AMD EPYC 9555P vs AMD EPYC 9655 Comparison

Head-to-Head Benchmarks

The comparison between the AMD EPYC 9655 and the AMD EPYC 9555P is unusually lopsided. Across the seventeen benchmark tests recorded in the database, the EPYC 9655 claims victory in every single one. The data shows a 100% win rate for the 96-core part, with the 64-core EPYC 9555P failing to take a single test. This is not a narrow advantage; the margins are substantial and consistent across both synthetic rendering workloads and real-world compute tasks.

Starting with the Cinebench suite, the EPYC 9655 posts a 15.2% lead in every iteration, from R15 through R23, in both single-core and multi-core tests. In Cinebench R23 multi-core, the 9655 scores 132672 against 115186 for the 9555P. The single-core results follow the same pattern: 18730 versus 16261, a 15.2% delta. This consistency across the entire Cinebench family indicates that the advantage is not workload-specific but rather a fundamental performance differential.

The Passmark results paint an even more dramatic picture. The largest single margin appears in Passmark Physics, where the EPYC 9655 delivers 25947 versus 15474, a staggering 67.7% advantage. Random string sorting shows a 56.8% gap (439682 vs 280398), while find prime numbers comes in at 49.8% (1598 vs 1067). Integer math demonstrates a 44.7% lead (1139161 vs 787106), and data encryption shows a 41.4% advantage (210555 vs 148896). Floating point math follows at 36.3% (662958 vs 486407), with multithread scoring 26.3% higher (156110 vs 123576). Data compression shows a 24% gap (3271896 vs 2639400), and extended instructions the narrowest Passmark margin at 6.4% (203285 vs 191082). The single-thread Passmark tests record a 12.8% advantage for the 9655 (3847 vs 3410).

The average benchmark score in the database reinforces this hierarchy. The EPYC 9655 averages 373479, while the EPYC 9555P averages 287066. The 9655 sits at the 100th percentile among all CPUs, whereas the 9555P ranks at the 99th percentile. Both are elite parts, but the 9655 occupies the absolute top tier.

Where Each One Wins

The EPYC 9655 wins everywhere, but the magnitude of its victories varies in ways that suggest different use-case strengths. The massive 67.7% lead in physics simulations and the 56.8% edge in random string sorting point to workloads that scale aggressively with core count and memory bandwidth. The 49.8% gap in prime number finding, a purely computational task, similarly rewards the 96-core configuration.

For memory-intensive operations, the 9655's advantage is pronounced. Data compression, which benefits from large caches and high throughput, shows a 24% lead. Encryption, which often scales with core count, shows a 41.4% gap. For integer and floating point math, the 9655 leads by 44.7% and 36.3% respectively, indicating that CPU-bound number crunching sees substantial gains from the additional 32 cores.

The single-core results are narrower but still favor the 9655. A 15.2% lead in Cinebench single-core and 12.8% in Passmark single-thread suggest that the 9655's 4.50 GHz boost clock, as opposed to the 9555P's 4.40 GHz, contributes to its advantage in lightly threaded tasks. However, the gap is far smaller than in multi-core tests, confirming that the core count difference is the primary driver of the performance separation.

There is no workload category in the recorded data where the EPYC 9555P comes out ahead. The 9555P does have a higher base clock at 3.20 GHz versus 2.60 GHz, but in practice this does not translate into any benchmark win. The 9655's combination of more cores, a higher boost clock, and larger L3 cache (384 MB vs 256 MB) proves decisive across the board.

The Verdict

The data is unambiguous: the AMD EPYC 9655 is the superior processor in every measured category. For workloads that demand maximum throughput, whether in multi-threaded rendering, data compression, encryption, or mathematical computation, the 9655 delivers a consistent and often massive advantage. The 67.7% lead in physics, the 56.8% in string sorting, and the 49.8% in prime number finding are the kinds of margins that translate directly into reduced job completion times.

Buyers who prioritize single-thread responsiveness will still find the 9655 ahead, though by a smaller margin. The 12.8% Passmark single-thread lead and 15.2% Cinebench single-core lead mean that even lightly threaded applications run faster on the 9655. There is no scenario in the recorded benchmarks where the 9555P is the recommended choice based on performance alone.

The EPYC 9555P remains a strong processor in its own right. It sits at the 99th percentile of all CPUs and outperforms its nearest rivals in the database, including the Intel Xeon 696X by 0.3% and the AMD EPYC 9565 by 0.6%. For users with a 64-core ceiling or those constrained by the lower TDP of 360 watts versus 400 watts, the 9555P is a capable alternative. But the benchmark data shows that the 9655 is the outright winner for anyone who can accommodate its higher core count and power envelope.

FAQ

Q: Which processor wins in Cinebench R23 multi-core?

A: The AMD EPYC 9655 scores 132672, while the AMD EPYC 9555P scores 115186. The 9655 leads by 15.2%.

Q: What is the largest performance gap between the two CPUs?

A: The largest margin is in Passmark Physics, where the EPYC 9655 scores 25947 versus 15474 for the EPYC 9555P, a 67.7% difference.

Q: Does the EPYC 9555P win any benchmark in the database?

A: No. The EPYC 9655 wins all 17 recorded head-to-head benchmark tests, with the EPYC 9555P recording zero wins.

Q: How do the two processors compare in single-thread performance?

A: The EPYC 9655 leads by 15.2% in Cinebench R23 single-core (18730 vs 16261) and by 12.8% in Passmark single-thread (3847 vs 3410).

Q: What are the core and thread counts for each processor?

A: The AMD EPYC 9655 has 96 cores and 192 threads. The AMD EPYC 9555P has 64 cores and 128 threads.

Q: How does the EPYC 9555P compare to its nearest rival?

A: The EPYC 9555P has an average score of 287066, which is 0.3% ahead of the Intel Xeon 696X (286102) and 0.6% ahead of the AMD EPYC 9565 (285471).

Architecture Differences

Both processors belong to the AMD EPYC 9005 series, share the Zen 5 architecture, and carry the Turin codename. Both are built on TSMC's 4 nm process node, use the AMD Socket SP5, and support DDR5 memory across a twelve-channel bus with 576.0 GB/s of bandwidth. Both support ECC memory and provide PCIe Gen 5 with 128 lanes (CPU only). Neither has integrated graphics, and both have a locked multiplier.

The core differences are structural. The EPYC 9655 packs 96 cores and 192 threads, while the EPYC 9555P offers 64 cores and 128 threads. This 32-core gap is the primary driver of the performance differences seen in the benchmarks. The 9655's base clock is 2.60 GHz with a boost of 4.50 GHz, whereas the 9555P runs at a higher base of 3.20 GHz but a lower boost of 4.40 GHz. The higher boost clock on the 9655 explains its narrow but consistent single-thread wins.

Cache allocation differs substantially. The 9655 has an L3 cache of 384 MB shared, while the 9555P has 256 MB shared. Both have the same per-core L1 (80 KB) and L2 (1 MB) cache. The transistor count reflects the core disparity: the 9655 contains 99,780 million transistors across 12 dies of 70.6 mm² each, while the 9555P uses 66,520 million transistors across 8 dies of the same size. The TDP also scales accordingly, with the 9655 rated at 400 watts and the 9555P at 360 watts.

The launch MSRP for the EPYC 9655 is $11852, and for the EPYC 9555P it is $7983. Both processors were released on 2024-10-09 and are currently in active production. The part numbers are 100-000000674 for the 9655 and 100-000001523 for the 9555P.