AMD EPYC 7713P vs Intel Core Ultra 7 165U Comparison
AMD EPYC 7713P
Core Ultra 7 165U
PERFORMANCE BENCHMARKS
Analysis: AMD EPYC 7713P vs Intel Core Ultra 7 165U
The Intel Core Ultra 7 165U and the AMD EPYC 7713P sit at opposite ends of the computing spectrum: one is a 15-watt mobile processor designed for thin-and-light laptops, the other a 225-watt server chip built for rack-mounted workloads. The benchmark data reflects that chasm, with the EPYC taking all six head-to-head victories, but the magnitudes of those victories tell a nuanced story about what each chip is optimized to do.
Head-to-Head Benchmarks
The AMD EPYC 7713P dominates every single benchmark in the comparison set, but the margins vary dramatically by workload type. In Cinebench R23 multi-core, the EPYC scores 69,229 against the Ultra 7’s 10,226 — a delta of -85.2% from the Intel part’s perspective. That means the EPYC delivers roughly 6.8 times the multi-threaded rendering throughput, a gap driven by its 64 cores versus the Ultra 7’s 12.
Single-core results are closer, but still decisively in AMD’s favor. In Cinebench R23 single-core, the EPYC scores 9,773 versus the Ultra 7’s 1,662, a -83% delta. The EPYC’s 2.00 GHz base clock and 3.68 GHz boost clock are both lower than the Ultra 7’s 1.70 GHz base and 4.90 GHz boost, yet the Zen 3 architecture’s higher instructions-per-clock still wins by a wide margin. This is not a close contest in any metric.
The Cinebench R20 results show similar patterns. Multi-core: EPYC 29,076 versus Ultra 7 5,943 (-79.6%). Single-core: EPYC 4,104 versus Ultra 7 839 (-79.6%). The R15 results are equally lopsided: multi-core 6,978 versus 1,605 (-77%), single-core 984 versus 231 (-76.5%). In every case, the EPYC’s advantage exceeds 75%, meaning the Intel chip trails by at least a factor of four in raw performance.
However, the average benchmark score paints a different picture. The Ultra 7 165U has an avgBenchmarkScore of 20,249, while the EPYC 7713P sits at 20,024 — the Intel part is actually 1.1% higher on this aggregate metric. This is because the average includes the Ultra 7’s PassMark scores, which cover tasks like data compression, encryption, and integer math where the mobile chip’s efficiency-oriented design competes more effectively. The head-to-head table only includes Cinebench tests, which heavily favor the server processor’s core count.
Architecture Differences
The architectural divide is stark. The Ultra 7 165U uses Intel’s Meteor Lake architecture, built on a 7 nm process at Intel’s foundry. It packs 12 cores and 14 threads, with a cache hierarchy of 112 KB L1 per core, 2 MB L2 per core, and 12 MB shared L3. The EPYC 7713P uses AMD’s Zen 3 architecture, codenamed Milan, also on a 7 nm node but fabricated by TSMC. It scales to 64 cores and 128 threads, with 64 KB L1 per core, 512 KB L2 per core, and a massive 256 MB shared L3 cache.
The transistor counts and die sizes highlight the scale difference. The EPYC packs 33,200 million transistors across 8 dies of 81 mm² each, while the Ultra 7’s transistor count and die size are not listed in the data. The EPYC’s 256 MB L3 is 21 times larger than the Ultra 7’s 12 MB, a critical advantage for server workloads that repeatedly access large datasets.
Memory support diverges completely. The Ultra 7 uses DDR5 with dual-channel memory and 89.6 GB/s bandwidth; the EPYC uses DDR4 with eight-channel memory and 204.8 GB/s bandwidth — more than double the throughput. The EPYC also supports ECC memory, while the Ultra 7 does not. PCIe connectivity similarly favors the server chip: the EPYC offers 128 Gen 4 lanes (CPU only) versus the Ultra 7’s 12 Gen 4 lanes.
Integrated graphics are exclusive to the Intel part, featuring Arc Xe-LPG 64EU. The EPYC has no integrated graphics, relying on discrete GPUs or no display output at all. The sockets are incompatible by design: Intel BGA 2049 for mobile, AMD Socket SP3 for servers.
Where Each One Wins
The EPYC 7713P wins every Cinebench test, making it the clear choice for CPU-bound rendering, simulation, and any workload that scales with core count. Its 64 cores and 128 threads deliver 6.8x the multi-core R23 performance, and its 256 MB L3 cache reduces memory latency for large working sets. The eight-channel DDR4 memory interface provides 204.8 GB/s of bandwidth, which is essential for database operations, virtualization hosts, and scientific computing. The 128 PCIe Gen 4 lanes allow massive I/O expansion for storage arrays and accelerators.
The Ultra 7 165U wins on efficiency and portability, though not on any listed benchmark. Its 15-watt TDP versus the EPYC’s 225-watt TDP means it can operate in fanless or low-noise designs, while the EPYC requires substantial cooling infrastructure. The integrated Arc Xe-LPG graphics provide display output and basic GPU acceleration without a discrete card, which is impossible on the EPYC. The Ultra 7’s higher boost clock of 4.90 GHz helps in bursty, lightly threaded tasks, but the benchmark data shows even single-core performance favors the EPYC by 76-83%.
The PassMark scores for the Ultra 7 (not compared head-to-head) show strengths in specific tasks: data compression at 188,105, integer math at 57,785, and floating-point math at 40,775. These are respectable mobile results, but they do not approach server-class throughput. The Ultra 7’s percentileVsAllCpus is 74, identical to the EPYC’s 74, meaning both chips outperform 74% of all CPUs in the database — a reminder that the Ultra 7 is a capable mobile part despite losing to a server flagship.
Specification Differences
The key differences are as follows: cores (12 vs 64), threads (14 vs 128), base clock (1.70 GHz vs 2.00 GHz), boost clock (4.90 GHz vs 3.68 GHz), TDP (15 W vs 225 W), socket (Intel BGA 2049 vs AMD Socket SP3), architecture (Meteor Lake vs Zen 3), codename (Meteor Lake vs Milan), generation (Ultra 7 (Meteor Lake) vs EPYC (Zen 3 (Milan))), process node (7 nm Intel vs 7 nm TSMC), foundry (Intel vs TSMC), L1 cache (112 KB per core vs 64 KB per core), L2 cache (2 MB per core vs 512 KB per core), L3 cache (12 MB shared vs 256 MB shared), memory support (DDR5 vs DDR4), memory bus (dual-channel vs eight-channel), memory bandwidth (89.6 GB/s vs 204.8 GB/s), ECC memory (false vs true), PCIe (Gen 4, 12 lanes vs Gen 4, 128 lanes), integrated graphics (Arc Xe-LPG 64EU vs none), market segment (Mobile vs Server/Workstation), release date (2023-12-13 vs 2021-03-14), part number (SRN6D vs 100-000000337100-000000337WOF), and launch MSRP ($448 vs $5010). The EPYC has more transistors (33,200 million) and a larger die (8x 81 mm²), while the Ultra 7’s transistor count and die size are absent from the data. The Ultra 7 has a higher boost clock, but the EPYC has a higher base clock.
FAQ
Q: Which processor is faster in multi-core Cinebench R23?
A: The AMD EPYC 7713P scores 69,229 versus the Intel Core Ultra 7 165U’s 10,226, a delta of -85.2% from the Intel part’s perspective.
Q: Does the Intel chip win any benchmark in the head-to-head table?
A: No. The EPYC 7713P wins all six head-to-head tests, covering Cinebench R15, R20, and R23 in both single-core and multi-core modes.
Q: What is the memory bandwidth difference?
A: The EPYC 7713P provides 204.8 GB/s via eight-channel DDR4, while the Ultra 7 165U provides 89.6 GB/s via dual-channel DDR5.
Q: Does the Ultra 7 have integrated graphics?
A: Yes, the Ultra 7 165U features Arc Xe-LPG 64EU integrated graphics. The EPYC 7713P has no integrated graphics.
Q: What are the TDP ratings?
A: The Ultra 7 165U has a TDP of 15 watts, while the EPYC 7713P has a TDP of 225 watts.
Q: Which chip has more L3 cache?
A: The EPYC 7713P has 256 MB shared L3 cache, while the Ultra 7 165U has 12 MB shared L3 cache.
The Verdict
The AMD EPYC 7713P is the overwhelming performance winner across every measured benchmark. Its 64 cores, 128 threads, and 256 MB L3 cache deliver 6.8x the multi-core rendering performance of the Ultra 7 in Cinebench R23, and it wins single-core tests by 76-83% despite a lower boost clock. For server workloads — virtualization, database processing, HPC simulations, and multi-tenant hosting — the EPYC’s 128 PCIe Gen 4 lanes, eight-channel memory, and ECC support make it the only viable choice in this pairing.
The Intel Core Ultra 7 165U is not a competitor in that arena. Its 12 cores and 14 threads are designed for mobile efficiency, and the 15-watt TDP allows deployment in laptops where the EPYC’s 225-watt TDP would be impossible. The integrated Arc Xe-LPG graphics add display capability that the EPYC lacks entirely. The Ultra 7’s higher boost clock (4.90 GHz) helps in short bursts, but the data shows even single-threaded performance favors the EPYC.
Choose the EPYC 7713P if you need maximum throughput, memory bandwidth, or I/O expansion in a server chassis. Choose the Ultra 7 165U if you need a low-power mobile processor with integrated graphics and can accept a 4-6x performance deficit in multi-core tasks. The average benchmark scores are nearly identical (20,249 vs 20,024), but that metric obscures the fact that the EPYC’s Cinebench results are an order of magnitude higher while the Ultra 7’s PassMark results carry its average — the two chips are optimized for entirely different workloads, and the benchmark data makes that split unambiguous.