AMD EPYC 9535 vs AMD EPYC 9555P Comparison

Head-to-Head Benchmarks

The head-to-head data shows a clear overall winner in the AMD EPYC 9555P, which takes 9 of the 11 recorded benchmark comparisons. The largest margin comes in the PassMark physics test, where the EPYC 9555P scores 15,474 against the EPYC 9535's 3,834, a delta of 75.2% in favor of the 9555P. This is an enormous gap, far exceeding the single-digit or low-double-digit differences seen in most other workloads, and it suggests the 9555P has a substantial advantage in simulation or physics-based processing.

In data compression, the EPYC 9555P records 2,639,400 versus 2,308,822 for the EPYC 9535, a 12.5% lead. Data encryption also favors the 9555P, with 148,896 against 127,372, a 14.5% advantage. The prime number search workload shows a notable 18.1% gap, with the 9555P scoring 1,067 and the 9535 scoring 874. Random string sorting goes to the 9555P by 11.7%, with scores of 280,398 and 247,506 respectively. The extended instructions test gives the 9555P an 8% edge, 191,082 to 175,784. Floating point math and integer math both favor the 9555P by 6% and 7.2% respectively, with scores of 486,407 versus 457,047 and 787,106 versus 730,281. The multithreaded PassMark score shows a 7.3% advantage for the 9555P, 123,576 versus 114,528.

The EPYC 9535 wins only two benchmarks, both of which are the single-thread PassMark tests (listed twice in the dataset with identical scores). The 9535 records a single-thread score of 3,720 versus 3,410 for the 9555P, a 9.1% advantage. This is the only category where the 9535 leads by a meaningful margin, and it indicates that while the 9555P dominates heavily threaded and parallel workloads, the 9535 has an edge in lightly threaded or latency-sensitive single-core tasks.

Looking at the average benchmark scores, the EPYC 9535 posts an average of 379,408 across its recorded benchmarks, while the EPYC 9555P averages 287,066. However, this comparison is skewed by the fact that the 9555P includes Cinebench R15, R20, and R23 scores in its dataset, which are not present for the 9535, so the averages are not directly comparable across the full suite. The head-to-head PassMark results are the more reliable apples-to-apples comparison, and they consistently favor the 9555P in multi-threaded work.

In terms of nearest rivals, the EPYC 9535 sits at the 100th percentile of all CPUs, with its closest competitor being the AMD EPYC 9655 at a 1.6% delta, followed by the Intel Xeon 6960P at 3.9%, the AMD EPYC 9754 at 4.1%, and the AMD EPYC 9655P at -4.6% (meaning the 9655P scores higher). The EPYC 9555P ranks at the 99th percentile, with the Intel Xeon 696X nearly matching it at a 0.3% delta, then the AMD EPYC 9565 at 0.6%, the Intel Xeon 6780E at 2.4%, and the AMD Ryzen Threadripper 9970X at 2.6%. Both parts are clearly at the top of the CPU hierarchy, but the 9535 edges out the 9555P in overall percentile ranking.

Where Each One Wins

The EPYC 9555P is the clear choice for heavily parallel, throughput-oriented workloads. Every multi-threaded PassMark test in the head-to-head comparison goes to the 9555P, including data compression, encryption, prime number search, floating point math, integer math, multithread, physics, and random string sorting. The physics test is the standout, with a 75.2% advantage, making the 9555P particularly well-suited for scientific simulation, finite element analysis, or any workload that relies on rigid body or particle physics calculations. The encryption and compression results also show strong double-digit leads, which points to advantages in database encryption, backup compression, and data pipeline workloads.

The EPYC 9535 wins only the single-thread PassMark test, with a 9.1% lead. This makes it the better option for workloads that are bound by single-core performance, such as legacy applications that are not well-parallelized, certain database queries that run on a single thread, or interactive server sessions where response time on a single core matters more than aggregate throughput. The 9535 also holds a higher percentile rank at 100 versus 99 for the 9555P, which reflects its position relative to all CPUs in the database, though this is a broader ranking metric rather than a direct head-to-head result.

For mixed workloads, the 9555P's consistent wins across the majority of tests make it the more versatile performer. Its smallest multi-threaded advantage is 6% in floating point math, which is still a meaningful lead, and its 7.2% to 7.3% advantages in integer math and multithread scores show balanced gains across general-purpose compute. The 9535's single-thread win is significant but narrow in scope: it only applies to the one benchmark category, and even there the margin is under 10%.

Architecture Differences

Both the AMD EPYC 9535 and the AMD EPYC 9555P are built on the Zen 5 architecture, with the Turin codename, and belong to the EPYC 9005 series. They share the same 4 nm process node from TSMC, the same transistor count of 66,520 million, and the same die size of 8x 70.6 mm². The socket is identical as well: AMD Socket SP5. Neither part has an unlocked multiplier, and both lack integrated graphics, as expected for server/workstation parts.

The core and cache configurations are also identical. Both processors have 64 cores and 128 threads, with an L1 cache of 80 KB per core, an L2 cache of 1 MB per core, and a shared L3 cache of 256 MB. There is no 3D V-Cache on either part. Memory support is the same: DDR5 with a twelve-channel bus and a memory bandwidth of 576.0 GB/s, with ECC memory support enabled. PCIe connectivity is also identical at Gen 5 with 128 lanes from the CPU.

The key architectural differences lie in clock speeds and power envelope. The EPYC 9535 has a base clock of 2.40 GHz and a boost clock of 4.30 GHz, with a TDP of 300 W. The EPYC 9555P has a higher base clock of 3.20 GHz and a boost clock of 4.40 GHz, with a TDP of 360 W. The 9555P's higher base clock, a 0.80 GHz increase, directly explains its dominance in multi-threaded workloads, as sustained all-core operation runs at base clock. The higher boost clock, 0.10 GHz over the 9535, also contributes to the 9555P's single-thread performance, though the 9535 manages to win that specific test despite the clock deficit, which suggests the 9535's boost behavior may be more aggressive or its thermal headroom allows for higher sustained single-core frequencies.

The 300 W versus 360 W TDP difference is a direct consequence of the clock speed disparity. The 9555P consumes more power to achieve its higher clocks, while the 9535 trades clock speed for a lower thermal envelope. This is the primary architectural trade-off between the two parts: identical silicon, different binning and power limits.

Specification Differences

The two processors differ in only a few recorded specifications. The base clock is 2.40 GHz for the EPYC 9535 and 3.20 GHz for the EPYC 9555P. The boost clock is 4.30 GHz for the 9535 and 4.40 GHz for the 9555P. The TDP is 300 W for the 9535 and 360 W for the 9555P. The launch MSRP is $8992 for the 9535 and $7983 for the 9555P. The part numbers differ as well: 100-000001147 for the 9535 and 100-000001523 for the 9555P.

All other specifications are identical: 64 cores, 128 threads, same cache hierarchy, same memory support, same PCIe configuration, same socket, same architecture, same process node, same foundry, same transistor count, same die size, and same release date of 2024-10-09. Both are active production parts in the server/workstation market segment. The benchmark scores differ, as detailed in the head-to-head section, and the percentile rankings differ by one point (100 for the 9535, 99 for the 9555P). The average benchmark scores also differ, though the dataset composition varies between the two parts.

FAQ

Q: Which processor has the higher base clock speed?

A: The AMD EPYC 9555P has a base clock of 3.20 GHz, while the AMD EPYC 9535 has a base clock of 2.40 GHz. The 9555P also has a slightly higher boost clock at 4.40 GHz versus 4.30 GHz.

Q: How much larger is the EPYC 9555P's win in the physics benchmark?

A: The EPYC 9555P scores 15,474 in the PassMark physics test, compared to 3,834 for the EPYC 9535, a delta of 75.2% in favor of the 9555P. This is the largest margin of any benchmark in the head-to-head comparison.

Q: Does the EPYC 9535 win any benchmarks against the EPYC 9555P?

A: Yes, the EPYC 9535 wins the PassMark single-thread test with a score of 3,720 versus 3,410 for the 9555P, a 9.1% advantage. This result appears twice in the dataset, but it is the only category where the 9535 leads.

Q: What is the TDP difference between the two processors?

A: The EPYC 9535 has a TDP of 300 W, while the EPYC 9555P has a TDP of 360 W. The higher TDP on the 9555P corresponds to its higher base and boost clock speeds.

Q: Are the cache configurations the same on both parts?

A: Yes, both the EPYC 9535 and the EPYC 9555P have 80 KB of L1 cache per core, 1 MB of L2 cache per core, and 256 MB of shared L3 cache. Neither part has 3D V-Cache.

Q: How do the two processors compare in their nearest rival rankings?

A: The EPYC 9535 is at the 100th percentile of all CPUs, with its closest rival being the AMD EPYC 9655 at a 1.6% delta. The EPYC 9555P is at the 99th percentile, with the Intel Xeon 696X nearly matching it at a 0.3% delta. The 9535 has a higher percentile ranking, but the 9555P wins nearly all direct head-to-head comparisons.