AMD EPYC 7513 vs Intel Xeon 6520P Comparison

Head-to-Head Benchmarks

The recorded benchmark data reveals a clear split in workload preferences between these two server processors. The Intel Xeon 6520P dominates the Cinebench suite, winning all six multicore and singlecore tests. In Cinebench R23 multicore, the Xeon scores 53495 against the EPYC 7513's 50431, a 5.7% advantage. The same 5.7% delta appears across Cinebench R15 multicore (5392 vs 5083), R20 multicore (22467 vs 21181), and every singlecore variant in the series. This consistency suggests a per-thread performance edge baked into the Intel architecture, not a workload-specific quirk.

The Passmark results tell a more complex story. The AMD EPYC 7513 wins four of the seventeen head-to-head tests, and its victories are substantial. Data encryption shows the largest gap: the EPYC scores 63628 versus the Xeon's 45188, a 40.8% advantage. Integer math follows with a 27% lead (272145 vs 214288). Data compression (932240 vs 841518, 10.8% ahead) and random string sorting (104660 vs 95736, 9.3% ahead) round out the AMD wins. These are not marginal victories; they indicate that the EPYC 7513 has a meaningful architectural edge in specific algorithmic patterns.

The Intel Xeon 6520P counters with wins in the remaining thirteen tests. Its largest margin comes in Passmark physics, where it scores 7209 against the EPYC's 5118, a 29% lead. Find prime numbers shows a 27.8% advantage (526 vs 380), and single-thread performance is 26.1% higher (3356 vs 2479). Extended instructions (64557 vs 56451) and floating point math (162862 vs 151713) add smaller but still decisive wins. The Passmark multithread score goes to Intel at 62936 versus 59331, a 5.7% margin that mirrors the Cinebench multicore results.

Interpreting the data, the Xeon 6520P wins on raw throughput in most multithreaded and single-threaded scenarios, but the EPYC 7513's encryption and integer math results suggest that AMD's design excels at data manipulation tasks where the workload is less dependent on raw clock speed. The physics test, with its 29% Intel advantage, is the starkest example of a workload where the Xeon's higher boost clock and newer architecture pay off. The EPYC's 40.8% encryption win is equally stark in the opposite direction, indicating that AMD's Zen 3 core has a specialized advantage in cryptographic operations.

FAQ

Q: Which processor has the higher average benchmark score?

A: The AMD EPYC 7513 has an average benchmark score of 102244, placing it in the 97th percentile of all CPUs. The Intel Xeon 6520P averages 93786, which sits at the 96th percentile. Despite losing most head-to-head tests, the EPYC's larger average is driven by its massive wins in encryption and integer math.

Q: How does the Intel Xeon 6520P compare to its nearest rivals in the database?

A: The Xeon 6520P's average score of 93786 is essentially tied with the Intel Core Ultra 7 270K Plus, which has a delta of 0%. It trails the AMD EPYC 4564P by 1.5%, the AMD EPYC 9175F by 1.9%, and the AMD EPYC 4565P by 2.1%. None of these deltas exceed 2.1%, so the Xeon sits in a tightly contested performance band.

Q: What is the EPYC 7513's position relative to its nearest rivals?

A: The EPYC 7513 (102244 average) is 1.1% behind the AMD EPYC 8324P (103329), 1.2% ahead of the AMD Ryzen Threadripper PRO 9955WX (101041), and 2.9% ahead of the AMD EPYC 4585PX (99324). It also trails the Intel Xeon 6521P by 3.4%, which is notable because that rival is a different SKU in the same Xeon family as the 6520P.

Q: In which benchmark does the EPYC 7513 show its largest advantage over the Xeon 6520P?

A: The largest EPYC win is in Passmark data encryption, where it scores 63628 versus 45188, a 40.8% advantage. This is the single biggest delta in the entire head-to-head set.

Q: In which benchmark does the Intel Xeon 6520P show its largest advantage?

A: The largest Intel win is in Passmark physics, scoring 7209 against the EPYC's 5118, a 29% lead. The find prime numbers test is nearly as large at 27.8% (526 vs 380).

Q: How many total wins does each processor have in the head-to-head set?

A: The Intel Xeon 6520P wins 13 of the 17 recorded tests. The AMD EPYC 7513 wins 4 tests.

Architecture Differences

The two processors come from fundamentally different design generations. The AMD EPYC 7513 is built on Zen 3 architecture under the Milan codename, part of the EPYC 7003 series. It uses a 7 nm process at TSMC and packs 33,200 million transistors across a die configuration of 8x 81 mm². The Intel Xeon 6520P is Granite Rapids architecture, part of the Xeon 6 generation, built on Intel's 5 nm process with a single 598 mm² die. The process node difference alone suggests the Intel part has access to denser transistors, but the EPYC's multi-die approach allows for a larger aggregate transistor count.

Core counts differ significantly. The EPYC 7513 offers 32 cores and 64 threads, while the Xeon 6520P provides 24 cores and 48 threads. This 8-core, 16-thread advantage for AMD does not translate into consistent benchmark wins, which indicates that the Intel cores are individually more efficient in most recorded tests. Cache hierarchies also diverge. The EPYC has 64 KB of L1 per core, 512 KB of L2 per core, and 128 MB of shared L3 cache. The Xeon counters with 112 KB of L1 per core, 2 MB of L2 per core, and 144 MB of shared L3. Intel's larger per-core L1 and L2 allocations may explain its single-thread and physics test advantages.

Memory support is another generational split. The EPYC 7513 uses DDR4 with an eight-channel bus delivering 204.8 GB/s of bandwidth. The Xeon 6520P uses DDR5 with an eight-channel bus at 409.6 GB/s, exactly double the bandwidth. That memory bandwidth gap likely contributes to the Xeon's wins in floating point math and extended instructions, where data movement speed matters. Both support ECC memory. PCIe connectivity also differs: the EPYC offers Gen 4 with 128 lanes (CPU only), while the Xeon offers Gen 5 with 88 lanes (CPU only). The EPYC has more lanes but an older standard; the Xeon has fewer lanes at a faster standard.

Clock speeds favor Intel. The Xeon 6520P boosts to 4.00 GHz from a 2.40 GHz base, while the EPYC 7513 boosts to 3.65 GHz from a 2.60 GHz base. The Xeon's higher boost clock aligns with its single-thread benchmark dominance. Thermal design power is close: 200 W for the EPYC, 210 W for the Xeon. The EPYC uses AMD Socket SP3, the Xeon uses Intel Socket 4710. Neither processor has an unlocked multiplier. The integrated graphics field is null for AMD and listed as N/A for Intel, indicating no on-chip GPU for either.

The Verdict

The data supports a split decision based on workload type. The Intel Xeon 6520P is the stronger general-purpose compute processor. It wins the Cinebench multicore and singlecore tests across every version in the database, plus Passmark multithread, physics, floating point math, extended instructions, find prime numbers, and single-thread tests. Its 13 wins out of 17 head-to-head benchmarks make it the clear winner for rendering, physics simulation, and most general server workloads.

The AMD EPYC 7513 is the specialist. Its 40.8% encryption win and 27% integer math win are not anomalies; they represent a genuine architectural strength in data transformation tasks. Data compression and random string sorting also favor AMD. For workloads involving encryption, compression, or heavy integer operations, the EPYC 7513 is the better choice despite its overall loss in win count.

The average benchmark scores complicate the verdict. The EPYC's 102244 average beats the Xeon's 93786, a difference of roughly 9%, and the EPYC sits at the 97th percentile versus the Xeon's 96th. This means the EPYC's four wins are so large that they outweigh its thirteen losses in aggregate scoring. A buyer looking at average performance would favor the EPYC, while a buyer looking at test-by-test results would favor the Xeon. The database records show that neither processor is universally superior.

Specification Differences

The two processors differ in nearly every major specification category. The EPYC 7513 has 32 cores and 64 threads; the Xeon 6520P has 24 cores and 48 threads. Base clocks are 2.60 GHz for AMD and 2.40 GHz for Intel, but boost clocks flip the order: 3.65 GHz for AMD, 4.00 GHz for Intel. Thermal design power is 200 W for the EPYC and 210 W for the Xeon.

Process technology separates them clearly: 7 nm at TSMC for AMD, 5 nm at Intel for Intel. The EPYC uses a multi-die design with 8x 81 mm² chiplets and 33,200 million transistors; the Xeon is a monolithic 598 mm² die with no transistor count recorded in the database.

Cache configurations differ at every level. L1 is 64 KB per core for AMD versus 112 KB per core for Intel. L2 is 512 KB per core for AMD versus 2 MB per core for Intel. L3 is 128 MB shared for AMD versus 144 MB shared for Intel.

Memory support moves from DDR4 (AMD) to DDR5 (Intel), with bandwidth doubling from 204.8 GB/s to 409.6 GB/s. Both use eight-channel buses. PCIe generation advances from Gen 4 (AMD, 128 lanes) to Gen 5 (Intel, 88 lanes). The sockets are incompatible: AMD Socket SP3 versus Intel Socket 4710. The EPYC 7513 launched on 2021-03-14, while the Xeon 6520P launched on 2025-02-23, a four-year gap that explains the architectural generational differences. The launch MSRP for the EPYC 7513 is $2840, and for the Xeon 6520P it is $1295.

Where Each One Wins

The Intel Xeon 6520P wins across the Cinebench family: R15 multicore (5392 vs 5083), R15 singlecore (761 vs 717), R20 multicore (22467 vs 21181), R20 singlecore (3171 vs 2989), R23 multicore (53495 vs 50431), and R23 singlecore (7552 vs 7119). These are consistent 5.7% to 5.8% margins, suggesting a uniform per-core advantage in rendering workloads.

In Passmark, the Xeon takes multithread (62936 vs 59331), physics (7209 vs 5118, a 29% lead), extended instructions (64557 vs 56451), floating point math (162862 vs 151713), find prime numbers (526 vs 380), single thread (3356 vs 2479), and the duplicate singlethread entry (3356 vs 2479). Workloads involving physics simulation, prime number computation, and extended instruction sets should favor the Xeon.

The AMD EPYC 7513 wins data encryption (63628 vs 45188, 40.8% ahead), integer math (272145 vs 214288, 27% ahead), data compression (932240 vs 841518, 10.8% ahead), and random string sorting (104660 vs 95736, 9.3% ahead). Encryption-heavy workloads, integer-heavy computation, and data compression pipelines will see the largest benefits from choosing the EPYC. Notably, the EPYC's higher average benchmark score (102244 vs 93786) means that in aggregate, despite fewer wins, the AMD part delivers a higher overall performance profile across the full test suite.