CPU Comparison

The Intel Core i3-10105 and the AMD EPYC 7502 occupy opposite ends of the processor spectrum, and the benchmark data reflects this stark divide. The EPYC 7502, a 32-core server behemoth, wins every head-to-head benchmark in the pack, while the i3-10105, a 4-core desktop chip, holds its ground only in areas not directly compared here. The data shows a 6-0 sweep for the AMD part, but the story is more nuanced than a simple victory tally, as the single-core and multi-core deltas reveal distinct performance profiles.

Head-to-Head Benchmarks

The Cinebench suite provides the clearest picture of the performance gap between these two processors. In the R15 multi-core test, the AMD EPYC 7502 scores 4,428 against the Intel Core i3-10105’s 708, a delta of -84% from the Intel part’s perspective. This translates to the EPYC delivering over six times the multi-threaded throughput, a direct consequence of its 32 cores versus the i3’s 4. The R20 multi-core test shows a similar chasm: 18,454 versus 2,952, again an -84% delta. The R23 multi-core result is even more lopsided in absolute terms, with the EPYC scoring 43,940 compared to the i3’s 7,029, maintaining that consistent -84% gap.

Single-core performance tells a different story about architectural efficiency, though the winner remains the same. In Cinebench R15 single-core, the EPYC 7502 scores 625, while the i3-10105 manages only 99, a -84.2% delta. The R20 single-core test shows 2,605 versus 416, and R23 single-core shows 6,203 versus 992, each with the same -84% delta. This consistency across the board is striking: the EPYC 7502 is not just faster in multi-core; it also leads in single-core by the same proportional margin. The data implies that the Zen 2 architecture’s per-core performance, even at lower clock speeds, outpaces the older Comet Lake design in these specific workloads.

The head-to-head table lists six benchmarks, and the EPYC 7502 wins all six. There are no tests in the pack where the i3-10105 comes out ahead. The deltas are uniform at approximately -84%, suggesting that the EPYC’s advantage is structural, a fundamental difference in core count, memory bandwidth, and cache size, rather than workload-specific. For a potential buyer, this means the EPYC 7502 is categorically superior in every measured Cinebench scenario, making the i3-10105 a non-competitive option for anyone prioritizing these rendering tasks.

FAQ

Q: Which processor wins the most benchmarks in the head-to-head comparison?

A: The AMD EPYC 7502 wins all six head-to-head benchmarks listed, covering Cinebench R15, R20, and R23 in both single-core and multi-core tests. The Intel Core i3-10105 records zero wins in this direct comparison.

Q: How much faster is the EPYC 7502 in multi-core Cinebench R23?

A: The EPYC 7502 scores 43,940 in Cinebench R23 multi-core, while the i3-10105 scores 7,029. This represents a -84% delta, meaning the AMD processor delivers roughly six times the multi-threaded performance in this specific test.

Q: Is the single-core performance gap as large as the multi-core gap?

A: Yes, the data shows a consistent gap. In Cinebench R23 single-core, the EPYC 7502 scores 6,203 versus the i3-10105’s 992, a -84% delta. The R15 and R20 single-core tests also show similar proportional differences at -84.2% and -84%, respectively.

Q: What is the average benchmark score for each processor?

A: The Intel Core i3-10105 has an average benchmark score of 12,316, while the AMD EPYC 7502 has an average score of 12,709. Both processors sit at the 67th percentile compared to all CPUs in the database.

Q: How do the two processors compare to their nearest rivals?

A: The i3-10105 is 0.6% ahead of the Intel Core i7-6700K, 2.2% ahead of the Intel Core i3-12100, and 1.8% behind the AMD EPYC 9174F. The EPYC 7502 is 0.5% ahead of the Intel Core i3-10105F, 0.3% behind the Intel Xeon Gold 6348, and 0.6% behind the AMD Ryzen Threadripper 3990X.

Q: Does the i3-10105 have any benchmark advantage in the pack?

A: No. The head-to-head data shows zero wins for the i3-10105. Its only listed benchmarks are the six Cinebench tests, all of which are won by the EPYC 7502.

Where Each One Wins

The EPYC 7502 wins every benchmark where both processors are measured, which makes a use-case split straightforward from the data. For any workload that relies on Cinebench-style rendering, which typically stresses multi-threaded throughput, the EPYC 7502 is the clear choice. Its 32 cores and 64 threads, combined with 128 MB of shared L3 cache, provide the resources needed for heavy parallel computation. The i3-10105, with its 4 cores and 8 threads, cannot compete in these scenarios, as the -84% deltas demonstrate.

The i3-10105’s potential strengths lie outside the measured benchmarks. It is a Desktop market segment part with integrated UHD Graphics 630, which the EPYC 7502 lacks entirely. For a system that requires a display output without a discrete GPU, the i3-10105 is the only option between the two. Its 65W TDP also suggests a much lower power draw than the EPYC’s 180W, though the pack does not provide direct power consumption test results. The i3-10105 also uses the Intel Socket 1200 platform, which is a mainstream consumer socket, while the EPYC 7502 requires the AMD Socket SP3 server platform.

For server and workstation workloads, the EPYC 7502 is the definitive winner. Its eight-channel memory bus, with 204.8 GB/s of bandwidth, dwarfs the i3-10105’s dual-channel 42.7 GB/s. The EPYC also supports ECC memory, a feature the i3 lacks. The data implies that the EPYC 7502 is designed for data-heavy, memory-sensitive applications, while the i3-10105 is a general-purpose desktop chip. The absence of any i3 wins in the benchmark pack reinforces this division: the EPYC 7502 is the performance leader, and the i3-10105 is only relevant for its unique desktop features like integrated graphics.

Specification Differences

The core specification gap is the defining difference. The Intel Core i3-10105 has 4 cores and 8 threads, while the AMD EPYC 7502 has 32 cores and 64 threads. Clock speeds also differ, with the i3-10105 running at a 3.70 GHz base and 4.40 GHz boost, versus the EPYC’s 2.50 GHz base and 3.35 GHz boost. The i3’s higher clocks suggest it is tuned for single-thread responsiveness, but the benchmark data shows the EPYC still wins single-core tests despite lower frequencies.

Cache configurations are vastly different. The i3-10105 has 64 KB of L1 cache per core, 256 KB of L2 per core, and 6 MB of shared L3. The EPYC 7502 has 96 KB of L1 per core, 512 KB of L2 per core, and 128 MB of shared L3. This 21x difference in L3 cache is a major factor in the EPYC’s favor for large datasets. Memory support also diverges: both use DDR4, but the i3 is dual-channel with 42.7 GB/s bandwidth, while the EPYC is eight-channel with 204.8 GB/s. ECC memory is supported only on the EPYC.

The process node and foundry differ: the i3-10105 is built on Intel’s 14 nm process, while the EPYC 7502 uses TSMC’s 7 nm node. The EPYC’s transistor count is listed as 3,800 million with a 74 mm² die size, while the i3’s transistor count and die size are not provided. PCIe support also differs, with the i3 offering Gen 3 with 16 lanes (CPU only), and the EPYC offering Gen 4. The i3 has integrated graphics, the EPYC does not. The i3 has a launch MSRP of $122; the EPYC has no launch MSRP listed.

Architecture Differences

The two processors come from fundamentally different architectural eras. The Intel Core i3-10105 uses the Comet Lake architecture, specifically the Comet Lake-R codename, built on a 14 nm process. This is a refined version of Intel’s older Skylake-derived design, optimized for desktop efficiency but limited in core count. The EPYC 7502 uses the Zen 2 architecture, with the Rome codename, built on TSMC’s 7 nm process. Zen 2 introduced a chiplet design, which the data suggests through the 74 mm² die size and 3,800 million transistors, though the pack does not explicitly describe chiplets.

The generation gap is clear from the naming: the i3-10105 is from Core 10th Gen, while the EPYC 7502 is from the EPYC 7002 series. The i3’s generation is listed as “Core i3 (Comet Lake),” and the EPYC’s as “EPYC (Zen 2 (Rome)).” The production status for both is “Active,” meaning neither is discontinued. The release dates differ, with the i3-10105 released in March 2021 and the EPYC 7502 in August 2019, placing the EPYC as the earlier design that still outperforms the newer Intel part in every measured benchmark.

Architectural features also diverge in memory and I/O. The EPYC’s eight-channel memory controller and 204.8 GB/s bandwidth are server-class features, supporting ECC for data integrity. The i3’s dual-channel controller and 42.7 GB/s bandwidth are consumer-grade. The EPYC’s PCIe Gen 4 support is a generation ahead of the i3’s PCIe Gen 3. The absence of integrated graphics on the EPYC is a deliberate trade-off for server workloads, while the i3’s UHD Graphics 630 is a desktop convenience. The benchmark results, with the EPYC winning single-core despite lower clocks, suggest the Zen 2 architecture has higher instructions-per-clock efficiency than Comet Lake in these tests.

The Verdict

The data unequivocally favors the AMD EPYC 7502 for any performance-focused workload. Across all six Cinebench tests, the EPYC wins with a consistent -84% delta over the Intel Core i3-10105. The multi-core scores, 4,428 vs 708 in R15, 18,454 vs 2,952 in R20, and 43,940 vs 7,029 in R23, show that the EPYC offers roughly six times the throughput in rendering tasks. Even in single-core tests, the EPYC leads, scoring 6,203 vs 992 in R23, which is surprising given the i3’s higher boost clock of 4.40 GHz versus 3.35 GHz. The EPYC’s 128 MB of L3 cache, 204.8 GB/s memory bandwidth, and 32 cores provide a structural advantage that clock speed cannot overcome.

Who should pick the Intel Core i3-10105? The pack offers no benchmark justification, as the i3 has zero wins. The only rational reasons would be its integrated UHD Graphics 630, which the EPYC lacks, or its lower 65W TDP versus 180W, suggesting lower power consumption. The i3 also uses the Intel Socket 1200, which is a mainstream platform. However, for anyone measuring performance via the provided benchmarks, the i3-10105 is non-competitive. The EPYC 7502 is the choice for server, workstation, or any multi-threaded rendering workload, with its eight-channel memory and ECC support adding to its enterprise appeal. The i3-10105 is only defensible for a basic desktop build where integrated graphics are required and performance is not the priority. Both processors sit at the 67th percentile, but the EPYC’s average benchmark score of 12,709 edges out the i3’s 12,316, confirming the EPYC as the stronger overall performer in this comparison.