AMD Ryzen Threadripper PRO 9985WX vs Intel Xeon 6960P Comparison

The Intel Xeon 6960P and AMD Ryzen Threadripper PRO 9985WX represent two distinct philosophies for high-end workstation and server processing. The Xeon 6960P, built on Intel’s Granite Rapids architecture, offers 72 cores and 144 threads, while the Threadripper PRO 9985WX, using AMD’s Zen 5 "Shimada Peak" design, provides 64 cores and 128 threads. Benchmark data reveals a clear split in workload preferences: the AMD part wins 10 of 14 head-to-head tests, while the Intel part secures 4 decisive victories in specific computational areas.

Where Each One Wins

The AMD Ryzen Threadripper PRO 9985WX dominates the broad spectrum of productivity and rendering workloads. In Cinebench tests across all three versions (R15, R20, and R23), the Threadripper consistently outperforms the Xeon by a margin of 16.4%. This consistent delta across generations of the Cinebench suite indicates a fundamental advantage in sustained multi-threaded rendering performance. The AMD processor also leads in PassMark’s integer math test, scoring 872710 against Intel's 727750, a 16.6% advantage, and in extended instructions (225340 vs 193404, a 14.2% lead). Multithreaded performance in PassMark also favors AMD, with a score of 150071 compared to 130659, a 12.9% difference. Even in floating-point math, where Intel’s architecture is traditionally strong, the Threadripper edges ahead by 4.7%, scoring 553348 to 527473.

The Intel Xeon 6960P, despite having more cores, wins where raw core count and specific instruction patterns are less relevant. Its most significant victory is in PassMark’s physics test, where it scores 24937 against AMD’s 13783, a staggering 80.9% advantage. This suggests Intel’s architecture handles the physics simulation workload with far greater efficiency. The Xeon also leads in random string sorting, scoring 371795 to 329014, a 13% margin, and in data encryption, where it scores 162013 versus 154824, a 4.6% lead. Finally, in the specific task of finding prime numbers, the Xeon wins decisively with a score of 1484 compared to 1138, a 30.4% advantage. These wins paint a picture of Intel excelling in tasks that may rely on specific instruction sets or memory access patterns rather than pure core throughput.

Architecture Differences

The two processors diverge fundamentally in their silicon design. The Intel Xeon 6960P is built on a 5 nm process node by Intel’s own foundry, while the AMD Ryzen Threadripper PRO 9985WX uses a 4 nm process from TSMC. This difference in process technology contributes to the AMD part’s higher clock speeds and efficiency. The Intel chip has a base clock of 2.70 GHz and a boost clock of 3.90 GHz, whereas the AMD chip operates at a base of 3.20 GHz and boosts to 5.40 GHz. The higher boost clock on the Threadripper is a key factor in its single-threaded performance, where it scores 4482 vs Intel’s 3287, a 26.7% lead.

Core and cache configurations also differ markedly. The Xeon 6960P packs 72 cores with 144 threads across a die size of 3x 598 mm². Its L1 cache is 112 KB per core, L2 is 2 MB per core, and it has a massive 432 MB of shared L3 cache. In contrast, the Threadripper PRO 9985WX has 64 cores and 128 threads spread across 8x 70.6 mm² chiplets, with a total transistor count of 66,520 million. Its cache hierarchy is smaller per core (64 KB L1, 1 MB L2) and its total L3 is 256 MB. Despite having fewer cores and less cache, the AMD part’s higher clocks and newer process node allow it to outpace the Intel chip in most benchmarks. The physical design also informs power characteristics, with the Xeon rated at a 500 W TDP versus the AMD’s 350 W, a difference that is reflected in their respective cooling and power delivery requirements.

Memory and I/O are also differentiators. The Intel Xeon 6960P supports a twelve-channel memory bus with a theoretical bandwidth of 614.4 GB/s, while the AMD Threadripper PRO 9985WX uses an eight-channel configuration with 409.6 GB/s. This gives Intel a substantial bandwidth advantage for memory-intensive workloads. However, the AMD part provides more PCIe lanes: 128 Gen 5 lanes (CPU only) compared to Intel’s 96. Both support DDR5 memory and ECC, but the platform specifics differ, with the Intel using Socket 7529 and the AMD using Socket sTR5.

Head-to-Head Benchmarks

The largest win for the AMD Ryzen Threadripper PRO 9985WX comes in single-threaded performance. In the PassMark single-thread test, the AMD scores 4482 against Intel’s 3287, a delta of 26.7%. This is a substantial gap driven by the AMD’s 5.40 GHz boost clock. This advantage carries over to the Cinebench R23 single-core test, where AMD scores 18756, a figure that is not directly comparable to the Xeon’s multicore score but highlights the architectural efficiency of Zen 5.

In multi-threaded workloads, the AMD part’s lead is consistent but narrower. In Cinebench R23 multicore, the Threadripper scores 132859 versus the Xeon’s 111060. This 16.4% lead is mirrored in Cinebench R20 (55800 vs 46645) and Cinebench R15 (13392 vs 11194). The consistency of this 16.4% delta across all three Cinebench versions indicates a stable performance advantage that scales with thread count, despite the Intel having 12.5% more cores (72 vs 64). The PassMark integer math test shows a similar story, with AMD ahead by 16.6% (872710 vs 727750).

The Intel Xeon 6960P’s most dominant win is in the PassMark physics test. With a score of 24937, it beats the AMD’s 13783 by a massive 80.9%. This is the single largest delta in the entire comparison and suggests that the physics simulation workload is highly sensitive to the Intel architecture’s specific strengths, possibly related to its memory bandwidth or cache design. The Xeon also wins in prime number finding by 30.4% (1484 vs 1138) and in random string sorting by 13% (371795 vs 329014). In data encryption, the Intel part’s 162013 score edges out AMD’s 154824 by 4.6%, a narrow but notable victory for Intel in a security-related task.

The Verdict

From the data, the AMD Ryzen Threadripper PRO 9985WX is the superior choice for general-purpose high-performance computing. Its wins in all Cinebench versions, integer math, floating-point math, extended instructions, and multithreaded PassMark tests show that it is faster for rendering, compilation, and most scientific computing tasks. The 16.4% lead in Cinebench across the board is a decisive factor for anyone running 3D rendering or video encoding. Its single-threaded advantage of 26.7% also makes it more responsive for interactive work and legacy software that relies on one or two cores.

The Intel Xeon 6960P, while losing the majority of benchmarks, holds clear niches. Its 80.9% win in physics is not trivial; for simulation workloads that resemble that PassMark physics test, the Xeon would be dramatically faster. Similarly, its 30.4% lead in prime number finding and 13% lead in random string sorting suggest specific algorithm classes where it excels. The Xeon’s higher memory bandwidth (614.4 GB/s vs 409.6 GB/s) and larger L3 cache (432 MB vs 256 MB) may benefit workloads that are memory-bandwidth-bound, even if those advantages do not translate to wins in the listed benchmarks. For a user prioritizing maximum general performance across a wide range of tasks, the Threadripper is the data-backed winner. For a user with a workload that specifically matches the Xeon’s winning patterns—particularly physics simulations—the Intel part is the one to choose.

FAQ

Q: Which processor has more cores and threads?

A: The Intel Xeon 6960P has 72 cores and 144 threads, while the AMD Ryzen Threadripper PRO 9985WX has 64 cores and 128 threads.

Q: What is the performance difference in Cinebench R23 multicore?

A: The AMD Ryzen Threadripper PRO 9985WX scores 132859, which is 16.4% higher than the Intel Xeon 6960P’s score of 111060.

Q: In which test does the Intel Xeon 6960P have its largest advantage?

A: In the PassMark physics test, the Intel Xeon 6960P scores 24937, which is 80.9% higher than the AMD’s 13783.

Q: How do the single-threaded performances compare?

A: The AMD Ryzen Threadripper PRO 9985WX scores 4482 in PassMark single-thread, which is 26.7% higher than the Intel Xeon 6960P’s score of 3287.

Q: What is the difference in memory bandwidth support?

A: The Intel Xeon 6960P supports a twelve-channel memory bus with 614.4 GB/s bandwidth, while the AMD Ryzen Threadripper PRO 9985WX supports an eight-channel bus with 409.6 GB/s.

Q: Which processor has a higher boost clock speed?

A: The AMD Ryzen Threadripper PRO 9985WX has a boost clock of 5.40 GHz, compared to the Intel Xeon 6960P’s 3.90 GHz.

Specification Differences

The following table lists only the fields where the Intel Xeon 6960P and AMD Ryzen Threadripper PRO 9985WX differ.

| Field | Intel Xeon 6960P | AMD Ryzen Threadripper PRO 9985WX |

| :--- | :--- | :--- |

| Cores | 72 | 64 |

| Threads | 144 | 128 |

| Base Clock | 2.70 GHz | 3.20 GHz |

| Boost Clock | 3.90 GHz | 5.40 GHz |

| TDP | 500 W | 350 W |

| Socket | Intel Socket 7529 | AMD Socket sTR5 |

| Architecture | Granite Rapids | Zen 5 |

| Codename | Granite Rapids | Shimada Peak |

| Process Node | 5 nm | 4 nm |

| Foundry | Intel | TSMC |

| Transistors | Not specified | 66,520 million |

| Die Size | 3x 598 mm² | 8x 70.6 mm² |

| L1 Cache | 112 KB (per core) | 64 KB (per core) |

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

| L3 Cache | 432 MB (shared) | 256 MB |

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

| Memory Bandwidth | 614.4 GB/s | 409.6 GB/s |

| PCIe | Gen 5, 96 Lanes (CPU only) | Gen 5, 128 Lanes (CPU only) |

| Launch MSRP | $9625 | $7999 |

| Multiplier Unlocked | false | true |

| Part Number | SRPKX | 100-000000722 |

| Release Date | 2024-09-23 | 2025-07-22 |

| Percentile vs All CPUs | 100 | 99 |

| Average Benchmark Score | 365194 | 320749 |

| Nearest Rival (avgScore, deltaPct) | AMD EPYC 9754 (364371, 0.2%) | AMD Ryzen Threadripper 9980X (321753, -0.3%) |