AMD EPYC 9275F vs AMD Ryzen 9 PRO 9965 Comparison

The AMD Ryzen 9 PRO 9965 and AMD EPYC 9275F are both Zen 5 parts, but they target very different corners of the server and workstation market. The data shows a clear split: the EPYC 9275F dominates nearly every multithreaded and throughput-oriented benchmark, while the Ryzen 9 PRO 9965 takes a decisive lead in single-thread performance. The benchmark results show the EPYC 9275F wins 9 of the 11 head-to-head comparisons, with the Ryzen 9 PRO 9965 taking the remaining 2.

Head-to-Head Benchmarks

The most lopsided result is in the PassMark physics test, where the EPYC 9275F scores 12,089 versus the Ryzen 9 PRO 9965's 3,256. That is a 73.1% advantage for the EPYC part — by far the largest gap in the entire comparison. The find prime numbers test tells a similar story, with the EPYC scoring 991 against the Ryzen's 364, a 63.3% deficit for the smaller chip. These two results indicate that the EPYC's additional cores and threads translate directly into raw computational throughput in integer-heavy workloads.

The gap narrows considerably in other tests, but the EPYC still holds a commanding lead. In data compression, the EPYC scores 1,212,560 versus 908,293, a 25.1% edge. Data encryption shows a 28.3% advantage (62,664 vs 44,920), and extended instructions come in at 25% (94,889 vs 71,210). Integer math is 23.4% higher on the EPYC (317,777 vs 243,280), while floating point math shows a 20.4% lead (201,888 vs 160,746). Random string sorting favors the EPYC by 34.1% (144,037 vs 94,915), and the multithread score is 21.2% higher (84,620 vs 66,655). Every one of these results follows the same pattern: the 24-core EPYC outmuscles the 16-core Ryzen, but the margins are not uniform, ranging from roughly 20% to over 70%.

The Ryzen 9 PRO 9965's only wins come in the single-thread benchmarks, where it scores 4,682 versus the EPYC's 3,810. That is a 22.9% advantage — a substantial margin that suggests the Ryzen's higher boost clock of 5.50 GHz (versus 4.80 GHz on the EPYC) plays a significant role. This is consistent with the fact that the Ryzen is a dual-chiplet part (2x 70.6 mm² die size) while the EPYC is an eight-chiplet design (8x 70.6 mm²), meaning the EPYC has more cores but each core may be operating under different thermal and power constraints.

Architecture Differences

Both processors are built on TSMC's 4 nm process node and use the Zen 5 architecture, but they diverge significantly in almost every other physical characteristic. The Ryzen 9 PRO 9965 is codenamed Granite Ridge and belongs to the Ryzen 9 generation, while the EPYC 9275F is codenamed Turin and is part of the EPYC 9005 series. The Ryzen uses AMD Socket AM5, whereas the EPYC requires AMD Socket SP5 — these are not interchangeable platforms.

Core counts are a major differentiator: the Ryzen has 16 cores and 32 threads, while the EPYC has 24 cores and 48 threads. The EPYC also carries far more cache — 256 MB of shared L3 versus 64 MB on the Ryzen — though both have 80 KB of L1 and 1 MB of L2 per core. Transistor counts reflect the scale difference: the EPYC has 66,520 million transistors versus 16,630 million on the Ryzen, and the EPYC's die is eight chiplets versus two on the Ryzen.

Memory subsystems are another major split. The Ryzen supports dual-channel DDR5 with 89.6 GB/s of bandwidth, while the EPYC supports twelve-channel DDR5 with 576.0 GB/s — a 6.4x difference in theoretical memory bandwidth. Both support ECC memory, but the EPYC's platform is clearly built for memory-intensive server workloads. PCIe lanes also differ dramatically: the Ryzen provides Gen 5 with 24 lanes (CPU only), while the EPYC provides Gen 5 with 128 lanes (CPU only). The Ryzen includes integrated Radeon Graphics; the EPYC has none. The Ryzen's TDP is 170 W, and the EPYC's is 320 W.

FAQ

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

A: The AMD Ryzen 9 PRO 9965 wins single-thread performance with a PassMark score of 4,682, which is 22.9% higher than the AMD EPYC 9275F's score of 3,810.

Q: How do the two compare in multithreaded performance?

A: The EPYC 9275F leads with a PassMark multithread score of 84,620 versus 66,655 for the Ryzen 9 PRO 9965, a 21.2% advantage. The EPYC also wins every other multithreaded workload in the data, including physics, data compression, and encryption.

Q: What is the difference in memory bandwidth?

A: The EPYC 9275F has 576.0 GB/s of memory bandwidth across a twelve-channel DDR5 bus, while the Ryzen 9 PRO 9965 has 89.6 GB/s across a dual-channel DDR5 bus.

Q: Do both CPUs use the same manufacturing process?

A: Yes, both are built on TSMC's 4 nm process node and use the Zen 5 architecture, but the Ryzen 9 PRO 9965 is codenamed Granite Ridge while the EPYC 9275F is codenamed Turin.

Q: How many PCIe lanes does each CPU provide?

A: The Ryzen 9 PRO 9965 provides 24 Gen 5 lanes (CPU only), while the EPYC 9275F provides 128 Gen 5 lanes (CPU only).

Q: Which CPU has more L3 cache?

A: The EPYC 9275F has 256 MB of shared L3 cache, whereas the Ryzen 9 PRO 9965 has 64 MB of L3 cache.

The Verdict

The data is unambiguous about workload distribution. The AMD EPYC 9275F is the superior processor for heavily multithreaded, memory-bandwidth-bound, and I/O-heavy server workloads. Its 24 cores and 48 threads, combined with 256 MB of L3 cache and 576.0 GB/s of memory bandwidth, yield decisive wins in every throughput benchmark — from physics simulation (73.1% ahead) to data compression (25.1% ahead) to integer math (23.4% ahead). The EPYC also sits in the 97th percentile of all CPUs, with an average benchmark score of 133,174, and its nearest rival, the Intel Xeon 6710E, trails by 2.5%.

The AMD Ryzen 9 PRO 9965 is the pick when single-thread performance matters most. Its 22.9% lead in single-thread score over the EPYC is substantial, and it achieves this while drawing 150 W less TDP (170 W vs 320 W). The Ryzen also has an integrated GPU, which the EPYC lacks, and it fits on the AM5 platform, which is a more accessible socket. The Ryzen sits in the 98th percentile of all CPUs, slightly higher than the EPYC's 97th, but its average benchmark score of 145,728 is actually higher than the EPYC's 133,174 — a reflection of how the single-thread win and strong multithread results balance out across all tests. Its nearest rival, the AMD Ryzen Threadripper PRO 9965WX, is only 0.9% ahead, while the Intel Xeon w9-3575X trails by 1%.

Specification Differences

• Cores: 16 (Ryzen 9 PRO 9965) vs 24 (EPYC 9275F)
• Threads: 32 vs 48
• Base Clock: 4.30 GHz vs 4.10 GHz
• Boost Clock: 5.50 GHz vs 4.80 GHz
• TDP: 170 W vs 320 W
• Socket: AMD Socket AM5 vs AMD Socket SP5
• Codename: Granite Ridge vs Turin
• Transistors: 16,630 million vs 66,520 million
• Die Size: 2x 70.6 mm² vs 8x 70.6 mm²
• L3 Cache: 64 MB vs 256 MB (shared)
• Memory Bus: Dual-channel vs Twelve-channel
• Memory Bandwidth: 89.6 GB/s vs 576.0 GB/s
• PCIe Lanes: 24 Gen 5 (CPU only) vs 128 Gen 5 (CPU only)
• Integrated Graphics: Radeon Graphics vs N/A
• Release Date: 2026-06-29 vs 2024-10-09
• Launch MSRP: None vs $3439

Where Each One Wins

The AMD EPYC 9275F wins in every scenario that scales with core count, memory bandwidth, or cache capacity. The data shows it is 63.3% faster in find prime numbers and 73.1% faster in physics — both tests that reward raw parallel compute. It is also the choice for data compression (25.1% ahead), encryption (28.3% ahead), and random string sorting (34.1% ahead). If the workload involves large datasets that benefit from 576.0 GB/s of memory bandwidth or 128 PCIe lanes for peripheral connectivity, the EPYC is the only option between these two. Its 24 cores and 48 threads provide a 50% core advantage over the Ryzen.

The AMD Ryzen 9 PRO 9965 wins where single-thread responsiveness dominates, which the 22.9% lead in PassMark single-thread confirms. It also offers integrated Radeon Graphics, eliminating the need for a discrete GPU in basic display output scenarios. The lower 170 W TDP makes it easier to cool and power in a workstation chassis, and the AM5 socket platform is more conventional than the SP5 server platform. For workloads that are lightly threaded but latency-sensitive — such as certain database queries, legacy applications, or interactive development tasks — the Ryzen's higher boost clock of 5.50 GHz provides a measurable edge. The Ryzen also holds a higher percentile rank (98 vs 97) and a higher average benchmark score (145,728 vs 133,174), meaning that in mixed workloads that combine single-thread and multithread tasks, the Ryzen can come out ahead overall.