AMD EPYC 7601 vs Intel Core i5-8350U Comparison

Head-to-Head Benchmarks

The recorded data leaves no ambiguity: the AMD EPYC 7601 wins all six head-to-head comparisons in the database. The largest margin appears in Cinebench R23 multi-core, where the EPYC 7601 scores 29,800 against the Intel Core i5-8350U’s 5,173, a delta of -82.6%. That means the EPYC delivers roughly 5.8 times the multi-threaded rendering throughput in that test. The single-core gap is proportionally similar: 4,207 versus 730 in Cinebench R23 single-core, also a -82.6% delta. The EPYC leads by the same -82.7% margin in Cinebench R15 multi-core (3,003 vs. 521) and single-core (423 vs. 73). Cinebench R20 repeats the pattern, with the EPYC scoring 12,516 multi-core and 1,766 single-core, compared to 2,172 and 306 for the Intel part, deltas of -82.6% and -82.7% respectively.

What these deltas reveal is that the EPYC 7601 is not merely ahead, it is ahead by a near-constant factor across every Cinebench iteration. The consistency of the -82.6% to -82.7% range suggests a fundamental throughput advantage tied to core count and thread count, not a workload-specific quirk. The i5-8350U’s best showing in the head-to-head is its Cinebench R23 multi-core score, where it reaches 5,173, but that still trails the EPYC by more than fivefold. Conversely, the EPYC’s smallest relative advantage appears in the same test, but the absolute gap remains enormous.

For context, the database places both processors at the 65th percentile among all CPUs, yet their average benchmark scores differ: the i5-8350U averages 8,998 across its recorded tests, while the EPYC 7601 averages 8,619. That near-parity in average score, despite the EPYC’s dominance in Cinebench, reflects the i5’s strong showings in PassMark workloads not covered in the head-to-head set. The i5-8350U records 79,781 in data compression, 13,422 in floating point math, and 22,244 in integer math, while the EPYC has no comparable PassMark entries in the database. The head-to-head table only includes Cinebench, so the EPYC’s 6-0 win record does not capture the full picture of the i5’s single-thread and memory-latency-sensitive strengths.

FAQ

Q: Which processor wins more head-to-head benchmarks in the database?

A: The AMD EPYC 7601 wins all six recorded head-to-head comparisons. The Intel Core i5-8350U has zero wins in that table.

Q: How large is the multi-core performance gap in Cinebench R23?

A: The EPYC 7601 scores 29,800 versus the i5-8350U’s 5,173, a delta of -82.6%. The EPYC is approximately 5.8 times faster in that specific test.

Q: Does the Intel Core i5-8350U have any benchmark where it beats the EPYC 7601?

A: In the head-to-head benchmark set, no. However, the database includes PassMark tests for the i5 only, such as 79,781 in data compression and 22,244 in integer math, but these have no corresponding EPYC scores for direct comparison.

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

A: The Intel Core i5-8350U has an average benchmark score of 8,998. The AMD EPYC 7601 has an average of 8,619. Despite the EPYC’s Cinebench dominance, the i5’s average is slightly higher because of its PassMark results.

Q: Are both processors in the same performance percentile?

A: Yes, both the i5-8350U and the EPYC 7601 sit at the 65th percentile among all CPUs in the database.

Q: What are the nearest rivals for each processor?

A: For the i5-8350U, the nearest rivals include the Intel Core i5-1135G7 (average score 9,053, delta -0.6%) and the Intel Xeon Gold 6330 (8,939, delta 0.7%). For the EPYC 7601, the nearest rivals are the Intel Core i7-10510U (8,580, delta 0.5%) and the Intel Core i5-8250U (8,577, delta 0.5%).

Architecture Differences

The two processors come from entirely different design philosophies. The Intel Core i5-8350U uses the Kaby Lake architecture, specifically the Kaby Lake-R refresh, built on Intel’s 14 nm process at Intel’s foundry. Its die measures 123 mm². The AMD EPYC 7601 uses the Zen architecture, codenamed Naples, also on a 14 nm process but fabricated by GlobalFoundries. Its die is larger at 213 mm², and it packs 4,800 million transistors, a figure the Intel part does not disclose in the database.

Core organization differs fundamentally. The i5-8350U has 4 cores and 8 threads, with a per-core L1 cache of 64 KB, per-core L2 of 256 KB, and a shared 6 MB L3. The EPYC 7601 has 32 cores and 64 threads, with a per-core L1 of 96 KB, per-core L2 of 512 KB, and a shared 64 MB L3. That 64 MB L3 is over ten times larger than the i5’s entire L3, which directly benefits server-style workloads with large working sets. The EPYC also supports ECC memory, while the i5 does not. The i5 includes integrated UHD 620 graphics; the EPYC has no integrated graphics at all, reflecting its server/workstation target.

Memory architecture diverges sharply. The i5 supports DDR4 but the database lists no memory bus width or bandwidth. The EPYC lists an eight-channel memory bus with 170.6 GB/s of bandwidth, a figure that indicates massive parallel memory throughput. The EPYC also specifies PCIe Gen 3 support, while the i5’s PCIe details are absent from the database. Both are built on 14 nm, but the EPYC’s larger die and transistor count point to a much more complex memory and core fabric.

The release dates differ by roughly two months: the i5-8350U launched on 2017-08-20, the EPYC 7601 on 2017-06-28. Both are marked as Active in production status. The i5 is classified as Mobile, while the EPYC is Server/Workstation. The i5’s multiplier is locked; the EPYC’s is unlocked, allowing user-controlled overclocking in compatible platforms.

Specification Differences

The two parts differ across nearly every core specification. The i5-8350U has 4 cores and 8 threads; the EPYC 7601 has 32 cores and 64 threads. Base clocks are 1700.00 MHz for the i5 and 2.20 GHz for the EPYC. Boost clocks are 3.60 GHz for the i5 and 3.20 GHz for the EPYC, meaning the i5 actually has a higher single-core boost, but the EPYC’s base clock is higher. TDP is a stark contrast: 15 watts for the i5 versus 180 watts for the EPYC, a twelvefold difference that reflects the mobile versus server positioning.

Sockets are incompatible: Intel BGA 1356 for the i5, AMD Socket SP3 for the EPYC. Cache configurations differ in every level, with the EPYC’s per-core L1 at 96 KB versus 64 KB, per-core L2 at 512 KB versus 256 KB, and shared L3 at 64 MB versus 6 MB. The EPYC’s eight-channel memory bus and 170.6 GB/s bandwidth have no counterpart in the i5’s database entry. ECC memory is present on the EPYC, absent on the i5. Integrated graphics exist only on the i5 (UHD 620). The EPYC’s part number is PS7601BDVIHAF; the i5’s is SR3L9. The EPYC is an unlocked multiplier part; the i5 is locked. The EPYC belongs to the EPYC 7001 series, while the i5 has no series designation in the database.

Where Each One Wins

The AMD EPYC 7601 is the clear winner for any workload that scales with core count and thread count. The Cinebench results, all six of them, show the EPYC at least 5.8 times faster in multi-core and about 5.7 times faster in single-core, despite its lower boost clock. The 64 MB L3 cache and eight-channel memory bandwidth make it suited for server virtualization, database workloads, and compute-heavy rendering where data sets exceed the i5’s 6 MB L3. The EPYC’s ECC support and 180-watt TDP align with always-on server environments where reliability and sustained throughput matter more than power efficiency. The unlocked multiplier offers flexibility for tuning in workstation chassis.

The Intel Core i5-8350U wins in scenarios the head-to-head table does not capture. Its PassMark results, while not compared directly to the EPYC, show strong absolute numbers: 79,781 in data compression, 22,244 in integer math, 13,422 in floating point math, and 6,129 in multithread. These scores, combined with a 15-watt TDP, make the i5 the appropriate choice for battery-powered mobile devices where the EPYC’s 180-watt envelope is physically impossible. The i5’s integrated UHD 620 graphics eliminate the need for a discrete GPU in basic display tasks, another feature the EPYC lacks. The i5 also holds a higher boost clock at 3.60 GHz versus 3.20 GHz, which helps in lightly threaded, latency-sensitive applications such as web browsing or office productivity.

In terms of average benchmark scores, the i5 actually edges the EPYC: 8,998 versus 8,619. That means the i5’s PassMark workloads, which include single-threaded and mixed operations, compensate for its massive Cinebench deficit. For a user prioritizing a balanced mobile experience, the i5 is the data-backed choice. For a user prioritizing raw parallel compute, the EPYC is the only option between these two. The database places both at the 65th percentile, but that percentile masks how differently they achieve that standing: the i5 through broad, moderate scores, the EPYC through extreme multi-core peaks and no recorded PassMark data. The wins table says 6-0 for the EPYC, but the full benchmark record shows the i5’s strengths live outside that narrow comparison.