AMD Ryzen 9 PRO 9965 vs Intel Xeon 674X Comparison

The AMD Ryzen 9 PRO 9965 and Intel Xeon 674X represent two fundamentally different approaches to high-end workstation computing, and the benchmark data reflects that divergence clearly. The Intel Xeon 674X wins 9 of the 11 head-to-head comparisons, often by substantial margins, while the AMD Ryzen 9 PRO 9965 takes the remaining 2 wins, both in single-threaded performance. The data shows a 19% advantage for AMD in single-thread tests, but Intel counters with dominance across nearly every multi-threaded and specialized workload, ranging from a 20.8% lead in multithread to a commanding 57.1% lead in physics. The average benchmark scores are remarkably close—145,728 for AMD versus 143,103 for Intel—placing both processors in the 98th percentile of all CPUs, but the distribution of strengths could not be more different.

The Verdict

The data presents a clear split: the Intel Xeon 674X is the pick for multi-threaded intensive workloads, while the AMD Ryzen 9 PRO 9965 is the choice for single-threaded responsiveness. With 28 cores and 56 threads, the Xeon 674X leverages its 47.5% advantage in prime number finding and 34.1% lead in floating-point math to dominate computational tasks. The AMD chip, with 16 cores and 32 threads, compensates with a higher boost clock of 5.50 GHz compared to the Intel's 4.90 GHz, translating to a 19% win in PassMark's single-thread test.

For users running physics simulations, the Xeon 674X is the unambiguous pick—its 7,586 score versus the Ryzen's 3,256 represents a 57.1% delta, the largest margin in the entire comparison. Similarly, data compression workloads favor Intel by 26.5%, with scores of 1,236,272 versus 908,293. The Xeon's 144 MB of shared L3 cache and 409.6 GB/s of memory bandwidth provide the infrastructure for these wins.

Conversely, the Ryzen 9 PRO 9965 is the pick for single-threaded applications, legacy software, or any workload that scales poorly across cores. Its 4,682 single-thread score is not just a win; it's a 19% improvement over the Xeon's 3,933. The AMD chip also offers integrated Radeon Graphics, whereas the Xeon has no integrated graphics, making the Ryzen a more self-contained option for basic display output.

The closest rival comparisons reinforce this split: the Ryzen 9 PRO 9965's nearest rival is the AMD EPYC 7643P at 0.6% delta, while the Xeon 674X sits near the Intel Xeon 6732P at -0.2%. Both processors are within 1.2% of their nearest rivals, indicating tight competition at this tier. Ultimately, the choice comes down to thread scaling: if the workload uses all available cores, the Xeon 674X wins decisively; if it relies on single-core speed, the Ryzen 9 PRO 9965 is superior.

Architecture Differences

The architectural divergence is foundational. The AMD Ryzen 9 PRO 9965 uses the Granite Ridge codename, built on a 4 nm process by TSMC, with 16,630 million transistors across a dual-die design (2x 70.6 mm²). The Intel Xeon 674X uses Granite Rapids, fabricated on Intel's 5 nm node, with a much larger dual-die footprint of 2x 598 mm². This size difference reflects the Xeon's higher core count: 28 cores versus AMD's 16, and 56 threads versus 32.

Cache hierarchies differ substantially. The AMD chip provides 80 KB of L1 per core, 1 MB of L2 per core, and 64 MB of L3 cache. The Intel chip offers 112 KB of L1 per core, 2 MB of L2 per core, and a massive 144 MB of shared L3 cache—more than double AMD's total. This cache advantage directly supports the Xeon's wins in data compression and random string sorting, which benefit from larger working sets held on-chip.

Memory subsystems also diverge. Both support DDR5 and ECC memory, but the Xeon 674X uses an eight-channel bus with a theoretical bandwidth of 409.6 GB/s, while the Ryzen 9 PRO 9965 uses dual-channel with 89.6 GB/s. This 4.5x bandwidth advantage for Intel is a critical enabler for its multi-threaded dominance. PCIe connectivity mirrors this: the Xeon offers Gen 5 with 128 lanes, versus AMD's Gen 5 with 24 lanes (CPU only).

The Xeon 674X has an unlocked multiplier, while the Ryzen 9 PRO 9965 does not, though the TDP figures tell the power story: 270 W for Intel versus 170 W for AMD. The Xeon's higher power envelope is consistent with its additional cores and bandwidth. The AMD chip includes integrated Radeon Graphics; the Xeon has N/A for integrated graphics, requiring a discrete GPU for display. The Xeon 674X is a server/workstation part with a launch MSRP of $2199, while the Ryzen 9 PRO 9965 also targets the server/workstation segment without a listed launch MSRP.

Head-to-Head Benchmarks

The Intel Xeon 674X dominates the multi-threaded benchmarks. The largest win is in PassMark physics, where Intel scores 7,586 against AMD's 3,256—a 57.1% delta. This is followed by prime number finding, where Intel's 693 beats AMD's 364 by 47.5%. Floating-point math shows Intel ahead by 34.1% (243,877 versus 160,746), and extended instructions by 26.9% (97,373 versus 71,210). Data encryption shows Intel at 61,195 versus AMD's 44,920, a 26.6% delta.

Data compression, a key server workload, goes to Intel at 1,236,272 versus 908,293—a 26.5% lead. Random string sorting follows a similar pattern: Intel at 127,529 versus AMD's 94,915, a 25.6% delta. Integer math shows Intel at 308,968 versus AMD's 243,280, a 21.3% advantage. The multithread score, the broadest measure of parallel performance, has Intel at 84,196 versus AMD's 66,655, a 20.8% lead.

The AMD Ryzen 9 PRO 9965 wins only two benchmarks, both single-threaded, with identical scores of 4,682 versus Intel's 3,933. This 19% delta is the only category where AMD leads, but it is a significant one. The single-thread score appears twice in the dataset (once as passmark_single_thread, once as passmark_singlethread), and both confirm the same result. No other benchmark in the comparison shows AMD ahead.

The overall win count is decisive: Intel wins 9, AMD wins 2. However, the average benchmark scores are nearly identical—145,728 for AMD versus 143,103 for Intel, a difference of just 1.8%. This suggests that AMD's single-thread wins carry disproportionate weight in the average calculation, while Intel's multi-threaded wins, though numerous, are spread across more tests.

FAQ

Q: Which processor has the higher single-thread performance?

A: The AMD Ryzen 9 PRO 9965 wins both single-thread benchmarks with a score of 4,682, compared to the Intel Xeon 674X's 3,933, representing a 19% advantage for AMD.

Q: How much faster is the Intel Xeon 674X in physics simulations?

A: The Xeon 674X scores 7,586 in PassMark physics, while the Ryzen 9 PRO 9965 scores 3,256. This is a 57.1% delta in favor of Intel, the largest margin in the entire comparison.

Q: What memory bandwidth does each processor support?

A: The Intel Xeon 674X supports eight-channel DDR5 with a theoretical bandwidth of 409.6 GB/s. The AMD Ryzen 9 PRO 9965 supports dual-channel DDR5 with 89.6 GB/s.

Q: Does either processor have integrated graphics?

A: The AMD Ryzen 9 PRO 9965 includes Radeon Graphics. The Intel Xeon 674X lists N/A for integrated graphics, meaning a discrete GPU is required for display output.

Q: What are the core and thread counts?

A: The Intel Xeon 674X has 28 cores and 56 threads. The AMD Ryzen 9 PRO 9965 has 16 cores and 32 threads.

Q: Which processor has a higher average benchmark score?

A: The AMD Ryzen 9 PRO 9965 has an average benchmark score of 145,728, slightly higher than the Intel Xeon 674X's 143,103. Both are in the 98th percentile of all CPUs.

Where Each One Wins

The Intel Xeon 674X wins in every multi-threaded and specialized workload category, making it the clear choice for compute-heavy tasks. Its 57.1% lead in physics is ideal for engineering simulations, molecular modeling, and game physics. The 47.5% advantage in prime number finding suggests strength in cryptography and number theory applications. Floating-point math, where Intel leads by 34.1%, benefits scientific computing, financial modeling, and any workload relying on heavy arithmetic. The 20.8% multithread win covers general parallel workloads, while the 26.5% data compression lead suits database, archival, and streaming applications.

The Xeon's 26.6% encryption win and 21.3% integer math lead further cement its position for server-side workloads. Memory bandwidth is a key enabler—the Xeon's 409.6 GB/s versus AMD's 89.6 GB/s explains why Intel dominates in memory-sensitive benchmarks like random string sorting (25.6% lead) and extended instructions (26.9% lead). The 56 threads versus 32 threads, combined with 144 MB of L3 cache versus 64 MB, gives Intel a structural advantage in throughput.

The AMD Ryzen 9 PRO 9965 wins only in single-threaded performance, with a 19% lead. This makes it the pick for lightly threaded applications, legacy codebases that don't scale, or any scenario where per-core speed matters more than total throughput. Its higher boost clock of 5.50 GHz (versus Intel's 4.90 GHz) directly supports this strength. The integrated Radeon Graphics also gives AMD an edge for systems that need display output without a discrete GPU.

The power profile favors AMD as well—170 W TDP versus Intel's 270 W—suggesting lower power draw for single-threaded or lightly threaded workloads. The Ryzen's 4 nm process versus Intel's 5 nm process likely contributes to its efficiency advantage, though the data does not directly measure power efficiency. For users who prioritize single-core responsiveness, the Ryzen 9 PRO 9965 is the data-backed choice; for everything else, the Xeon 674X wins.