AMD EPYC 9115 vs AMD Ryzen 9 9950X Comparison

The AMD Ryzen 9 9950X and AMD EPYC 9115 share the same Zen 5 architecture, 4 nm process, 16 cores, 32 threads, and identical cache layout, yet they are built for entirely different platforms. The data shows the Ryzen 9 9950X dominates nearly every benchmark in this comparison, winning 16 of 17 head-to-head tests, while the EPYC 9115 counters with a single physics win and massive platform-level advantages in memory bandwidth and PCIe connectivity. Both CPUs sit at the 96th percentile among all processors, with average benchmark scores of 70,420 and 69,887 respectively, making them close in aggregate performance but divergent in design intent.

FAQ

Q: Which CPU has the higher boost clock?

A: The AMD Ryzen 9 9950X boosts to 5.70 GHz, while the AMD EPYC 9115 boosts to 4.10 GHz.

Q: How do their memory bandwidth figures compare?

A: The EPYC 9115 offers 576.0 GB/s of memory bandwidth through its twelve-channel memory bus, versus 89.6 GB/s for the Ryzen 9 9950X’s dual-channel bus.

Q: Which processor is unlocked for overclocking?

A: The Ryzen 9 9950X has an unlocked multiplier; the EPYC 9115 does not.

Q: Do both CPUs support ECC memory?

A: Yes, both support ECC memory.

Q: What is the difference in their TDP ratings?

A: The Ryzen 9 9950X has a TDP of 170 W, while the EPYC 9115 has a TDP of 125 W.

Q: Which CPU has an integrated GPU?

A: The Ryzen 9 9950X includes Radeon Graphics, whereas the EPYC 9115 has no integrated graphics.

Architecture Differences

Both processors are built on the same TSMC 4 nm process and use the Zen 5 architecture, with identical transistor counts of 16,630 million and a die size of 2x 70.6 mm². The L1 cache is 80 KB per core, L2 is 1 MB per core, and L3 is 64 MB shared on both. The core and thread counts are also identical at 16 and 32. The fundamental differences lie in platform integration and clock behavior.

The Ryzen 9 9950X, codenamed Granite Ridge, is a desktop part on AMD Socket AM5. It runs at a base clock of 4.30 GHz and a boost clock of 5.70 GHz, with a TDP of 170 W. It uses a dual-channel DDR5 memory bus delivering 89.6 GB/s, and provides 24 PCIe Gen 5 lanes from the CPU. It includes integrated Radeon Graphics and an unlocked multiplier, making it a consumer-oriented part.

The EPYC 9115, codenamed Turin, is a server/workstation chip on AMD Socket SP5. Its base clock is 2.60 GHz and boost clock is 4.10 GHz, with a lower TDP of 125 W. The memory subsystem is vastly different: a twelve-channel DDR5 bus provides 576.0 GB/s of bandwidth, and the CPU exposes 128 PCIe Gen 5 lanes. It has no integrated graphics, no unlocked multiplier, and is designed for high-throughput, multi-socket server environments.

Despite the same core architecture, the clock speeds and platform capabilities create starkly different performance profiles. The Ryzen 9 9950X relies on high frequency for single-thread and lightly threaded workloads, while the EPYC 9115 prioritizes memory bandwidth and I/O scalability over raw clock speed.

The Verdict

The benchmark data is unambiguous: the Ryzen 9 9950X outperforms the EPYC 9115 in 16 of 17 head-to-head tests, including all Cinebench R15, R20, and R23 multi-core and single-core runs, and nearly all Passmark workloads. The average benchmark score of the Ryzen 9 9950X is 70,420, which is 0.8% higher than the EPYC 9115’s 69,887. In the nearest-rival context, the Ryzen 9 9950X sits 0.5% above the Intel Core i7-14700KF and 0.5% above the i7-14700K, while the EPYC 9115 sits 0.3% below both Intel parts.

For users building a desktop or workstation where raw CPU performance per socket is the priority, the Ryzen 9 9950X is the clear choice. It delivers a 33.6% advantage in Cinebench R23 multi-core (56,122 vs 42,003) and a 40.3% lead in Passmark single-thread (4,737 vs 3,377). The only test the EPYC 9115 wins is Passmark physics, where it scores 4,198 against 3,279, a 21.9% margin.

However, the EPYC 9115 is not meant to compete in the same arena. Its twelve-channel memory bus and 128 PCIe Gen 5 lanes are features the Ryzen 9 9950X cannot match. For server deployments that require massive memory bandwidth, high-density I/O, or ECC reliability across many sockets, the EPYC 9115 provides platform capabilities that the desktop part lacks. The choice hinges on whether the workload is bound by CPU compute speed (choose the Ryzen 9 9950X) or by memory and I/O scalability (choose the EPYC 9115).

Specification Differences

| Field | AMD Ryzen 9 9950X | AMD EPYC 9115 |

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

| Base clock | 4.30 GHz | 2.60 GHz |

| Boost clock | 5.70 GHz | 4.10 GHz |

| TDP | 170 W | 125 W |

| Socket | AMD Socket AM5 | AMD Socket SP5 |

| Codename | Granite Ridge | Turin |

| Memory bus | Dual-channel | Twelve-channel |

| Memory bandwidth | 89.6 GB/s | 576.0 GB/s |

| PCIe lanes (CPU only) | Gen 5, 24 Lanes | Gen 5, 128 Lanes |

| Integrated graphics | Radeon Graphics | None |

| Market segment | Desktop | Server/Workstation |

| Multiplier unlocked | Yes | No |

| Launch MSRP | $649 | $726 |

Both share the same core count, thread count, process node, transistor count, die size, cache hierarchy, and ECC support. The differences above are the only fields where the two diverge.

Head-to-Head Benchmarks

The Ryzen 9 9950X wins every Cinebench test by a consistent 33.6% margin. In Cinebench R23 multi-core, it scores 56,122 versus 42,003; in single-core, 7,923 versus 5,929. The same 33.6% delta appears in R15 and R20, indicating a uniform clock-driven advantage across all Cinebench versions.

Passmark results show a wider spread. The largest win for the Ryzen 9 9950X is in extended instructions, where it scores 71,564 against 45,686, a 56.6% lead. Data compression follows at 48.4% (896,544 vs 604,120), and single-thread performance is 40.3% higher (4,737 vs 3,377). Floating-point math shows a 37.2% gap (159,063 vs 115,955), while integer math and random string sorting each show a 32% advantage. Data encryption is 30.8% higher (44,366 vs 33,914), and multithread is 33.6% higher (66,030 vs 49,416). Even the find prime numbers test, which often favors server parts, shows a 21.1% lead for the Ryzen 9 9950X (345 vs 285).

The sole EPYC 9115 victory is Passmark physics, where it scores 4,198 against 3,279, a 21.9% margin. This is the only head-to-head test where the EPYC 9115 outperforms the Ryzen 9 9950X. Across the full set of 17 tests, the Ryzen 9 9950X holds an average lead of roughly 34% in the tests it wins, while the EPYC 9115’s single win is substantial but isolated.

Where Each One Wins

The Ryzen 9 9950X is the winner for any workload that depends on CPU compute throughput. Its higher base and boost clocks translate directly to superior performance in single-threaded applications, lightly threaded tasks, and even heavily threaded multi-core workloads like Cinebench. The data shows consistent 30–57% advantages across encryption, compression, math, and sorting tasks. For desktop users, content creators, and workstation professionals who need the fastest possible execution per core, the Ryzen 9 9950X is the definitive pick.

The EPYC 9115 wins in one specific benchmark—Passmark physics—but its real strengths are not captured by CPU-only tests. Its twelve-channel memory bus provides 576.0 GB/s of bandwidth, over six times that of the Ryzen 9 9950X, and its 128 PCIe Gen 5 lanes enable massive storage and networking configurations. For server environments where memory bandwidth and I/O throughput are the bottlenecks, the EPYC 9115 is the appropriate choice, even though its raw CPU scores are lower. The physics win may indicate better behavior in certain simulation or physics-based workloads, but the overall pattern is clear: the Ryzen 9 9950X dominates CPU-bound benchmarks, while the EPYC 9115 excels in platform-level scalability.