AMD Ryzen 9 PRO 9965X3D vs Intel Xeon 674X Comparison

# AMD Ryzen 9 PRO 9965X3D vs Intel Xeon 674X

The AMD Ryzen 9 PRO 9965X3D and Intel Xeon 674X sit at opposite ends of the workstation spectrum, despite landing within 0.4% of each other in average benchmark score. The Ryzen 9 PRO wins only the single-threaded tests, while the Xeon 674X dominates every multi-threaded and throughput-oriented workload in the head-to-head comparison. The data paints a clear picture: this is a contest between a 16-core, 32-thread Zen 5 part with a massive L3 cache and a 28-core, 56-thread Granite Rapids Xeon with eight-channel memory bandwidth. The Xeon takes 9 of 11 benchmark wins, but the Ryzen’s single-thread advantage of 17.3% is the kind of margin that matters in lightly threaded workflows.

Where Each One Wins

The Intel Xeon 674X wins every heavily parallel workload in the head-to-head set. Its largest victories come in floating-point math (35.6% ahead), physics (36.7% ahead), and data encryption (28.3% ahead). The Xeon also leads in integer math by 23%, data compression by 27.6%, extended instructions by 27.3%, and random string sorting by 27.2%. These are not narrow margins—they are consistent, double-digit advantages across the board. The Xeon’s multithread score of 84196 versus the Ryzen’s 68225 (a 19% gap) reinforces that this processor is built for sustained, all-core throughput.

The AMD Ryzen 9 PRO 9965X3D wins exactly one type of test: single-threaded performance. Its score of 4614 beats the Xeon’s 3933 by 17.3%. That is a substantial lead for any application that relies on a single core, such as older software, certain simulation tools, or interactive workloads. The Ryzen also carries a 98th percentile ranking among all CPUs, matching the Xeon’s percentile, but its wins are concentrated rather than spread across the benchmark suite. If your work is dominated by single-core responsiveness, the Ryzen is the clear choice; if it scales across cores, the Xeon is unbeatable in this matchup.

Architecture Differences

The two processors come from fundamentally different design philosophies. The Ryzen 9 PRO 9965X3D uses AMD’s Zen 5 architecture, codenamed Granite Ridge, built on a 4nm process at TSMC. It features 16 cores and 32 threads, with a base clock of 4.30 GHz and a boost clock of 5.50 GHz. The Xeon 674X uses Intel’s Granite Rapids architecture, also codenamed Granite Rapids but for the Xeon 600 series, built on a 5nm process at Intel. It doubles the core count to 28 cores and 56 threads, but clocks lower at 3.00 GHz base and 4.90 GHz boost.

Cache configurations diverge sharply. The Ryzen has 80 KB of L1 per core, 1 MB of L2 per core, and 128 MB of L3 cache—a figure that hints at the 3D V-Cache design lineage even though the field does not explicitly label it as such. The Xeon counters with 112 KB of L1 per core, 2 MB of L2 per core, and 144 MB of shared L3. The Xeon’s per-core cache is larger, but the Ryzen’s total L3 is close, and the Ryzen’s higher clocks give it an edge in latency-sensitive single-thread work. The Xeon’s die size of 2x 598 mm² dwarfs the Ryzen’s 2x 70.6 mm², reflecting the massive difference in transistor budget and core complexity.

Memory and I/O also separate the two. The Ryzen supports dual-channel DDR5 with 89.6 GB/s of bandwidth, while the Xeon uses eight-channel DDR5 with 409.6 GB/s—a 4.6x advantage. The Xeon also provides 128 PCIe Gen 5 lanes versus the Ryzen’s 24 lanes. Both support ECC memory, but the Xeon is clearly designed for memory-intensive servers and workstations. The Ryzen includes integrated Radeon Graphics; the Xeon has no integrated graphics. The Ryzen’s TDP is 170W versus the Xeon’s 270W, and the Xeon is multiplier-unlocked while the Ryzen is not.

Head-to-Head Benchmarks

The largest single win for the Xeon 674X is in floating-point math, where it scores 243877 against the Ryzen’s 157123—a 35.6% advantage. This is a massive gap that will show up in scientific computing, financial modeling, and any workload heavy on FP operations. Physics simulation follows closely, with the Xeon at 7586 versus 4801, a 36.7% lead. These two tests alone make the Xeon the obvious pick for compute-heavy rendering or simulation tasks.

The Xeon’s data encryption score of 61195 versus 43905 (28.3% ahead) and data compression score of 1236272 versus 895563 (27.6% ahead) indicate strong performance in data-centric enterprise workloads. Extended instructions (97373 vs 70797) and random string sorting (127529 vs 92886) both favor the Xeon by roughly 27%. Integer math goes to the Xeon at 308968 versus 237955, a 23% edge. Even the prime number test, a more latency-sensitive workload, favors the Xeon by 13.6% (693 vs 599).

The only bright spot for the Ryzen is single-thread performance. Its score of 4614 versus 3933 gives it a 17.3% win, and that same margin appears in both the single_thread and singlethread test entries. This is not a trivial advantage; it means the Ryzen will feel snappier in day-to-day use and in applications that cannot utilize more than one or two cores. The multithread score of 68225 for the Ryzen versus 84196 for the Xeon (19% deficit) confirms that the Ryzen’s core count disadvantage is real, but the single-thread lead partially compensates in mixed workloads.

FAQ

Q: Which processor has better single-thread performance?

A: The AMD Ryzen 9 PRO 9965X3D wins single-thread benchmarks with a score of 4614, beating the Intel Xeon 674X’s 3933 by 17.3%.

Q: How large is the Xeon’s lead in multi-threaded workloads?

A: The Xeon 674X scores 84196 in the multithread test versus the Ryzen’s 68225, a 19% advantage. Its largest wins are 35.6% in floating-point math and 36.7% in physics.

Q: Do both processors support ECC memory?

A: Yes, both the Ryzen 9 PRO 9965X3D and the Xeon 674X support ECC memory, which is critical for workstation reliability.

Q: What are the core and thread counts?

A: The Ryzen has 16 cores and 32 threads, while the Xeon has 28 cores and 56 threads—a 75% core advantage for Intel.

Q: Which processor has more PCIe lanes?

A: The Xeon 674X provides 128 PCIe Gen 5 lanes, far exceeding the Ryzen’s 24 lanes. This favors the Xeon for multi-GPU or high-speed storage configurations.

Q: Is there a clock speed difference?

A: The Ryzen boosts to 5.50 GHz with a 4.30 GHz base, while the Xeon boosts to 4.90 GHz with a 3.00 GHz base. The Ryzen’s higher clocks explain its single-thread dominance.

Specification Differences

The two CPUs differ in nearly every measurable specification. The Ryzen 9 PRO 9965X3D has 16 cores and 32 threads, while the Xeon 674X has 28 cores and 56 threads. Base clocks are 4.30 GHz versus 3.00 GHz, and boost clocks are 5.50 GHz versus 4.90 GHz, both favoring AMD. TDP is 170W for the Ryzen and 270W for the Xeon. The Ryzen uses AMD Socket AM5; the Xeon uses Intel Socket 4710. The Ryzen is built on a 4nm TSMC process, while the Xeon uses a 5nm Intel process. Cache per core: the Ryzen has 80 KB L1 and 1 MB L2; the Xeon has 112 KB L1 and 2 MB L2. Total L3 is 128 MB for the Ryzen and 144 MB for the Xeon.

Memory support: both use DDR5, but the Ryzen is dual-channel with 89.6 GB/s bandwidth, while the Xeon is eight-channel with 409.6 GB/s. PCIe: the Ryzen has 24 Gen 5 lanes; the Xeon has 128 Gen 5 lanes. The Ryzen includes Radeon Graphics; the Xeon has no integrated graphics. The Xeon is multiplier-unlocked; the Ryzen is locked. The Ryzen’s die size is 2x 70.6 mm²; the Xeon’s is 2x 598 mm². The Ryzen has 16,630 million transistors; the Xeon’s transistor count is not listed. Release dates are 2026-06-29 for AMD and 2026-02-01 for Intel.

The Verdict

The data supports a clear split: pick the Intel Xeon 674X for throughput, and pick the AMD Ryzen 9 PRO 9965X3D for single-thread responsiveness. The Xeon wins 9 of 11 head-to-head benchmarks, with margins ranging from 13.6% to 36.7%. Its 28 cores, 56 threads, eight-channel memory bandwidth, and 128 PCIe lanes make it the superior choice for rendering, simulation, data compression, encryption, and any workload that scales with cores and memory bandwidth. The Ryzen’s 17.3% single-thread lead and lower TDP (170W vs 270W) make it attractive for interactive work, legacy software, or environments where power is a constraint.

The Xeon’s average benchmark score of 143103 is within 0.4% of the Ryzen’s 143735, but that near-tie masks the skew: the Xeon’s wins are broad and deep, while the Ryzen’s wins are narrow. If your application is single-threaded, the Ryzen is the only rational choice. If your workload is multi-threaded, the Xeon is the decisive winner, and its 409.6 GB/s memory bandwidth versus 89.6 GB/s is a structural advantage that no clock speed can overcome. The Xeon’s launch MSRP is $2199, which you may consider as a reference point, but the Ryzen’s price is not listed. Ultimately, the Ryzen 9 PRO 9965X3D is a workstation part that excels where cores are less important than speed; the Xeon 674X is a server-class workhorse that punishes any workload that leaves cores idle. Choose accordingly.