AMD Ryzen Threadripper PRO 9985WX vs Intel Xeon 6747P Comparison

Head-to-Head Benchmarks

The recorded benchmark data presents a decisive overall result: the AMD Ryzen Threadripper PRO 9985WX wins 13 of the 14 head-to-head comparisons, with the Intel Xeon 6747P taking a single narrow victory. The largest margins are not subtle. In Passmark integer math, the AMD part scores 872710 against 468518, a delta of 86.3 percent. Random string sorting tells a similar story, with the AMD processor at 329014 versus 180382, an 82.4 percent advantage. These are workloads where core count and memory subsystem throughput dominate, and the data reflects that.

The Cinebench multicore tests are uniformly lopsided. Across Cinebench R15, R20, and R23 multicore, the AMD Ryzen Threadripper PRO 9985WX holds a consistent 53.7 percent lead over the Intel Xeon 6747P. The scores are 13392 versus 8712, 55800 versus 36301, and 132859 versus 86432 respectively. This consistency across three generations of the Cinebench test suggests a stable architectural advantage rather than a workload-specific quirk. The AMD processor's 64 cores and 128 threads are clearly being leveraged effectively in these heavily parallel rendering tasks.

Single-thread performance also favors the AMD part, though by a smaller margin. In Passmark single-thread and single-thread (recorded as two identical entries), the AMD scores 4482 against 3236, a 38.5 percent delta. This is notable because it shows the Zen 5 architecture is not merely scaling with core count; the individual cores themselves are faster. The Intel Xeon 6747P, despite its lower core count, cannot match this per-thread performance.

The only benchmark where the Intel Xeon 6747P comes out ahead is Passmark find prime numbers. Here, Intel scores 1151 versus AMD's 1138, a 1.1 percent margin. It is a narrow win, and it indicates that the Intel architecture holds a slight edge in this specific integer-heavy, latency-sensitive workload. The difference is small enough that it could be attributed to cache behavior or instruction scheduling, but it is a genuine data point in Intel's favor.

Other Passmark metrics show the AMD part winning by substantial but varying margins. Data compression: 2912972 versus 1833378, a 58.9 percent lead. Extended instructions: 225340 versus 142557, a 58.1 percent lead. Floating point math: 553348 versus 365904, a 51.2 percent lead. Data encryption: 154824 versus 90789, a 70.5 percent lead. Multithread: 150071 versus 101685, a 47.6 percent lead. Physics: 13783 versus 13398, a 2.9 percent lead, which is the closest non-prime-number result.

The overall average benchmark score for the AMD Ryzen Threadripper PRO 9985WX sits at 320749, placing it in the 99th percentile of all CPUs. The Intel Xeon 6747P also reaches the 99th percentile, but its average score is 238263. The gap between the two average scores is roughly 34.6 percent, which aligns with the pattern seen in the individual tests.

Where Each One Wins

The AMD Ryzen Threadripper PRO 9985WX is the clear choice for multi-threaded, throughput-oriented workloads. The Cinebench multicore results, the Passmark multithread score, and the integer math and compression tests all point to a processor that excels when all 64 cores and 128 threads can be kept busy. This includes rendering, video encoding, simulation, and large-scale data processing. The 53.7 percent lead in Cinebench multicore and the 86.3 percent lead in integer math are the standout figures for this use case.

The AMD part also wins in single-thread and lightly threaded tasks, which is important for workstation responsiveness and for applications that are not perfectly parallelized. A 38.5 percent lead in Passmark single-thread means that even legacy or poorly threaded software will run faster on the AMD processor. The data encryption and extended instructions results further reinforce the AMD advantage in security-heavy and vectorized workloads.

The Intel Xeon 6747P's only recorded win is in Passmark find prime numbers, where it edges out the AMD part by 1.1 percent. This is a very specific workload involving prime number generation, which is often bound by integer division and memory latency. For a user whose primary application is this type of mathematical computation, the Intel part offers a slight edge, but it is a narrow and isolated victory.

Looking at the broader picture, the Intel Xeon 6747P sits within a competitive field of its own. Its nearest rivals include the AMD EPYC 9634, which is 2.5 percent slower on average, and the Intel Xeon 6980P, which is 5.3 percent slower. The AMD Ryzen Threadripper PRO 9985WX, by contrast, is only 0.3 percent behind the AMD Ryzen Threadripper 9980X and ahead of the Intel Xeon 6781P by 1.7 percent. This suggests that the Threadripper PRO is positioned near the top of its own peer group, while the Xeon 6747P is a mid-tier performer in its segment.

FAQ

Q: Which processor has the higher multicore performance in Cinebench?

A: The AMD Ryzen Threadripper PRO 9985WX leads the Intel Xeon 6747P by 53.7 percent in Cinebench R15, R20, and R23 multicore tests.

Q: Is there any benchmark where the Intel Xeon 6747P wins?

A: Yes, in Passmark find prime numbers, the Intel Xeon 6747P scores 1151 versus 1138 for the AMD part, a 1.1 percent margin.

Q: How do the single-thread scores compare?

A: The AMD Ryzen Threadripper PRO 9985WX scores 4482 in Passmark single-thread, which is 38.5 percent higher than the Intel Xeon 6747P's 3236.

Q: What is the average benchmark score difference?

A: The AMD part has an average benchmark score of 320749, while the Intel part averages 238263, a gap of roughly 34.6 percent.

Q: Where does each processor rank among all CPUs?

A: Both processors sit in the 99th percentile of all CPUs, according to the database.

Q: Which processor has more cores and threads?

A: The AMD Ryzen Threadripper PRO 9985WX has 64 cores and 128 threads, while the Intel Xeon 6747P has 48 cores and 96 threads.

Specification Differences

The two processors differ in nearly every core specification category. The AMD Ryzen Threadripper PRO 9985WX is built on a 4 nm process by TSMC, while the Intel Xeon 6747P uses a 5 nm process from Intel. The AMD part has a die size listed as 8x 70.6 mm², whereas the Intel part is 2x 598 mm². Transistor count is recorded for the AMD processor at 66,520 million, but no transistor figure is listed for the Intel part.

Clock speeds show a clear separation. The AMD processor has a base clock of 3.20 GHz and a boost clock of 5.40 GHz. The Intel processor operates at a base clock of 2.70 GHz and a boost clock of 3.90 GHz. Power envelopes are close: the AMD part has a TDP of 350 watts, and the Intel part is rated at 330 watts. The AMD processor has an unlocked multiplier, while the Intel processor is locked.

Socket compatibility is distinct. The AMD part uses AMD Socket sTR5, and the Intel part uses Intel Socket 4710. Cache layouts also diverge significantly. The AMD processor has 64 KB of L1 cache per core, 1 MB of L2 cache per core, and 256 MB of L3 cache. The Intel processor has 112 KB of L1 cache per core, 2 MB of L2 cache per core, and 288 MB of shared L3 cache. The Intel part has a larger L3 pool, but the AMD part has a larger total cache footprint when considering per-core L1 and L2 allocations.

PCIe lane counts differ. The AMD processor supports Gen 5 with 128 lanes (CPU only), while the Intel processor supports Gen 5 with 88 lanes (CPU only). Both support DDR5 memory with an eight-channel memory bus and a memory bandwidth of 409.6 GB/s. Both also support ECC memory and have no integrated graphics. The AMD part was released on 2025-07-22, and the Intel part was released earlier on 2025-02-23.

Architecture Differences

The architectural split is fundamental. The AMD Ryzen Threadripper PRO 9985WX uses the Zen 5 architecture, codenamed Shimada Peak, and belongs to the 9000 series. The Intel Xeon 6747P uses the Granite Rapids architecture, also codenamed Granite Rapids, and belongs to the Xeon 6 (Granite Rapids-SP) generation. The AMD part is manufactured by TSMC on a 4 nm node, while the Intel part is manufactured by Intel on a 5 nm node.

Core and thread counts are the most visible architectural difference. The AMD processor offers 64 cores and 128 threads, which is 33.3 percent more cores and 33.3 percent more threads than the Intel processor's 48 cores and 96 threads. This explains much of the multicore performance gap, though the single-thread advantage of the AMD part suggests the core design itself is also faster.

Cache architecture reflects different design philosophies. The AMD part uses a per-core L1 and L2 allocation (64 KB and 1 MB respectively) with a large 256 MB L3 cache. The Intel part uses a larger per-core L1 (112 KB) and L2 (2 MB) but a slightly larger shared L3 pool of 288 MB. The Intel part's L3 is explicitly shared, while the AMD part's L3 is not specified as shared in the data. The AMD part's total cache per core is effectively higher when summing L1, L2, and a proportional share of L3.

The process node difference is notable. The TSMC 4 nm node used for the AMD part is a more advanced process than the Intel 5 nm node, which can contribute to higher clock speeds and better power efficiency. The AMD part reaches a boost clock of 5.40 GHz, while the Intel part tops out at 3.90 GHz. This 1.5 GHz gap in boost clock is a major factor in the single-thread performance difference.

PCIe connectivity also differs architecturally. The AMD part provides 128 Gen 5 lanes, while the Intel part provides 88 Gen 5 lanes. This gives the AMD processor more headroom for expansion cards, GPUs, and storage controllers. Both parts target the Server/Workstation market segment and are marked as Active in production status. The AMD part has a part number of 100-000000722, and the Intel part is listed as SRVEZ. The launch MSRP for the AMD part is $7999, and for the Intel part it is $6497.