math (31.3%). Its 48 cores, 256 MB of L3 cache, and twelve-channel memory give it an insurmountable advantage in parallel workloads. The Threadripper PRO 9975WX, despite having a 5.40 GHz boost clock and an unlocked multiplier, cannot overcome the core deficit. The single-thread PassMark win (15%) is real but narrow, and it does not translate to Cinebench single-core victories.

Looking at the nearest rivals, the EPYC 9455P sits 3.8% above the Intel Xeon w9-3595X and 11.8% above the Intel Xeon 6741P, while trailing the Intel Xeon 6747P by 8.6%. The Threadripper PRO 9975WX, by contrast, is 1.5% above the AMD EPYC 7763 but 2.7% below the Intel Xeon 6740E. This contextual data shows the EPYC competing at the top of the server heap, while the Threadripper sits in a more crowded mid-field.

Where Each One Wins

The EPYC 9455P is the clear choice for server and workstation deployments where multi-threaded performance is paramount. Its wins in data compression, encryption, floating-point math (357,783 vs 304,833), and random string sorting make it ideal for database workloads, scientific computing, and any application that can utilize 96 threads. The physics test result — a 137.6% lead — suggests exceptional performance in simulation and rendering tasks that scale with core count. The 576.0 GB/s memory bandwidth further cements its position for memory-intensive server applications.

The Threadripper PRO 9975WX wins specifically in single-thread PassMark tests and extended instructions (by a negligible 0.2%). Its 4.00 GHz base clock and 5.40 GHz boost clock, combined with an unlocked multiplier, make it more suitable for lightly-threaded applications that depend on raw clock speed. The 128 MB L3 cache and 409.6 GB/s bandwidth are still substantial, but the 32-core count limits its ceiling in heavily parallel tasks. For users who need moderate multi-threading but prioritize single-thread responsiveness and overclocking headroom, the Threadripper offers a distinct — albeit narrower — set of advantages. The benchmark data ultimately favors the EPYC 9455P for any workload that scales across cores, while the Threadripper’s wins are confined to specific, single-threaded scenarios.