AMD Ryzen Threadripper PRO 9995WX vs Intel Xeon 696X Comparison

The AMD Ryzen Threadripper PRO 9995WX sweeps this comparison in the recorded data: 14 benchmark wins to zero. The Intel Xeon 696X is a capable Granite Rapids workstation processor in its own right, sitting in the 99th percentile of all CPUs in the database, but the Threadripper PRO 9995WX sits in the 100th percentile and beats it in every single head-to-head test, by margins ranging from 21.4 percent in single-threaded work to 398.5 percent in physics simulation. For anyone choosing strictly on benchmark performance, the data points one way, and it is not close.

The Verdict

The Threadripper PRO 9995WX is the pick for maximum throughput in every workload category the database measures. Its average benchmark score of 412068 outpaces the Xeon 696X's 286102 by a wide gap, and it wins all 14 recorded head-to-head tests. Even in single-threaded Passmark performance, where core counts are irrelevant, the Threadripper leads 4542 to 3742, a 21.4 percent advantage driven by its higher boost clock and newer Zen 5 architecture.

The Xeon 696X has a defensible position only in contexts the benchmarks do not measure directly: a different platform (Intel Socket 4710 versus AMD Socket sTR5), an earlier launch MSRP of $5599 versus $11700 for the Threadripper PRO, and a launch date of 2026-02-01 versus 2025-07-22. Within the database's own measurements, however, there is no scenario in which the Xeon finishes ahead. Buyers prioritizing raw recorded performance should choose the 9995WX; those already invested in the Intel platform, or constrained on launch MSRP, are the only profiles for which the 696X makes sense from this data.

Architecture Differences

The two processors come from opposite ends of the design spectrum. The Threadripper PRO 9995WX is built on AMD's Zen 5 architecture, codenamed Shimada Peak, on a 4 nm TSMC process. It carries 96 cores and 192 threads with a base clock of 2.50 GHz and a boost clock of 5.40 GHz. Its chiplet design is explicit in the data: 12 dies of 70.6 mm² each, with 99,780 million transistors total.

The Xeon 696X uses Intel's Granite Rapids architecture (Granite Rapids-WS) on Intel's own 5 nm process, with 64 cores and 128 threads, a 2.40 GHz base clock, and a 4.80 GHz boost clock. Its package is 2 dies of 598 mm² each; the database records no transistor count for it.

Cache strategies diverge sharply. The Threadripper gives each core 64 KB of L1 and 1 MB of L2, with 384 MB of L3 across the package. The Xeon offers fatter per-core allocations, 112 KB of L1 and 2 MB of L2 per core, but less total L3 at 336 MB shared. Both parts support DDR5 on an eight-channel bus with identical 409.6 GB/s of memory bandwidth, both support ECC memory, and both expose 128 lanes of PCIe Gen 5 from the CPU. Neither has integrated graphics. Both ship with a 350 W TDP, and both have unlocked multipliers for overclocking.

FAQ

Q: Which CPU is faster overall?

A: The Threadripper PRO 9995WX, decisively. Its average benchmark score of 412068 exceeds the Xeon 696X's 286102, and it wins all 14 head-to-head benchmark comparisons in the database.

Q: How big is the multi-core gap?

A: Very large. Cinebench R23 multi-core reads 148601 for the Threadripper versus 89227 for the Xeon, a 66.5 percent margin. The R15 and R20 multi-core results show the same 66.5 percent delta.

Q: Is the Xeon faster in any workload?

A: No. The database shows zero benchmark wins for the Xeon 696X across all recorded tests, including single-threaded Passmark, where it scores 3742 against 4542.

Q: How does each compare to its own nearest rivals?

A: The Threadripper PRO 9995WX trades blows with EPYC parts: it trails the AMD EPYC 9745 by 3.3 percent, leads the EPYC 9655P by 3.6 percent, and is ahead of the EPYC 9535 and EPYC 9655 by 8.6 and 10.3 percent respectively. The Xeon 696X sits almost exactly level with the AMD EPYC 9565 (0.2 percent ahead) and the EPYC 9555P (0.3 percent behind), while leading the Intel Xeon 6780E by 2 percent and the Ryzen Threadripper 9970X by 2.3 percent.

Q: Do they have the same platform requirements?

A: No. The Threadripper uses AMD Socket sTR5; the Xeon uses Intel Socket 4710. Both are Server/Workstation segment parts with active production status, and both platforms provide eight-channel DDR5 with ECC.

Q: What are the launch MSRPs?

A: The Threadripper PRO 9995WX launched at an MSRP of $11700, while the Xeon 696X launched at an MSRP of $5599.

Specification Differences

Only the fields where the two differ:

• Cores/Threads: 96/192 (Threadripper PRO 9995WX) vs 64/128 (Xeon 696X)
• Base clock: 2.50 GHz vs 2.40 GHz
• Boost clock: 5.40 GHz vs 4.80 GHz
• Architecture: Zen 5 (Shimada Peak) vs Granite Rapids
• Process node: 4 nm (TSMC) vs 5 nm (Intel)
• Transistors: 99,780 million vs not recorded
• Die configuration: 12x 70.6 mm² vs 2x 598 mm²
• L1 cache: 64 KB per core vs 112 KB per core
• L2 cache: 1 MB per core vs 2 MB per core
• L3 cache: 384 MB vs 336 MB shared
• Socket: AMD Socket sTR5 vs Intel Socket 4710
• Release date: 2025-07-22 vs 2026-02-01
• Launch MSRP: $11700 vs $5599
• Part number: 100-000001361 vs SRWQ7

Everything else in the recorded specifications is identical: DDR5 support, eight-channel memory bus, 409.6 GB/s bandwidth, ECC, Gen 5 PCIe with 128 CPU lanes, no integrated graphics, 350 W TDP, unlocked multiplier, Server/Workstation segment, and active production status.

Head-to-Head Benchmarks

The Threadripper PRO 9995WX wins every recorded test, but the size of the win varies enormously by workload, and that variation tells the real story.

Cinebench multi-core results are almost eerily consistent: 14978 vs 8994 in R15, 62412 vs 37475 in R20, and 148601 vs 89227 in R23, all exactly 66.5 percent in AMD's favor. That figure roughly tracks the core-count difference (96 vs 64, a 50 percent advantage) plus the architectural and clock advantages of Zen 5, suggesting the Threadripper scales efficiently across its full die stack rather than relying on bursty behavior.

Passmark's multi-threaded suite confirms the picture: 171200 vs 104974 (63.1 percent). Individual subtests swing wider. Integer math is the Threadripper's standout, 1203634 vs 572072, a 110.4 percent margin that far exceeds the core-count ratio, indicating a genuine per-core advantage in integer workloads on Zen 5. Random string sorting lands at 418973 vs 180392, a 132.3 percent gap. Data encryption shows 206662 vs 112529 (83.7 percent), find-prime-numbers 1476 vs 853 (73 percent), data compression 3723652 vs 2264907 (64.4 percent), extended instructions 270647 vs 172975 (56.5 percent), and floating point math 725066 vs 450164 (61.1 percent).

The single most lopsided result is Passmark Physics: 16860 vs 3382, a 398.5 percent margin for the Threadripper. This is nearly five times the Xeon's score and is far beyond anything core counts alone explain, marking physics simulation as the most architecture-sensitive test in the set and the clearest differentiator between Zen 5 and Granite Rapids in this database.

Single-threaded Passmark reads 4542 vs 3742, a 21.4 percent Threadripper win that aligns with its 5.40 GHz boost clock against the Xeon's 4.80 GHz, plus the newer architecture. The Xeon records no Cinebench single-core entries in the database, so the Passmark single-thread result is the only per-core comparison available, and it too goes to AMD.

Where Each One Wins

The honest answer from the data: the Threadripper PRO 9995WX wins everywhere. There is no recorded workload, threaded or single-threaded, integer, floating point, cryptographic, or memory-sensitive, in which the Xeon 696X finishes ahead.

Within that sweep, the gradations still matter for buyers. The Threadripper's largest advantages come in physics simulation (398.5 percent), string sorting (132.3 percent), and integer math (110.4 percent), making it the clear choice for simulation-heavy engineering, scientific integer workloads, and large-scale data processing. Its 60 to 85 percent leads in encryption, compression, floating point, and the Cinebench suite cover rendering and content creation comprehensively.

The Xeon 696X's case rests entirely on context outside the benchmark scores: it competes in a rival cluster that includes the AMD EPYC 9565, EPYC 9555P, Intel Xeon 6780E, and Ryzen Threadripper 9970X, all within 2.3 percent of its average score, meaning it delivers genuine workstation-class throughput, just not at the 9995WX's altitude. Its $5599 launch MSRP versus $11700, its Socket 4710 platform, and its denser per-core cache (112 KB L1 and 2 MB L2 per core) are the factors a buyer would weigh against a database that, on pure performance, awards every single contest to the Threadripper PRO 9995WX.