AMD EPYC 9115 vs Intel Xeon 6515P Comparison

Head-to-Head Benchmarks

The head-to-head record between the AMD EPYC 9115 and the Intel Xeon 6515P is decisively lopsided. The AMD part wins 14 of the 17 recorded comparisons, while the Intel part manages only 3 victories. The breadth of AMD's advantage spans nearly every category of workload, from single-threaded responsiveness to multi-threaded throughput.

Starting with the Cinebench suite, the EPYC 9115 outperforms the Xeon 6515P consistently across all six tests. In Cinebench R15 multicore, the AMD chip scores 4233 against Intel's 3885, a 9% advantage. The single-core version of that same test shows a nearly identical gap: 597 versus 548, or 8.9%. Moving to Cinebench R20, the multicore result is 17641 versus 16189, again a 9% difference, while the single-core score of 2490 versus 2285 represents another 9% lead. Cinebench R23 follows the same pattern, with the AMD part posting 42003 multicore and 5929 single-core, compared to Intel's 38547 and 5442. Both of those represent 9% and 8.9% deltas respectively. These results indicate that the EPYC 9115's architectural efficiency translates directly into measurable performance gains across rendering and CPU-bound compute tasks.

The Passmark suite tells a more nuanced story. The EPYC 9115 wins the integer math test by a substantial margin, scoring 181807 against Intel's 147047, a 23.6% lead. This is the largest single advantage recorded in either direction. The AMD part also takes the single-threaded test decisively, posting 3360 versus 2855, a 17.7% gap. Data encryption favors AMD at 33489 versus 30476, a 9.9% difference, while physics simulation shows a 9.4% edge (4188 versus 3829). Multithread performance lands at 48936 versus 45350, a 7.9% win for AMD, and random string sorting goes to AMD at 70151 versus 64425, an 8.9% margin. Data compression is the closest contest, with the EPYC 9115 edging out the Xeon 6515P by just 1.3%, scoring 600099 versus 592645.

The Intel Xeon 6515P's three wins are concentrated in specific computational domains. The extended instructions test goes to Intel by a wide margin: 53383 versus 45477, a 14.8% advantage. Floating point math also favors Intel at 128954 versus 113853, an 11.7% lead. The largest Intel victory comes in the find prime numbers test, where the Xeon scores 385 versus the EPYC's 289, a 24.9% gap. These results suggest that Intel's architecture retains a clear strength in certain mathematical and instruction-heavy workloads, even while trailing in most general-purpose benchmarks.

Looking at the aggregate benchmark scores, the EPYC 9115 averages 69288 across all recorded tests, while the Xeon 6515P averages 67006. The AMD part's percentile ranking sits at 94 compared to Intel's 93. The nearest rival data reinforces this positioning: the EPYC 9115's closest competitor is the Intel Core i7-14700K with an average score of 69355, a negligible 0.1% delta. The Xeon 6515P sits near the AMD EPYC 4465P, which scores 66925, just 0.1% behind.

Architecture Differences

The two processors represent fundamentally different design philosophies. The AMD EPYC 9115 is built on the Zen 5 architecture, codenamed Turin, and belongs to the EPYC 9005 series. It uses a 4 nm process node fabricated by TSMC, with a transistor count of 16,630 million spread across a dual-chip design measuring 2x 70.6 mm². The Intel Xeon 6515P, by contrast, uses the Granite Rapids architecture, also known as Granite Rapids-SP, built on Intel's 5 nm process node. The database does not record transistor counts or die sizes for the Intel part.

Both processors feature 16 cores and 32 threads, but their cache hierarchies differ substantially. The AMD chip provides 80 KB of L1 cache per core and 1 MB of L2 per core, with 64 MB of shared L3. Intel's design offers 112 KB of L1 per core and 2 MB of L2 per core, with a larger 72 MB of shared L3. This means Intel provides more per-core cache at every level, yet the AMD part still manages to outperform it in most benchmarks, suggesting that Zen 5's cache utilization and memory subsystem are more efficient overall.

Memory support shows another significant divergence. Both processors support DDR5 and ECC memory, but the AMD EPYC 9115 features a twelve-channel memory bus delivering 576.0 GB/s of bandwidth. The Intel Xeon 6515P uses an eight-channel configuration with 409.6 GB/s. That difference in memory bandwidth, roughly 40% in AMD's favor, helps explain the EPYC's advantages in memory-sensitive workloads like data compression and random string sorting.

PCIe connectivity also differs. The AMD part provides Gen 5 with 128 lanes from the CPU, while Intel offers Gen 5 with 88 lanes. Both are server-class parts with no integrated graphics and both target the server and workstation market segment. The AMD processor uses Socket SP5, while Intel uses Socket 4710. Both are currently active production parts, with the EPYC 9115 released on October 9, 2024, and the Xeon 6515P released on February 23, 2025.

Clock speeds reveal another dimension of the AMD advantage. The EPYC 9115 runs at a base clock of 2.60 GHz and boosts to 4.10 GHz. The Xeon 6515P operates at 2.30 GHz base and 3.80 GHz boost. This 0.30 GHz advantage in base clock and 0.30 GHz in boost clock contributes to the AMD part's strong single-threaded showing. Interestingly, the Intel part carries a higher TDP of 150 watts compared to AMD's 125 watts, meaning the EPYC 9115 achieves better performance while drawing less power, an efficiency gap that the benchmark data consistently reflects.

The Verdict

The data paints a clear picture. The AMD EPYC 9115 is the superior processor in 14 of 17 head-to-head comparisons, with wins spanning Cinebench multicore and single-core, integer math, encryption, physics, multithread, random string sorting, data compression, and single-threaded performance. The Intel Xeon 6515P wins only in extended instructions, floating point math, and prime number finding. For general server and workstation workloads, the EPYC 9115 is the stronger choice, offering higher memory bandwidth, more PCIe lanes, higher clock speeds, and lower TDP.

The Intel part does hold specific advantages that matter for niche applications. Its floating point math score of 128954 versus 113853 indicates a 13.3% advantage in that domain. The extended instructions result, 53383 versus 45477, points to strengths in SIMD-heavy or specialized instruction workloads. The prime number finding test, where Intel leads by 24.9%, suggests particular efficiency in certain algorithmic patterns. Any workload that depends heavily on these specific capabilities would favor the Xeon 6515P.

Specification Differences

The two processors differ across several key specification fields. The AMD EPYC 9115 uses the Zen 5 architecture with the Turin codename, built on a 4 nm TSMC process with 16,630 million transistors and a die size of 2x 70.6 mm². The Intel Xeon 6515P uses Granite Rapids architecture on Intel's 5 nm process, with no transistor or die size data recorded.

Clock speeds differ significantly: the AMD part runs at 2.60 GHz base and 4.10 GHz boost, while Intel runs at 2.30 GHz base and 3.80 GHz boost. TDP also diverges, with AMD at 125 watts and Intel at 150 watts. The cache layouts differ, with Intel offering more per-core L1 (112 KB versus 80 KB) and L2 (2 MB versus 1 MB), plus larger shared L3 (72 MB versus 64 MB). Memory bandwidth heavily favors AMD at 576.0 GB/s versus Intel's 409.6 GB/s, reflecting the twelve-channel versus eight-channel memory bus difference. PCIe lanes also differ: AMD provides 128 Gen 5 lanes, Intel provides 88. The AMD part uses Socket SP5, Intel uses Socket 4710. Release dates are October 9, 2024, for AMD and February 23, 2025, for Intel. The launch MSRP for the AMD EPYC 9115 is $726, and for the Intel Xeon 6515P it is $740.

FAQ

Q: Which processor has higher single-threaded performance?

A: The AMD EPYC 9115 leads in every single-threaded benchmark. In Cinebench R23 single-core, it scores 5929 versus 5442, an 8.9% advantage. Passmark single-thread shows a larger gap at 3360 versus 2855, a 17.7% lead.

Q: Does the Intel Xeon 6515P win any benchmarks?

A: Yes, the Intel part wins three tests: extended instructions (53383 versus 45477, a 14.8% lead), floating point math (128954 versus 113853, an 11.7% lead), and find prime numbers (385 versus 289, a 24.9% lead).

Q: How do the processors compare in memory bandwidth?

A: The AMD EPYC 9115 provides 576.0 GB/s over a twelve-channel DDR5 bus, while the Intel Xeon 6515P delivers 409.6 GB/s over an eight-channel DDR5 bus. This gives AMD a substantial bandwidth advantage.

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

A: Both processors have 16 cores and 32 threads. They are identical in this regard, making performance differences attributable to architecture, clocks, and memory subsystems rather than core count.

Q: What is the TDP difference?

A: The AMD EPYC 9115 has a TDP of 125 watts, while the Intel Xeon 6515P has a TDP of 150 watts. The AMD part delivers higher performance in most benchmarks while drawing less power.

Q: Which processor has more PCIe lanes?

A: The AMD EPYC 9115 provides 128 Gen 5 lanes from the CPU, whereas the Intel Xeon 6515P provides 88 Gen 5 lanes. This affects expansion capability for GPUs, NVMe storage, and networking.

Where Each One Wins

The AMD EPYC 9115 is the clear winner for general-purpose server and workstation tasks. Its 23.6% advantage in integer math makes it the better choice for database operations, compilation, and general compute. The 17.7% single-threaded lead benefits latency-sensitive applications and lightly threaded workloads. The 9.9% encryption advantage supports secure communications and cryptographic workloads. The 9.4% physics win suits simulation and gaming server environments. Data compression, with a 1.3% edge, favors AMD for storage and analytics workloads. Multithread performance at 7.9% ahead helps in heavily parallelized tasks like video rendering, scientific computing, and virtualization.

The Intel Xeon 6515P's wins define its niche. The 24.9% prime number finding advantage suggests strength in number-theoretic computations and certain cryptography algorithms. The 11.7% floating point math lead benefits scientific simulations, weather modeling, and engineering analysis that rely heavily on floating point operations. The 14.8% extended instructions advantage indicates superior performance in workloads using advanced SIMD instructions, such as media encoding and signal processing.

The aggregate data places the EPYC 9115 at the 94th percentile of all CPUs, while the Xeon 6515P sits at the 93rd percentile. The average benchmark score difference, 69288 versus 67006, represents a 3.4% overall advantage for AMD. When selecting between these two processors, the decision comes down to whether the workload leans toward general-purpose compute, favoring AMD, or toward specialized floating point and instruction-heavy tasks, where Intel retains its strengths. For most server and workstation deployments, the AMD EPYC 9115's broader benchmark dominance, higher memory bandwidth, and lower power draw make it the more versatile choice.