AMD EPYC 9565 vs AMD EPYC 9755 Comparison

Where Each One Wins

The benchmark data splits this comparison into two very different personalities. The AMD EPYC 9755 wins 14 of the 17 recorded head-to-head tests, establishing itself as the clear multi-threaded workhorse. Its victories are not marginal: in integer math it leads by 115.9%, in data encryption by 100.7%, and in random string sorting by 95.7%. These are transformative margins, not incremental gains.

The AMD EPYC 9565, by contrast, wins exactly three tests, and each win reveals a different strength. It takes the single-thread PassMark test by 5.2%, and the find prime numbers test by 15.5%. These wins point to a processor that extracts more from each individual core despite its lower core count. The 9565 also carries a higher base clock at 3.15 GHz versus 2.70 GHz, and a higher boost clock at 4.30 GHz versus 4.10 GHz, which explains its single-thread advantage.

For workloads that scale across cores, the 9755 is the answer. The Cinebench suite shows a uniform 23% advantage across R15, R20, and R23, both single and multi-core. The 9755 also dominates in floating-point math by 68%, physics by 54.2%, and data compression by 75.1%. The pattern is consistent: the 9755 converts its 128 cores and 256 threads into overwhelming throughput.

The 9565 is not a weak processor; it sits at the 99th percentile of all CPUs in the database. Its average benchmark score of 285,471 is credible. But the 9755 sits at the 100th percentile with an average score of 505,778, which is 77% higher. The 9565 wins only where clock speed and per-core efficiency matter more than raw core count.

Architecture Differences

Both processors share the same fundamental design: Zen 5 architecture, Turin codename, 4 nm process node from TSMC, and the AMD Socket SP5. They belong to the same EPYC 9005 series, released on the same date. The differences emerge in the physical implementation.

The 9755 packs 128 cores and 256 threads, built from 16 chiplets of 70.6 mm² each, totaling 133,040 million transistors. The 9565 uses 72 cores and 144 threads, assembled from 12 chiplets of the same 70.6 mm² size, with 99,780 million transistors. The 9755 therefore uses 33% more chiplets and 33% more transistors, which directly explains its larger L3 cache.

L3 cache scales with the chiplet count. The 9755 has 512 MB of shared L3 cache, while the 9565 has 384 MB. Both share the same per-core L1 and L2 allocations: 80 KB and 1 MB per core respectively. The cache difference is purely a function of how many chiplets each processor integrates.

Memory support is identical: DDR5, twelve-channel, 576.0 GB/s of memory bandwidth, with ECC support on both. PCIe is also the same: Gen 5 with 128 lanes (CPU only). Neither has integrated graphics. Both are server and workstation parts with active production status.

The TDP differs by 100 watts: 500 for the 9755, 400 for the 9565. This is consistent with the core count difference, but the 9565 achieves its lower TDP while still running higher clocks, which indicates a more efficient per-core power profile.

The Verdict

The data directs different buyers to different processors. If the workload is dense multi-threading, the 9755 is the only rational choice. Its 115.9% lead in integer math and 100.7% lead in data encryption are not merely statistical; they represent near-doubling of throughput in those specific tasks. For database workloads, scientific computing, large-scale virtualization, or any environment where every additional core translates into more finished work per hour, the 9755 dominates.

The 9565 makes sense for workloads that are latency-sensitive or single-thread-bound. Its 5.2% lead in PassMark single-thread and 15.5% lead in prime number finding indicate that it extracts more performance from each core. The higher base clock of 3.15 GHz versus 2.70 GHz is the key differentiator. For applications that cannot use more than a few dozen cores but demand the fastest possible response from each one, the 9565 is the better fit.

The database also shows percentile placement: the 9755 is at the 100th percentile, the 9565 at the 99th. Both are elite parts, but the 9755 is at the absolute ceiling. Its nearest rival, the AMD EPYC 9845, beats it by 3.4%, while the 9755 leads the EPYC 9745 by 18.7% and the Threadripper PRO 9995WX by 22.7%. The 9565 trades blows with its nearest rivals: it trails the Intel Xeon 696X by 0.2%, trails the EPYC 9555P by 0.6%, and leads the Xeon 6780E by 1.8% and the Threadripper 9970X by 2%.

Neither processor is a wrong choice for its intended role. The 9755 is the throughput king; the 9565 is the per-core specialist.

FAQ

Q: Which processor has more cores?

A: The AMD EPYC 9755 has 128 cores and 256 threads, while the AMD EPYC 9565 has 72 cores and 144 threads.

Q: What is the single-thread performance difference?

A: The 9565 wins the PassMark single-thread test by 5.2% (3696 versus 3503). However, the 9755 wins all Cinebench single-core tests by 23%.

Q: How do they compare in memory bandwidth?

A: Both processors offer identical memory support: DDR5, twelve-channel, with 576.0 GB/s of bandwidth and ECC support.

Q: What is the L3 cache difference?

A: The 9755 has 512 MB of shared L3 cache, while the 9565 has 384 MB. Both share the same per-core L1 (80 KB) and L2 (1 MB) cache.

Q: Which processor has the higher boost clock?

A: The 9565 boosts to 4.30 GHz, while the 9755 boosts to 4.10 GHz. The 9565 also has a higher base clock at 3.15 GHz versus 2.70 GHz.

Q: What is the launch MSRP for each?

A: The 9755 has a launch MSRP of $12984, and the 9565 has a launch MSRP of $10486.

Head-to-Head Benchmarks

The Cinebench results are remarkably consistent. Across R15, R20, and R23, the 9755 wins both single and multi-core tests by exactly 23%. The multi-core scores are 14,250 versus 11,585 (R15), 59,378 versus 48,273 (R20), and 141,378 versus 114,937 (R23). The single-core scores follow the same pattern: 2,011 versus 1,635 (R15), 8,382 versus 6,814 (R20), and 19,959 versus 16,226 (R23). This uniformity suggests the 9755's advantage is structural, rooted in core count rather than any workload-specific behavior.

The PassMark suite reveals where the 9755 truly separates itself. Integer math shows a 115.9% lead (1,549,946 versus 717,948). Data encryption nearly doubles the 9565's output, winning by 100.7% (284,927 versus 141,936). Random string sorting is close behind at 95.7% (571,185 versus 291,941). Data compression follows at 75.1% (4,517,407 versus 2,579,631). Floating-point math wins by 68% (922,900 versus 549,422), and physics by 54.2% (27,806 versus 18,036).

The 9565's wins are narrower but real. In find prime numbers, it scores 2,422 versus 2,047, a 15.5% lead. In PassMark single-thread, it scores 3,696 versus 3,503, a 5.2% lead. The extended instructions test goes to the 9755 by 44.7% (303,321 versus 209,595), and the multithread test by 23% (166,328 versus 135,221).

The overall win count is 14 to 3 in favor of the 9755. The three losses for the 9755 are all in tests that reward per-core efficiency over raw throughput.

Specification Differences

| Specification | AMD EPYC 9755 | AMD EPYC 9565 |

|---|---|---|

| Cores | 128 | 72 |

| Threads | 256 | 144 |

| Base Clock | 2.70 GHz | 3.15 GHz |

| Boost Clock | 4.10 GHz | 4.30 GHz |

| TDP | 500 W | 400 W |

| Transistors | 133,040 million | 99,780 million |

| Die Size | 16x 70.6 mm² | 12x 70.6 mm² |

| L3 Cache | 512 MB (shared) | 384 MB (shared) |

| Launch MSRP | $12984 | $10486 |

Both processors share the same socket (AMD Socket SP5), architecture (Zen 5), codename (Turin), process node (4 nm), foundry (TSMC), memory support (DDR5, twelve-channel, 576.0 GB/s), PCIe configuration (Gen 5, 128 lanes), ECC support, and market segment (Server/Workstation). They also share the same release date and production status. The differences are entirely in core count, clock speeds, power envelope, transistor count, die configuration, L3 cache, and launch MSRP.