AMD EPYC 7351 vs Intel Core i9-9920X Comparison

The Intel Core i9-9920X and AMD EPYC 7351 are both 14 nm processors aimed at high-core-count workloads, but they target different markets: one is a desktop enthusiast part, the other a server/workstation chip. The benchmark data shows a consistent, if narrow, performance advantage for the Intel part across all tested Cinebench workloads, yet the EPYC counters with more physical cores, double the memory channels, and ECC support. This comparison hinges on whether raw benchmark scores or platform capabilities matter more for your specific use case.

Where Each One Wins

The Intel Core i9-9920X wins every single benchmark in the head-to-head comparison, taking all 6 recorded tests. Its margins are uniform at 6.6% across all multi-core and single-core Cinebench R15, R20, and R23 tests, with the single-core R15 test showing a slightly larger 6.8% lead. This makes the i9-9920X the clear choice for applications that rely on Cinebench-style rendering performance, where its higher clock speeds and per-core efficiency pay off.

The AMD EPYC 7351 does not win any benchmark in this dataset, but its strengths lie elsewhere. With 16 cores and 32 threads (versus 12 and 24 on the Intel), it offers more parallel processing headroom for heavily threaded server workloads that scale beyond what the benchmarks capture. Its eight-channel memory bus and 170.6 GB/s bandwidth (double the Intel's 85.3 GB/s) make it better suited for memory-bandwidth-intensive tasks like large database operations or virtualization, even if Cinebench scores do not reflect that advantage. The EPYC also supports ECC memory, a critical feature for data integrity in server environments that the Intel part lacks.

Architecture Differences

The architectural split is stark. The i9-9920X uses Intel's Skylake-X architecture on a 484 mm² die, while the EPYC 7351 is built on AMD's Zen (Naples) architecture with a much smaller 213 mm² die. Both are fabricated on a 14 nm process, but Intel uses its own foundry while AMD uses GlobalFoundries. The EPYC packs 4,800 million transistors into that smaller die, giving it a higher transistor density.

Cache layouts differ significantly. The Intel part has 64 KB of L1 and 1 MB of L2 per core, with 19.25 MB of shared L3 cache. The EPYC offers 96 KB of L1 per core (larger), but only 512 KB of L2 per core (half the Intel's), and compensates with a massive 64 MB of shared L3 cache — over three times the Intel's L3. This larger L3 pool helps the EPYC with data reuse across its 16 cores, while the Intel's larger per-core L2 benefits single-threaded responsiveness.

Memory architecture is another major divide. The i9-9920X runs quad-channel DDR4 with 85.3 GB/s bandwidth, while the EPYC 7351 runs eight-channel DDR4 with exactly double that at 170.6 GB/s. The EPYC also supports ECC memory; the Intel does not. For PCIe, the Intel provides Gen 3 with 44 lanes (CPU only), while the EPYC lists only "Gen 3" without a lane count in the data, suggesting the Intel offers more direct expansion lanes for GPU or NVMe setups.

Head-to-Head Benchmarks

The Cinebench results tell a consistent story. In Cinebench R15 multicore, the i9-9920X scores 2121 against the EPYC's 1989, a 6.6% lead. The single-core R15 test shows 299 vs 280, a 6.8% advantage for Intel. These margins hold steady through newer versions: R20 multicore shows 8839 vs 8291 (6.6% lead), and R20 single-core shows 1247 vs 1170 (6.6% lead). The pattern repeats in R23 with a multicore score of 21046 vs 19742 and a single-core score of 2971 vs 2787 — each a 6.6% gap.

The consistency of the deltaPct across every test (6.6%, except the R15 single-core at 6.8%) suggests the performance difference is architectural and clock-driven rather than workload-specific. The Intel's 3.50 GHz base and 4.50 GHz boost clocks vastly outpace the EPYC's 2.40 GHz base and 2.90 GHz boost, explaining why the Intel wins even though the EPYC has 33% more cores. In Geekbench, the i9-9920X scores 9422 multicore and 1393 single-core; the EPYC has no Geekbench scores in the data, so no comparison is possible there. The i9-9920X also holds a higher average benchmark score of 5917 versus the EPYC's 5710, though both sit at the 61st percentile of all CPUs.

FAQ

Q: Which processor has more cores and threads?

A: The AMD EPYC 7351 has 16 cores and 32 threads, while the Intel Core i9-9920X has 12 cores and 24 threads. The EPYC offers 4 more cores and 8 more threads.

Q: Does the Intel i9-9920X win all benchmark tests?

A: Yes. The head-to-head data shows the Intel part winning all 6 Cinebench tests (R15, R20, and R23, both multicore and single-core), with a consistent 6.6% margin in each, and the R15 single-core test at 6.8%.

Q: Which processor supports ECC memory?

A: The AMD EPYC 7351 supports ECC memory. The Intel Core i9-9920X does not support ECC memory.

Q: What is the memory bandwidth difference between them?

A: The EPYC 7351 has 170.6 GB/s of memory bandwidth across eight channels, exactly double the i9-9920X's 85.3 GB/s across four channels.

Q: Are both processors unlocked for overclocking?

A: Yes, both the Intel Core i9-9920X and AMD EPYC 7351 have the multiplier unlocked.

Q: What are the release dates for these chips?

A: The Intel Core i9-9920X was released on 2018-10-18, while the AMD EPYC 7351 was released earlier on 2017-06-28. The Intel part is end-of-life, while the EPYC is still active in production.

Specification Differences

| Specification | Intel Core i9-9920X | AMD EPYC 7351 |

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

| Cores | 12 | 16 |

| Threads | 24 | 32 |

| Base Clock | 3.50 GHz | 2.40 GHz |

| Boost Clock | 4.50 GHz | 2.90 GHz |

| TDP | 165 W | 170 W |

| Socket | Intel Socket 2066 | AMD Socket SP3 |

| Architecture | Skylake | Zen |

| Codename | Skylake-X | Naples |

| Process Node | 14 nm | 14 nm |

| Foundry | Intel | GlobalFoundries |

| Die Size | 484 mm² | 213 mm² |

| Transistors | Not specified | 4,800 million |

| L1 Cache | 64 KB (per core) | 96 KB (per core) |

| L2 Cache | 1 MB (per core) | 512 KB (per core) |

| L3 Cache | 19.25 MB (shared) | 64 MB (shared) |

| Memory Bus | Quad-channel | Eight-channel |

| Memory Bandwidth | 85.3 GB/s | 170.6 GB/s |

| ECC Memory | No | Yes |

| PCIe | Gen 3, 44 Lanes (CPU only) | Gen 3 |

| Market Segment | Desktop | Server/Workstation |

| Production Status | End-of-life | Active |

| Release Date | 2018-10-18 | 2017-06-28 |

| Launch MSRP | $1189 | Not specified |

| Part Number | SREZ6 | PS7351BEVGPAF |

The Verdict

The data points to a decisive but narrow victory for the Intel Core i9-9920X in raw rendering benchmarks, with a 6.6% edge across every Cinebench test. If your workload is Cinebench-style rendering and you need the highest scores, the i9-9920X is the better choice, despite having fewer cores. Its higher clock speeds (4.50 GHz boost vs 2.90 GHz) and larger per-core L2 cache clearly drive this win. It also offers more PCIe lanes (44 vs unspecified on the EPYC) and a lower TDP of 165 W versus 170 W.

However, the AMD EPYC 7351 is not without merit. It has 16 cores and 32 threads, which is 33% more parallel resources than the Intel. Its eight-channel memory and 170.6 GB/s bandwidth are double the Intel's, and it supports ECC memory — a non-negotiable feature for many server and workstation reliability requirements. The EPYC also remains in active production, while the i9-9920X is end-of-life. For a server platform where memory bandwidth, ECC, and core count matter more than single-thread clock speed, the EPYC 7351 is the rational pick. For a desktop workstation where Cinebench scores are the priority, the i9-9920X takes the win. Choose based on platform needs: rendering performance favors Intel, server capabilities favor AMD.