AMD Ryzen Threadripper 9970X vs Intel Xeon w9-3595X Comparison

Head-to-Head Benchmarks

The recorded data splits this matchup into two distinct narratives: the AMD Ryzen Threadripper 9970X dominates in lightly threaded and latency-sensitive workloads, while the Intel Xeon w9-3595X flexes its massive core count in raw parallel throughput. Out of 11 shared PassMark tests, the AMD part takes 6 wins, while the Intel part takes 5, but the margins tell a more complex story than the split suggests.

The single-thread test is the clearest blowout. The Threadripper 9970X scores 4530 against 3720 for the Xeon w9-3595X, a 21.8% advantage. That gap repeats in the physics test, where the AMD chip posts 6835 against 5842, a 17% lead. These are the two largest deltas in the entire comparison, and they point to a fundamental difference in per-core efficiency.

The AMD chip also wins the multithread test, scoring 107399 versus 99576, a 7.9% edge, despite having only 32 cores and 64 threads. The Intel part counters with 60 cores and 120 threads, yet cannot convert that raw hardware advantage into a PassMark multithread victory. The prime number test also goes to AMD, 615 to 580, a 6% margin, which indicates a decided advantage in integer-heavy iterative workloads. Random string sorting is nearly a tie: AMD edges ahead with 191445 versus 190745, a 0.4% difference.

The Intel Xeon w9-3595X takes its wins in the throughput-oriented tests. Floating point math is its strongest showing: 379008 versus 309719, an 18.3% lead. Data compression also goes to Intel, 1831962 versus 1757998, a 4% margin. Data encryption follows with a 5.9% lead, 92249 versus 86765. Integer math is closer, 473507 versus 465378, a 1.7% advantage for Intel. Extended instructions is the tightest result on the board, with Intel ahead by only 0.3%, 142785 versus 142342.

The average benchmark scores in the database reflect these workloads differently. The Threadripper 9970X has an average benchmark score of 279778, while the Xeon w9-3595X sits at 209881. That discrepancy is largely because the AMD part has more recorded PassMark entries, which skew the average toward its strengths. The nearest rivals for the AMD part include the Intel Xeon 6780E at 280438, just 0.2% ahead, and the AMD EPYC 9565 at 285471, 2% ahead. The Xeon w9-3595X's nearest rivals are the AMD EPYC 9455P at 217854, 3.7% ahead, and the Intel Xeon 6741P at 194901, which the Xeon w9-3595X beats by 7.7%.

Where Each One Wins

The AMD Ryzen Threadripper 9970X is the pick for workloads that are latency-sensitive or rely on high clock speeds. Its 5.40 GHz boost clock, compared to 4.80 GHz for the Intel part, directly explains the single-thread and physics results. The 21.8% single-thread advantage means tasks like code compilation, scripting, and interactive rendering previews will feel snappier. The 17% physics lead suggests the AMD chip is better suited to simulation steps that cannot be parallelized efficiently. The multithread win, despite fewer cores, also indicates that the AMD part's Zen 5 architecture delivers more usable throughput per thread, which benefits workloads with moderate thread counts but high per-thread demands.

The Intel Xeon w9-3595X wins where core count is the dominant factor. Its 60 cores and 120 threads, versus 32 cores and 64 threads for AMD, show up clearly in floating point math, where it leads by 18.3%. The data compression and encryption wins, at 4% and 5.9% respectively, point to workloads that scale almost linearly with thread count and memory bandwidth. The Intel part also has a significant memory bandwidth advantage: 307.2 GB/s over eight channels, compared to 204.8 GB/s over four channels for the AMD chip. That 102.4 GB/s difference helps in large dataset processing, video encoding, and scientific simulations that saturate memory subsystems.

The integer math result, a narrow 1.7% lead for Intel, is notable because it is a test where core count usually dominates. The AMD part's 6% win in prime number finding, which is typically integer-bound, suggests that the AMD chip's per-core performance compensates for its lower core count in specific integer patterns. The extended instructions test, nearly identical at 0.3% apart, shows that both architectures handle AVX-style workloads at parity.

The Verdict

Pick the AMD Ryzen Threadripper 9970X if your workloads are dominated by single-threaded performance, physics simulation, or moderate-thread-count multithreading. The 21.8% single-thread lead and 17% physics lead are decisive. The multithread win at 7.9%, despite having 28 fewer cores and 56 fewer threads, is a strong signal that the AMD architecture extracts more from each thread. The 128 MB L3 cache, versus 112.5 MB for Intel, also helps in scenarios where data fits in cache. This is the right processor for workstations running interactive modeling, real-time rendering, and developer toolchains.

Pick the Intel Xeon w9-3595X if your workloads are massively parallel and memory-bandwidth-hungry. The 18.3% floating point lead and 5.9% encryption lead are entirely convincing in those domains. The 307.2 GB/s memory bandwidth and 112 PCIe Gen 5 lanes, versus 204.8 GB/s and 80 lanes for AMD, matter for systems driving multiple high-speed devices or processing datasets far larger than cache. The data compression win at 4% also favors Intel for archival and database workloads. This is the part for batch rendering, scientific computing, and server-side data processing where thread count is the limiting factor.

The database shows both parts at the 99th percentile versus all CPUs, so neither is a weak choice. The decision is strictly workload fit. For a mixed-use desktop workstation with emphasis on responsiveness, the AMD part wins on the recorded data. For a dedicated throughput machine, the Intel part wins where it matters most.

FAQ

Q: Which CPU has a higher single-thread score?

A: The AMD Ryzen Threadripper 9970X scores 4530 in the PassMark single-thread test, which is 21.8% higher than the Intel Xeon w9-3595X's 3720.

Q: Does the Intel Xeon w9-3595X win any benchmark by a large margin?

A: Yes, the Intel part leads by 18.3% in floating point math, scoring 379008 versus 309719 for the AMD chip. It also leads by 5.9% in data encryption.

Q: How do the core counts compare?

A: The Intel Xeon w9-3595X has 60 cores and 120 threads, while the AMD Ryzen Threadripper 9970X has 32 cores and 64 threads.

Q: Which CPU has more memory bandwidth?

A: The Intel Xeon w9-3595X has 307.2 GB/s over an eight-channel memory bus, while the AMD Ryzen Threadripper 9970X has 204.8 GB/s over a quad-channel bus.

Q: Which CPU wins the multithread benchmark?

A: The AMD Ryzen Threadripper 9970X wins the PassMark multithread test with 107399, which is 7.9% ahead of the Intel Xeon w9-3595X's 99576.

Q: Are both CPUs at the same performance percentile?

A: Yes, both the AMD Ryzen Threadripper 9970X and the Intel Xeon w9-3595X are at the 99th percentile versus all CPUs in the database.

Architecture Differences

The AMD Ryzen Threadripper 9970X is built on a 4 nm process at TSMC, while the Intel Xeon w9-3595X uses a 10 nm process at Intel. The AMD part packs 33,260 million transistors across a die size of 4x 70.6 mm². The Intel part has a larger die footprint at 4x 477 mm², though its transistor count is not recorded in the database.

Both CPUs use DDR5 memory and support ECC. The AMD part is quad-channel with 204.8 GB/s bandwidth; the Intel part is eight-channel with 307.2 GB/s. The Intel part also provides more PCIe Gen 5 lanes: 112 versus 80 for AMD. Neither has integrated graphics.

Cache hierarchies differ. The AMD Ryzen Threadripper 9970X has 64 KB L1 per core, 1 MB L2 per core, and a 128 MB L3 cache. The Intel Xeon w9-3595X has 80 KB L1 per core, 2 MB L2 per core, and a 112.5 MB L3 cache. The AMD part's larger L3, 15.5 MB more, helps in single-thread and moderate-thread workloads that fit in cache.

The architectural lineage is distinct. AMD uses Zen 5 with the codename Shimada Peak, while Intel uses Sapphire Rapids. The AMD part is classified as a desktop market segment, while the Intel part is listed as server/workstation. Both have unlocked multipliers, and both are active production parts. The AMD part released on 2025-07-29, nearly a year after the Intel part's 2024-08-23 release. The launch MSRP for the AMD part is $2499, and for the Intel part it is $5889.

Clock speeds are a major differentiator. The AMD chip has a base clock of 4.00 GHz and a boost clock of 5.40 GHz. The Intel chip has a base clock of 2.00 GHz and a boost clock of 4.80 GHz. The AMD part's higher clocks, combined with a 4 nm node, explain its single-thread dominance. The Intel part's lower clocks but far higher core count explain its throughput advantages. The TDP ratings are 350 W for AMD and 385 W for Intel, though the database does not record cooler requirements or power draw in practical tests. The socket is AMD Socket sTR5 for the Threadripper and Intel Socket 4677 for the Xeon W.