AMD EPYC 7261 vs Intel Core i5-1345UE Comparison

The AMD EPYC 7261 and the Intel Core i5-1345UE occupy opposite corners of the CPU landscape: a socketed server chip built for sustained throughput, and a soldered mobile processor engineered for low-power portables. Despite that gulf in design intent, the recorded data shows them trading blows in a remarkably tight contest. Across the six recorded Cinebench entries, the EPYC 7261 wins all six, but every margin sits between roughly one and two percent. The two chips land at the same percentile against all CPUs in the database (50th), and their average benchmark scores, 2740 for the EPYC and 2698 for the Core i5, differ by well under two percent. On paper this is a statistical tie, and the interesting story lies in how differently each CPU arrives at nearly identical results.

Head-to-Head Benchmarks

Every recorded test in the database ends with the same winner. In Cinebench R15 multi-core, the EPYC 7261 scores 954 against 940 for the Core i5-1345UE, a 1.5 percent gap. The single-core R15 result is even closer: 134 versus 132, again a 1.5 percent margin. R20 follows the same pattern. The EPYC posts 3979 multi-core and 561 single-core, against 3918 and 553 for the Core i5, margins of 1.6 and 1.4 percent respectively.

Cinebench R23, the longest and most thermally demanding of the three renders, is where sustained throughput matters most. The EPYC 7261 scores 9476 in multi-core versus 9329, a 1.6 percent win, and 1337 versus 1317 in single-core, 1.5 percent apart. The consistency of these deltas is itself informative. A gap that stays flat from a short R15 run to a long R23 run indicates that neither chip is throttling relative to the other as the workload extends; the EPYC holds its lead under sustained load just as it does in burst workloads.

Context from each chip's rival set reinforces how evenly matched they are. The EPYC 7261's nearest rivals include the Intel Xeon E-2176M (average score 2751, 0.4 percent higher), the Xeon E5-2628L v3 (2724, 0.6 percent lower), the Ryzen 7 4700U (2722, 0.7 percent lower), and the Ryzen 3 4300GE (2759, 0.7 percent higher). The Core i5-1345UE sits among the Core i7-8700T (2701), the Core i7-1185G7E (2703), the Xeon E5-2640 v3 (2703), and the Xeon E-2226G (2709), all within 0.4 percent of its 2698 average. Both processors occupy the same performance neighborhood as mid-tier desktop and workstation parts, separated by a whisker in aggregate.

Architecture Differences

The EPYC 7261 belongs to the EPYC 7001 series and is built on the Zen microarchitecture, codenamed Naples. It uses an 8-core, 16-thread configuration on a 14 nm GlobalFoundries process, with a die size of 213 mm² and 4,800 million transistors. Its base clock is 2.50 GHz with a 2.90 GHz boost, and it carries a 170 W TDP in the AMD Socket SP3. Cache is a strength: 96 KB of L1 per core, 512 KB of L2 per core, and a generous 32 MB of shared L3. Memory support is DDR4 only, but through an eight-channel bus delivering 170.6 GB/s of bandwidth, with ECC support. PCIe is Gen 3. The chip has no integrated graphics, its multiplier is unlocked, and it remains in active production with part number PS7261BEV8RAF.

The Core i5-1345UE is a Raptor Lake-U part, fabricated on Intel's 10 nm process. It uses a hybrid 10-core, 12-thread layout (two more cores than the EPYC, but four fewer threads), with a base clock recorded at 1400 MHz and a boost clock of 4.60 GHz, dramatically higher than the EPYC's ceiling. Its TDP is 15 W, a fraction of the EPYC's envelope. Per-core L1 is 80 KB and per-core L2 is 1.25 MB, larger than the EPYC's per-core L2, but shared L3 is only 12 MB versus the EPYC's 32 MB. Memory support spans both DDR4 and DDR5 over a dual-channel bus, with no ECC. PCIe is Gen 4 with 8 lanes from the CPU. Integrated Iris Xe Graphics 80EU are present, the multiplier is locked, and the chip is soldered via Intel BGA 1744.

The generational gap is nearly five years: a 2018 server platform against a 2023 mobile platform. That the older chip still edges every test is a direct consequence of its power and platform advantages, while the Core i5's competitiveness comes from a newer node and much higher boost clocks.

Where Each One Wins

The benchmark ledger is unambiguous in direction but ambiguous in magnitude: the EPYC 7261 wins all six recorded tests, none by more than 1.6 percent. For rendering workloads specifically, the data shows no meaningful difference between the two. A 9476 versus 9329 R23 multi-core gap is within the range where run-to-run variance could plausibly flip a result.

Where the EPYC genuinely separates itself is in platform capability rather than raw score. Eight-channel DDR4 at 170.6 GB/s, ECC memory, and the SP3 socket make it the only option of the two suitable for memory-bound server tasks, data integrity-sensitive workloads, or deployments where the chip must be serviceable rather than soldered. Its unlocked multiplier also allows tuning, and its 16 threads give it more hardware parallelism than the Core i5's 12.

The Core i5-1345UE wins on everything the benchmarks cannot show. It achieves within two percent of the EPYC's performance at a 15 W TDP, which is the headline fact of this comparison. It adds integrated Iris Xe Graphics 80EU where the EPYC has none, supports both DDR4 and DDR5 including the newer standard, offers PCIe Gen 4 versus Gen 3, and boosts to 4.60 GHz versus 2.90 GHz for far better burst responsiveness in lightly threaded tasks. For any mobile, compact, or power-constrained design, it is the practical choice despite the clean sweep in the render column.

Specification Differences

• Cores/threads: EPYC 8 cores / 16 threads; Core i5 10 cores / 12 threads
• Base clock: EPYC 2.50 GHz; Core i5 1400 MHz
• Boost clock: EPYC 2.90 GHz; Core i5 4.60 GHz
• TDP: EPYC 170 W; Core i5 15 W
• Socket: AMD Socket SP3 versus Intel BGA 1744
• Architecture: Zen (Naples) versus Raptor Lake-U
• Process node: 14 nm GlobalFoundries versus 10 nm Intel
• L1 cache: 96 KB per core versus 80 KB per core
• L2 cache: 512 KB per core versus 1.25 MB per core
• L3 cache: 32 MB shared versus 12 MB shared
• Memory: DDR4 eight-channel at 170.6 GB/s versus DDR4/DDR5 dual-channel; ECC present only on the EPYC
• PCIe: Gen 3 versus Gen 4 with 8 CPU lanes
• Integrated graphics: none versus Iris Xe Graphics 80EU
• Multiplier: unlocked on the EPYC, locked on the Core i5
• Transistors and die size recorded for the EPYC (4,800 million, 213 mm²) but not for the Core i5

FAQ

Q: Which CPU is faster in Cinebench?

A: The AMD EPYC 7261 wins all six recorded tests. Its largest margin is 1.6 percent in Cinebench R20 and R23 multi-core (3979 versus 3918, and 9476 versus 9329); its smallest is 1.4 percent in R20 single-core (561 versus 553). The differences are marginal in practical terms.

Q: How do the two compare against other CPUs in the database?

A: Both sit at the 50th percentile against all CPUs. The EPYC's average score of 2740 places it within 0.7 percent of rivals like the Xeon E-2176M (2751) and Ryzen 3 4300GE (2759), while the Core i5's 2698 puts it within 0.4 percent of the Core i7-8700T (2701) and Xeon E-2226G (2709).

Q: Does the power difference matter?

A: Yes, it is the defining contrast. The Core i5-1345UE has a 15 W TDP while the EPYC 7261 is rated at 170 W, yet the EPYC's benchmark lead never exceeds 1.6 percent. The Core i5 delivers essentially equivalent render performance at a small fraction of the power budget.

Q: Which chip supports ECC memory?

A: Only the EPYC 7261. It supports DDR4 with ECC across an eight-channel bus rated at 170.6 GB/s. The Core i5-1345UE lacks ECC support but offers broader memory compatibility, handling both DDR4 and DDR5 on a dual-channel bus.

Q: Do either of these CPUs have integrated graphics?

A: The Core i5-1345UE includes Iris Xe Graphics 80EU. The EPYC 7261 has no integrated graphics and requires a discrete GPU for display output.

Q: Can either CPU be overclocked?

A: The EPYC 7261 has an unlocked multiplier, permitting adjustment. The Core i5-1345UE's multiplier is locked, and it is a soldered BGA 1744 part, so it cannot be swapped or tuned in the same way.