AMD Ryzen Threadripper PRO 9965WX vs Intel Xeon 6730P Comparison

The AMD Ryzen Threadripper PRO 9965WX and the Intel Xeon 6730P occupy the same server and workstation segment, but the recorded data shows they are not close competitors in throughput. The Threadripper PRO 9965WX wins 16 of 17 head-to-head benchmarks, including every Cinebench render test and nearly every PassMark workload, while the Xeon 6730P takes a single win in physics simulation. The gap is largest in single-threaded work, where AMD leads by a substantial margin, and the two chips sit at the 98th and 97th percentiles of all CPUs in the database respectively.

Head-to-Head Benchmarks

The Cinebench suite tells a consistent story. In Cinebench R15 multi-core, the Threadripper PRO 9965WX scores 8162 against 6349 for the Xeon 6730P, a 28.6 percent lead. R20 multi-core lands at 34009 versus 26458, a 28.5 percent gap, and R23 multi-core is 80976 versus 62996, again 28.5 percent in AMD's favor. Single-core results mirror this precisely: 1152 versus 896 in R15, 4801 versus 3735 in R20, and 11431 versus 8893 in R23, all roughly 28.5 percent advantages for the Threadripper. Rendering performance per core and per chip both favor AMD by a wide, uniform margin.

PassMark results broaden the picture. Integer math is a 20.1 percent win for the Threadripper (349195 to 290740), data compression runs 18.2 percent faster (1345230 to 1138470), and data encryption matches that 18.2 percent spread (66155 to 55964). Random string sorting is AMD's biggest PassMark victory at 31.3 percent (149617 versus 113919), while the overall multi-thread rating goes to the Threadripper 92604 to 74113, a 24.9 percent lead. Extended instructions (108753 versus 96204) and prime number searches (752 versus 686) also fall to AMD at 13.0 and 9.6 percent respectively.

The single-thread rating is the most lopsided result of the entire comparison: 4551 for the Threadripper PRO 9965WX against 2995 for the Xeon 6730P, a 52 percent advantage. Any workload bounded by one fast thread, from CAD interactions to script-driven scene assembly, will feel dramatically different on the AMD part.

The one Intel win is physics. The Xeon 6730P scores 8606 in PassMark physics against 7529 for the Threadripper, a 12.5 percent edge, and floating point math is nearly tied at 226838 versus 229685, a 1.3 percent AMD lead. Floating-point-heavy simulation is the only area where the two trade blows rather than one side dominating.

FAQ

Q: Which CPU is faster overall?

A: The AMD Ryzen Threadripper PRO 9965WX. It wins 16 of 17 recorded benchmarks, holds an average benchmark score of 147009 against 124756 for the Xeon 6730P, and sits in the 98th percentile versus Intel's 97th percentile across the database.

Q: Does the Intel Xeon 6730P win anything?

A: Yes. PassMark physics, where it scores 8606 against 7529 for the Threadripper PRO 9965WX, an advantage of 12.5 percent. It is Intel's only head-to-head win in the recorded data.

Q: How large is the single-threaded performance gap?

A: Very large. The Threadripper posts a PassMark single-thread score of 4551 versus 2995 for the Xeon, a 52 percent advantage, and leads by roughly 28.5 percent in every Cinebench single-core test.

Q: Do the two chips differ in memory capability?

A: No. Both support DDR5 on an eight-channel bus with 409.6 GB/s of bandwidth and ECC memory. Memory subsystems are effectively identical on paper.

Q: How do they compare on PCIe connectivity?

A: The Threadripper PRO 9965WX offers Gen 5 with 128 CPU lanes, while the Xeon 6730P offers Gen 5 with 88 CPU lanes. GPU-heavy or accelerator-heavy workstation builds benefit from AMD's additional lanes.

Q: Which chip is more power-hungry?

A: The Threadripper, with a TDP of 350 against 250 for the Xeon 6730P. The Intel part is the cooler-running of the two by a TDP difference of 100.

Where Each One Wins

The Threadripper PRO 9965WX is the clear choice for rendering, compilation, and throughput-driven content creation. It wins every Cinebench multi-core test by nearly 29 percent, the PassMark multi-thread rating by 24.9 percent, and integer-heavy workloads by 20.1 percent. Its 52 percent single-thread lead also makes it the better platform for applications that still lean on one main thread, a common trait in DCC tools and engineering software. Combined with 128 Gen 5 PCIe lanes, it suits multi-GPU rendering farms, local AI development boxes, and storage-dense workstations.

The Xeon 6730P carves out a narrower niche. Its 12.5 percent physics win and near-parity in floating point math (a 1.3 percent gap) suggest it performs competitively in simulation-adjacent workloads, and its lower 250 TDP means it fits thermal envelopes the 350-watt Threadripper cannot. Its larger core count of 32 against 24, delivered at lower clocks, points toward deployments where many modest parallel tasks, rather than fast threads, are the priority. For buyers anchored to Socket 4710 platforms, it is the only option of the two.

Context from the database reinforces the gap. The Threadripper's average score of 147009 places it within 0.1 percent of the AMD EPYC 8434P (146881) and 0.9 percent ahead of the AMD Ryzen 9 PRO 9965 (145728), putting it in genuinely elite company. The Xeon 6730P's 124756 average is comparable to the AMD Ryzen Threadripper PRO 5975WX (124171, a 0.5 percent gap) and the AMD EPYC 7642 (124006, 0.6 percent), a solid but visibly lower tier.

Specification Differences

The cores and threads split in Intel's favor: 32 cores and 64 threads for the Xeon 6730P versus 24 cores and 48 threads for the Threadripper PRO 9965WX. Clocks split the other way, with AMD running a 4.20 base and 5.40 boost against Intel's 2.50 base and 3.80 boost. That clock disparity explains much of the per-core benchmark gap.

TDP differs at 350 for AMD and 250 for Intel. Sockets are incompatible: AMD Socket sTR5 versus Intel Socket 4710. PCIe lane counts diverge sharply, 128 Gen 5 lanes for the Threadripper versus 88 for the Xeon. AMD ships an unlocked multiplier; Intel does not. Launch dates differ, with the Xeon arriving in February 2025 and the Threadripper in July 2025. The launch MSRP is $2899 for the Threadripper PRO 9965WX and $3726 for the Xeon 6730P.

Everything else on the spec sheet lines up: DDR5 support, eight-channel memory buses, 409.6 GB/s bandwidth, ECC memory, no integrated graphics, active production status, and a shared server/workstation market segment.

Architecture Differences

The two chips come from opposite design philosophies. AMD's Threadripper PRO 9965WX uses the Zen 5 architecture, codenamed Shimada Peak, built on TSMC's 4 nm process. Intel's Xeon 6730P uses Granite Rapids, specifically Granite Rapids-SP, on Intel's own 5 nm process.

Cache hierarchies differ in structure. The Threadripper gives each core 64 KB of L1 and 1 MB of L2, with 128 MB of L3. The Xeon gives each core 112 KB of L1 and 2 MB of L2, with a larger 288 MB shared L3. Intel's approach favors per-core cache capacity and a big shared pool, while AMD's favors clock speed and density.

Chip packaging also diverges. The Threadripper is a multi-chip design with four dies of 70.6 mm² each and 33,260 million transistors, while the Xeon uses two dies of 598 mm² each; transistor count is not recorded for the Intel part. The die-size contrast illustrates the split: AMD spreads its silicon across smaller, denser 4 nm chiplets, and Intel consolidates into larger 5 nm compute tiles. With 28.5 percent class leads in rendering, a 52 percent single-thread advantage, and doubled PCIe lanes, the recorded data positions the Threadripper PRO 9965WX as the performance leader of this pairing, with the Xeon 6730P answering only in physics and power efficiency.