AMD EPYC 9655 vs Intel Xeon 6960P Comparison

The AMD EPYC 9655 and Intel Xeon 6960P are flagship server processors, both commanding the top percentile of all CPUs in the database. The data reveals a consistent and often substantial performance advantage for the AMD part across every benchmark recorded, making this a surprisingly one-sided comparison in raw compute throughput.

Head-to-Head Benchmarks

The EPYC 9655 sweeps the board, winning all 14 head-to-head comparisons. The most dramatic victory comes in PassMark's integer math test, where the AMD chip scores 1,139,161 against the Intel's 727,750, a staggering 56.5% lead. This suggests a profound advantage in general-purpose arithmetic workloads that rely heavily on integer operations.

The margin narrows but remains decisive in floating-point math. Here, the EPYC 9655 posts 662,958, which is 25.7% ahead of the Xeon 6960P's 527,473. This is a significant gap for scientific and engineering simulations that depend on floating-point throughput.

Across the Cinebench suite, the pattern is remarkably consistent. In the R15, R20, and R23 multi-core tests, the AMD processor holds a uniform 19.5% advantage. The scores are 13,373 vs 11,194 in R15, 55,722 vs 46,645 in R20, and 132,672 vs 111,060 in R23. This suggests a scalable performance lead that persists regardless of the rendering workload's complexity.

Data-centric tasks also favor the AMD chip. In data compression, the EPYC 9655 scores 3,271,896, which is 16.9% higher than the Xeon's 2,797,724. The encryption test shows an even larger 30% gap, with the AMD part scoring 210,555 versus the Intel's 162,013. This could indicate a notable edge in security and database workloads.

Even in tests where the margin is smaller, the AMD part still comes out on top. In extended instructions, the EPYC 9655 leads by a modest 5.1% (203,285 vs 193,404). The physics test shows a 4.1% advantage (25,947 vs 24,937), and prime number finding shows a 7.7% lead (1,598 vs 1,484). The AMD processor also wins the multithreaded test by 19.5% (156,110 vs 130,659) and random string sorting by 18.3% (439,682 vs 371,795).

Single-threaded performance is another clear win for AMD. The EPYC 9655 scores 3,847 in both the PassMark single-thread and singlethread tests, while the Xeon 6960P trails at 3,287, a 17% deficit. This is a critical finding, as it means the AMD chip is not just faster when all cores are busy, but also when just one core is active.

Architecture Differences

The two processors are built on fundamentally different designs. The AMD EPYC 9655 is part of the EPYC 9005 series, codenamed Turin, and uses the Zen 5 architecture on a 4nm process from TSMC. It packs 96 cores and 192 threads, running at a base clock of 2.60 GHz and a boost clock of 4.50 GHz. The chip is composed of 12 dies, each measuring 70.6 mm², totaling 99,780 million transistors.

The Intel Xeon 6960P belongs to the Xeon 6 generation (Granite Rapids-AP) and is built on Intel's own 5nm process. It offers 72 cores and 144 threads, which explains much of the performance deficit in multi-threaded tests. Its clock speeds are lower as well, with a base of 2.70 GHz and a boost of 3.90 GHz. The Intel die is significantly larger at 598 mm², and there are three of them.

Cache configurations also differ substantially. The AMD EPYC 9655 provides 80 KB of L1 cache and 1 MB of L2 cache per core, plus a massive 384 MB of shared L3 cache. The Intel Xeon 6960P offers more per-core cache, with 112 KB of L1 and 2 MB of L2, and a larger L3 pool of 432 MB. Despite having fewer cores, the Intel chip's larger L3 could benefit certain cache-sensitive workloads, though the benchmark data suggests this does not overcome the core count disadvantage.

Memory support is comparable, with both using DDR5 and a twelve-channel memory bus. However, the Intel Xeon 6960P has a higher theoretical memory bandwidth of 614.4 GB/s compared to the AMD's 576.0 GB/s. Both support ECC memory. PCIe connectivity differs, with the AMD offering 128 Gen 5 lanes versus the Intel's 96 Gen 5 lanes.

The power envelopes are also distinct. The AMD EPYC 9655 has a TDP of 400W, while the Intel Xeon 6960P is rated higher at 500W. This suggests the Intel chip may require more robust cooling, though the performance results indicate it does not translate into a performance win. The AMD part uses the AMD Socket SP5, while the Intel uses the Intel Socket 7529.

Where Each One Wins

The benchmark data points to the AMD EPYC 9655 as the superior choice for nearly all compute-intensive scenarios. The 56.5% lead in integer math makes it the clear pick for database operations, financial modeling, and any workload that processes large volumes of discrete data. The 30% advantage in encryption further solidifies its position for security-focused tasks like VPN termination or secure data storage.

For rendering and content creation, the consistent 19.5% lead across all three Cinebench versions makes the EPYC 9655 the obvious choice for 3D animation studios or video production houses that rely on CPU rendering. The 25.7% lead in floating-point math also points to a strong fit for scientific computing, simulations, and AI inference tasks that require extensive matrix operations.

The Intel Xeon 6960P does not have a single benchmark win to claim. However, its architecture offers some theoretical advantages. The larger 432 MB L3 cache and higher 614.4 GB/s memory bandwidth could be beneficial in specific workloads that are heavily reliant on data reuse and memory throughput. The higher per-core L2 cache (2 MB vs 1 MB) might also help in certain latency-sensitive applications. However, the available data does not show a scenario where these features translate into a performance victory.

In terms of scalability, the AMD's 128 PCIe Gen 5 lanes versus the Intel's 96 provides more headroom for expanding storage or GPU clusters within a single socket. The AMD's lower TDP of 400W versus the Intel's 500W also suggests it could be easier to deploy in dense server environments with power constraints.

The Verdict

The data is unambiguous. The AMD EPYC 9655 wins every single head-to-head benchmark against the Intel Xeon 6960P. The average benchmark score for the AMD is 373,479, placing it at the 100th percentile. The Intel Xeon 6960P also sits at the 100th percentile with an average score of 365,194, but the delta is clear: the AMD is 2.2% ahead of the Intel based on the rival comparison data.

For buyers prioritizing raw compute performance, the AMD EPYC 9655 is the only logical choice. Its 96 cores and higher clock speeds deliver a commanding lead in multi-threaded tasks, while its single-thread performance advantage of 17% ensures it does not lag in lightly threaded applications. The launch MSRP for the AMD EPYC 9655 is $11,852, and for the Intel Xeon 6960P it is $9,625.

The Intel Xeon 6960P could still be considered by organizations that are deeply invested in Intel's server ecosystem, or if the lower launch MSRP is a determining factor. Its larger cache and higher memory bandwidth are features that might be leveraged in specific, cache-heavy enterprise applications. However, for any general-purpose server role, the AMD EPYC 9655's consistent and significant performance lead makes it the superior processor.

FAQ

Q: Which processor has more cores?

A: The AMD EPYC 9655 has 96 cores and 192 threads, while the Intel Xeon 6960P has 72 cores and 144 threads.

Q: What is the biggest performance difference between the two?

A: The largest delta is in the PassMark integer math test, where the AMD EPYC 9655 is 56.5% ahead of the Intel Xeon 6960P.

Q: Does the Intel Xeon 6960P win any benchmarks?

A: No. The AMD EPYC 9655 wins all 14 head-to-head benchmark comparisons recorded in the data.

Q: How do they compare in single-threaded performance?

A: The AMD EPYC 9655 is 17% faster, scoring 3,847 in the PassMark single-thread test compared to the Intel's 3,287.

Q: What are the memory specifications for each?

A: Both support DDR5 memory with a twelve-channel bus. The Intel Xeon 6960P has a higher theoretical bandwidth of 614.4 GB/s, while the AMD EPYC 9655 is rated at 576.0 GB/s.

Q: What is the difference in their TDP ratings?

A: The AMD EPYC 9655 has a TDP of 400W, while the Intel Xeon 6960P has a higher TDP of 500W.