AMD EPYC 9015 vs AMD Ryzen AI Max 390 Comparison

The AMD EPYC 9015 and AMD Ryzen AI Max 390 are both Zen 5 parts from AMD, but they target opposite ends of the computing spectrum. The EPYC 9015 is a server processor built for the SP5 platform with a 125 W TDP, while the Ryzen AI Max 390 is a mobile chip with a 55 W TDP and integrated graphics. The benchmark data shows a clear overall winner in raw compute, but the EPYC 9015 holds specific advantages that matter in its intended server role. This analysis walks through the head-to-head results, the verdict, use-case splits, frequency questions, and the architectural and specification differences between the two.

Head-to-Head Benchmarks

The PassMark suite reveals a decisive performance gap, with the Ryzen AI Max 390 winning 10 of the 11 compared tests. The largest margin comes in integer math, where the Ryzen AI Max 390 scores 146,519 against the EPYC 9015's 92,524, a 36.9% advantage. This is the single biggest delta in the entire comparison and indicates a substantial throughput difference for general-purpose arithmetic workloads. Floating-point math also shows a massive gap: the Ryzen AI Max 390 posts 90,594 versus 61,615 for the EPYC 9015, a 32% lead. These two results alone establish the mobile chip as the far more capable calculation engine in this matchup.

The data encryption test follows a similar pattern, with the Ryzen AI Max 390 scoring 25,097 compared to the EPYC 9015's 17,790, a 29.1% difference. Extended instructions show a 27.8% gap (38,716 vs. 27,970), and data compression is 27.5% higher for the Ryzen AI Max 390 (487,145 vs. 353,014). The multithread score, which is a broad measure of parallel performance, comes in at 41,737 for the Ryzen AI Max 390 versus 30,689 for the EPYC 9015, a 26.5% difference. Random string sorting also favors the Ryzen AI Max 390 by 25.1%, with scores of 53,113 and 39,772 respectively.

Single-thread performance is another clear win for the Ryzen AI Max 390, which scores 4,028 compared to the EPYC 9015's 3,265, an 18.9% margin. This is consistent with the clock speed difference between the two chips, as the mobile part boosts to 5.00 GHz while the server part reaches only 4.10 GHz. Prime number finding is the closest contest, with the Ryzen AI Max 390 edging out the EPYC 9015 by just 3.8% (316 vs. 304), suggesting that this particular workload does not scale as strongly with the other advantages.

The EPYC 9015 claims exactly one victory: the physics test. It scores 2,893 against the Ryzen AI Max 390's 2,761, a 4.8% advantage. This is a narrow win, and it stands in contrast to the large margins seen elsewhere. The result is noteworthy because it shows that despite losing the raw compute battles, the server chip retains an edge in a specific simulation-style workload, likely due to its memory architecture and bandwidth advantages. The overall average benchmark scores reflect this split: the EPYC 9015 averages 57,555, while the Ryzen AI Max 390 averages 56,273, putting them less than 2.3% apart in aggregate despite the individual test disparities.

The Verdict

The data is unambiguous for most workloads: the AMD Ryzen AI Max 390 is the faster processor in this head-to-head. It wins 10 of 11 benchmarks, and the margins are often substantial, ranging from double-digit percentage points in encryption, compression, and math tests to a near-tie in prime number finding. Its single-thread score is nearly 19% higher, and its multithread score is over 26% higher. For any user prioritizing raw computational throughput, whether in integer math, floating-point math, or memory-intensive sorting, the Ryzen AI Max 390 is the superior choice based on these results.

The AMD EPYC 9015 is not without its merits, but they are narrower. Its only benchmark win is in physics, where it leads by 4.8%. This suggests that for specific simulation logic, the server chip has a specialized strength. However, the overall benchmark average for the EPYC 9015 (57,555) is actually higher than that of the Ryzen AI Max 390 (56,273), which is a paradox explained by the fact that the EPYC 9015 has other benchmarks in its full profile that pull its average up, while the head-to-head list only covers the overlapping tests. The EPYC 9015 sits at the 92nd percentile of all CPUs, while the Ryzen AI Max 390 sits at the 91st, so they are effectively peers in overall standing.

The choice depends on the platform and workload. If the task is purely about benchmark scores and the system can accommodate a mobile processor, the Ryzen AI Max 390 is the clear pick. If the task involves server-grade reliability, memory bandwidth, or PCIe lane capacity, the EPYC 9015's platform advantages become decisive, even if its raw compute scores are lower. The EPYC 9015 is part of the EPYC 9005 series with a 125 W TDP and twelve-channel memory support, features that do not appear in the benchmark suite but are critical for certain deployments.

Where Each One Wins

The Ryzen AI Max 390 wins in every category of raw compute except physics. Its largest advantages are in integer math (36.9% ahead), floating-point math (32% ahead), and data encryption (29.1% ahead). These are the workloads that benefit most from its higher core count (12 vs. 8), higher thread count (24 vs. 16), and superior boost clock (5.00 GHz vs. 4.10 GHz). For general productivity, development, and content creation tasks that rely on these math and encryption primitives, the Ryzen AI Max 390 is the stronger performer. Its single-thread advantage also makes it better for lightly threaded applications that depend on one or two cores running at maximum speed.

The EPYC 9015 wins specifically in the physics test, where it leads by 4.8%. This is the only head-to-head test it wins, and it suggests a specialization in physics simulation logic that may be tied to its twelve-channel memory bus and higher memory bandwidth of 576.0 GB/s, compared to the Ryzen AI Max 390's quad-channel bus and 256.0 GB/s. The EPYC 9015 also offers 128 PCIe Gen 5 lanes versus the Ryzen AI Max 390's 16 PCIe Gen 4 lanes, making it the only viable option for systems requiring massive I/O expansion. The EPYC 9015's larger TDP of 125 W also indicates it is designed for sustained all-core operation in a server chassis, whereas the Ryzen AI Max 390's 55 W TDP is optimized for mobile efficiency.

FAQ

Q: Which processor has the higher single-thread benchmark score?

A: The AMD Ryzen AI Max 390 scores 4,028 in the PassMark single-thread test, which is 18.9% higher than the AMD EPYC 9015's score of 3,265.

Q: What is the largest performance gap between the two processors in any benchmark?

A: The largest gap is in the integer math test, where the AMD Ryzen AI Max 390 scores 146,519 versus the AMD EPYC 9015's 92,524, a 36.9% difference in favor of the Ryzen AI Max 390.

Q: Does the AMD EPYC 9015 win any benchmark against the Ryzen AI Max 390?

A: Yes, the EPYC 9015 wins the PassMark physics test with a score of 2,893, which is 4.8% higher than the Ryzen AI Max 390's score of 2,761.

Q: How do the two processors compare in overall benchmark average?

A: The AMD EPYC 9015 has an average benchmark score of 57,555, which is slightly higher than the AMD Ryzen AI Max 390's average of 56,273, despite the latter winning most individual head-to-head tests.

Q: What is the difference in multithread performance?

A: The AMD Ryzen AI Max 390 scores 41,737 in the PassMark multithread test, which is 26.5% higher than the AMD EPYC 9015's score of 30,689.

Q: Which processor has a higher percentile ranking among all CPUs?

A: The AMD EPYC 9015 is at the 92nd percentile, while the AMD Ryzen AI Max 390 is at the 91st percentile, making them nearly equivalent in overall standing.

Architecture Differences

Both processors are built on the Zen 5 architecture, but they come from different codenames and target different platforms. The AMD EPYC 9015 uses the Turin codename and is part of the EPYC 9005 series, while the AMD Ryzen AI Max 390 uses the Strix Halo codename and belongs to the Ryzen AI Max generation. Both are manufactured on a 4 nm process at TSMC, with the same die size of 2x 70.6 mm². The EPYC 9015 has a transistor count of 16,630 million, while the Ryzen AI Max 390's transistor count is not listed. The cache hierarchy is identical: 80 KB of L1 per core, 1 MB of L2 per core, and 64 MB of shared L3. Neither processor features 3D V-Cache.

The core configurations differ significantly. The EPYC 9015 has 8 cores and 16 threads, while the Ryzen AI Max 390 has 12 cores and 24 threads. This 50% increase in core and thread count is the primary driver of the Ryzen AI Max 390's multithread performance advantage. Clock speeds also differ: the EPYC 9015 has a base clock of 3.60 GHz and a boost clock of 4.10 GHz, while the Ryzen AI Max 390 has a lower base clock of 3.20 GHz but a much higher boost clock of 5.00 GHz. The higher boost clock explains the Ryzen AI Max 390's 18.9% single-thread advantage.

Memory support is a major architectural division. The EPYC 9015 supports DDR5 memory over a twelve-channel bus, providing a memory bandwidth of 576.0 GB/s. The Ryzen AI Max 390 supports LPDDR5X memory over a quad-channel bus, with a memory bandwidth of 256.0 GB/s. The EPYC 9015's memory bandwidth is more than double that of the Ryzen AI Max 390, which likely contributes to its physics test win. Both processors support ECC memory. The EPYC 9015 uses AMD Socket SP5, while the Ryzen AI Max 390 uses AMD Socket FP11. The EPYC 9015 has no integrated graphics, while the Ryzen AI Max 390 includes a Radeon 8050S iGPU. Neither processor has an unlocked multiplier.

Specification Differences

The table below lists only the specification fields where the two processors differ, based on the data provided.

| Specification | AMD EPYC 9015 | AMD Ryzen AI Max 390 |

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

| Cores | 8 | 12 |

| Threads | 16 | 24 |

| Base Clock | 3.60 GHz | 3.20 GHz |

| Boost Clock | 4.10 GHz | 5.00 GHz |

| TDP | 125 W | 55 W |

| Socket | AMD Socket SP5 | AMD Socket FP11 |

| Codename | Turin | Strix Halo |

| Generation | EPYC (Zen 5 (Turin)) | Ryzen AI Max (Zen 5 (Strix Halo)) |

| Transistors | 16,630 million | Not listed |

| Memory Support | DDR5 | LPDDR5X |

| Memory Bus | Twelve-channel | Quad-channel |

| Memory Bandwidth | 576.0 GB/s | 256.0 GB/s |

| PCIe | Gen 5, 128 Lanes (CPU only) | Gen 4, 16 Lanes (CPU only) |

| Integrated Graphics | N/A | Radeon 8050S |

| Market Segment | Server/Workstation | Mobile |

| Release Date | 2024-10-09 | 2025-01-05 |

| Launch MSRP | $527 | Not listed |

| Part Number | 100-000001553 | 100-000001423 |