AMD EPYC 7282 vs Intel Xeon Platinum 8260 Comparison

FAQ

Q: Which processor has more cores and threads?

A: The Intel Xeon Platinum 8260 has 24 cores and 48 threads, while the AMD EPYC 7282 has 16 cores and 32 threads. Despite this core count disadvantage, the EPYC 7282 wins all six head-to-head Cinebench tests.

Q: How much faster is the AMD EPYC 7282 in multi-core workloads?

A: The EPYC 7282 leads by 6% in Cinebench R15, R20, and R23 multi-core tests. For example, in R23 multi-core it scores 25,616 versus the Xeon's 24,173, a margin of 1,443 points.

Q: What is the single-core performance difference?

A: The EPYC 7282 wins every single-core test by roughly 6%. In Cinebench R20 single-core, it scores 1,518 versus 1,433 for the Xeon, a 5.9% advantage. The R15 and R23 single-core deltas are 6.1% and 6% respectively.

Q: Which processor has a smaller manufacturing process?

A: The AMD EPYC 7282 uses a 7 nm process from TSMC, while the Intel Xeon Platinum 8260 uses a 14 nm process from Intel. The EPYC also has a lower TDP of 120 watts compared to the Xeon's 165 watts.

Q: Do both processors support ECC memory?

A: Yes, both the AMD EPYC 7282 and the Intel Xeon Platinum 8260 support ECC memory. Both are designed for the server and workstation market segment.

Q: How do these processors compare in overall average benchmark score?

A: The AMD EPYC 7282 has an average benchmark score of 7,409, while the Intel Xeon Platinum 8260 averages 6,992. Both sit at the 63rd percentile among all CPUs in the database.

Architecture Differences

The AMD EPYC 7282 is built on the Zen 2 architecture, codenamed Rome, and is part of the EPYC 7002 series. It uses a 7 nm process manufactured by TSMC and contains 7,600 million transistors across two dies, each measuring 74 mm². The Intel Xeon Platinum 8260 uses the Cascade Lake architecture, codenamed Cascade Lake-SP, on Intel's 14 nm process with 8,000 million transistors.

Cache organization differs significantly. The EPYC 7282 provides 64 KB of L1 cache per core, 512 KB of L2 per core, and 32 MB of L3 per die, totaling 64 MB of L3. The Xeon Platinum 8260 also has 64 KB of L1 per core but doubles L2 to 1 MB per core, while offering 35.75 MB of shared L3 cache. The EPYC's split-die L3 arrangement with 32 MB per die contrasts with the Xeon's single shared pool.

Memory support also differs. The EPYC 7282 features eight-channel DDR4 memory with a bandwidth of 85.3 GB/s. The database records no memory bus width or bandwidth figure for the Xeon Platinum 8260. Both processors support ECC memory.

PCIe connectivity sets the EPYC apart: it offers PCIe Gen 4 with 128 lanes (CPU only). No PCIe specification is recorded for the Xeon Platinum 8260.

The EPYC 7282 fits the AMD Socket SP3, while the Xeon Platinum 8260 uses Intel Socket 3647. The EPYC's TDP is 120 watts, the Xeon's is 165 watts. Neither processor has an unlocked multiplier.

The EPYC 7282 was released on August 6, 2019, with a launch MSRP of $650. The Xeon Platinum 8260 was released on December 10, 2018; no launch MSRP is recorded.

Head-to-Head Benchmarks

The head-to-head data covers six Cinebench tests, and the AMD EPYC 7282 wins all six. The largest single margin is in Cinebench R15 single-core, where the EPYC scores 364 against the Xeon's 343, a 6.1% advantage. The remaining five tests show a remarkably consistent 5.9% to 6% lead for the EPYC.

In multi-core workloads, the EPYC 7282 scores 2,581 in Cinebench R15 versus 2,436 for the Xeon, a 6% lead. In Cinebench R20, the gap is 10,758 versus 10,152, again 6%. The R23 multi-core result shows 25,616 versus 24,173, also 6%. Across all three multi-core tests, the EPYC maintains a steady advantage despite having 8 fewer cores and 16 fewer threads than the Xeon.

Single-core results follow the same pattern. Cinebench R15 gives the EPYC 364 versus 343, a 6.1% win. Cinebench R20 shows 1,518 versus 1,433, a 5.9% margin. Cinebench R23 records 3,616 versus 3,412, a 6% lead.

The average benchmark scores reinforce the trend. The EPYC 7282 averages 7,409, while the Xeon Platinum 8260 averages 6,992. That difference of 417 points places the EPYC ahead by roughly 6%.

Looking at nearest rivals provides context. The EPYC 7282's average score of 7,409 sits within 0.2% of the Intel Core i9-9980XE (7,398) and 0.1% below the Intel Core i9-9990XE (7,415). The Xeon Platinum 8260's average of 6,992 is 0.3% above the Intel Pentium Gold G6400 (6,969) and 0.8% above the Intel Core i9-9960X (6,938). The Xeon also trails the Intel Core i9-7980XE (7,018) by 0.4%.

The Verdict

The data is unambiguous: the AMD EPYC 7282 outperforms the Intel Xeon Platinum 8260 in every recorded benchmark. The EPYC wins all six head-to-head Cinebench tests, with margins between 5.9% and 6.1%. It achieves this with 16 cores and 32 threads, compared to the Xeon's 24 cores and 48 threads, and with a lower TDP of 120 watts versus 165 watts.

For single-threaded workloads, the EPYC's higher base clock of 2.80 GHz versus 2.40 GHz and boost clock of 3.20 GHz versus 3.90 GHz still result in a consistent single-core win. The Zen 2 architecture and 7 nm process appear to deliver better instructions per clock than the 14 nm Cascade Lake design.

For multi-threaded workloads, the EPYC's 6% lead despite an 8-core deficit is notable. The 64 MB of total L3 cache and eight-channel memory with 85.3 GB/s bandwidth likely contribute to this efficiency. The Xeon's 35.75 MB shared L3 and unspecified memory configuration cannot close the gap.

The EPYC 7282 also offers PCIe Gen 4 with 128 lanes, a feature absent from the recorded data for the Xeon. Both processors target the server and workstation segment and both support ECC memory.

The verdict from the database is clear: choose the AMD EPYC 7282 for better measured performance in both single-core and multi-core Cinebench workloads. The Xeon Platinum 8260's only advantages in the recorded data are its higher core count, higher boost clock, and larger L2 cache per core, none of which translate into a benchmark win.

Specification Differences

The two processors differ on several recorded specifications:

Cores: AMD EPYC 7282 has 16 cores; Intel Xeon Platinum 8260 has 24 cores.

Threads: AMD EPYC 7282 has 32 threads; Intel Xeon Platinum 8260 has 48 threads.

Base Clock: AMD EPYC 7282 runs at 2.80 GHz; Intel Xeon Platinum 8260 runs at 2.40 GHz.

Boost Clock: AMD EPYC 7282 boosts to 3.20 GHz; Intel Xeon Platinum 8260 boosts to 3.90 GHz.

TDP: AMD EPYC 7282 is rated at 120 watts; Intel Xeon Platinum 8260 is rated at 165 watts.

Socket: AMD EPYC 7282 uses AMD Socket SP3; Intel Xeon Platinum 8260 uses Intel Socket 3647.

Process Node: AMD EPYC 7282 uses 7 nm (TSMC); Intel Xeon Platinum 8260 uses 14 nm (Intel).

Transistors: AMD EPYC 7282 has 7,600 million; Intel Xeon Platinum 8260 has 8,000 million.

Die Size: AMD EPYC 7282 uses 2x 74 mm² dies; no die size is recorded for the Xeon.

L2 Cache: AMD EPYC 7282 has 512 KB per core; Intel Xeon Platinum 8260 has 1 MB per core.

L3 Cache: AMD EPYC 7282 has 32 MB per die (64 MB total); Intel Xeon Platinum 8260 has 35.75 MB shared.

Memory Bus: AMD EPYC 7282 is eight-channel; no memory bus width is recorded for the Xeon.

Memory Bandwidth: AMD EPYC 7282 is rated at 85.3 GB/s; no bandwidth figure is recorded for the Xeon.

PCIe: AMD EPYC 7282 offers Gen 4 with 128 lanes (CPU only); no PCIe specification is recorded for the Xeon.

Release Date: AMD EPYC 7282 released on August 6, 2019; Intel Xeon Platinum 8260 released on December 10, 2018.

Launch MSRP: AMD EPYC 7282 has a launch MSRP of $650; no launch MSRP is recorded for the Xeon.

Where Each One Wins

The AMD EPYC 7282 wins every recorded benchmark category. In multi-core workloads, it leads by 6% across Cinebench R15, R20, and R23. This makes it the stronger choice for heavily threaded rendering, simulation, and compilation tasks, despite having fewer cores than the Xeon.

In single-core workloads, the EPYC 7282 also wins by approximately 6% across all three Cinebench versions. This suggests better per-thread efficiency, which benefits lightly threaded applications, database workloads, and general server responsiveness.

The EPYC 7282 further differentiates itself with platform features: PCIe Gen 4 with 128 lanes provides more modern I/O connectivity, and the eight-channel DDR4 memory bus with 85.3 GB/s bandwidth offers a defined memory throughput advantage. The 120 watt TDP is lower than the Xeon's 165 watts, which matters for dense server deployments.

The Intel Xeon Platinum 8260 has no benchmark wins in the recorded data. Its higher core count of 24 and thread count of 48, along with a 3.90 GHz boost clock and 1 MB L2 per core, are its distinguishing specifications. For workloads where those specific attributes matter more than the measured Cinebench results, such as certain database or virtualization scenarios that scale with raw thread count, the Xeon remains a viable platform choice. However, the recorded data provides no benchmark evidence of an advantage in any tested workload.

The average benchmark scores summarize the split: the EPYC 7282 at 7,409 sits in a competitive cluster with the Intel Core i9-9990XE and i9-9980XE, while the Xeon Platinum 8260 at 6,992 aligns with desktop-class parts like the Intel Pentium Gold G6400 and AMD Ryzen 3 3200G. The EPYC's placement among high-end desktop processors in the database underscores its performance efficiency. The Xeon's placement among entry-level desktop chips, despite its server positioning, indicates its measured performance trails the EPYC noticeably.

For builders prioritizing measured Cinebench performance, memory bandwidth, PCIe Gen 4 connectivity, and lower power draw, the AMD EPYC 7282 is the clear pick. For builders needing the Xeon's specific socket, higher core count, or Intel platform features, the Xeon Platinum 8260 remains an option, but the database shows no performance category where it comes out ahead.