AMD EPYC 9184X vs Intel Core Ultra 7 265K Comparison

The Intel Core Ultra 7 265K and AMD EPYC 9184X occupy different ends of the processor spectrum, yet both sit at the 94th percentile of all CPUs in the database. The Intel part, a 20-core desktop chip on Arrow Lake, wins 12 of 17 head-to-head benchmarks against the AMD part, a 16-core server processor on Zen 4 Genoa-X. The AMD part counters with a 768 MB L3 cache, twelve-channel memory, and wins in physics, integer math, and Cinebench R23. The data shows a clear split: Intel for single-thread and floating-point workloads, AMD for cache-hungry server tasks.

Where Each One Wins

The Intel Core Ultra 7 265K wins 12 of the 17 recorded head-to-head benchmarks. Its wins span single-thread performance, floating-point math, encryption, extended instructions, data compression, and multithreaded Cinebench R15 and R20. The AMD EPYC 9184X wins 5 benchmarks: Cinebench R23 multi-core and single-core, Passmark integer math, Passmark physics, and random string sorting.

For desktop use, the Intel part is the clear choice. It leads by 74.6% in Passmark single-thread (4928 vs 2822) and by 22.9% in Cinebench R15 single-core (708 vs 576). For floating-point math, the Intel part nearly doubles the AMD part, scoring 189629 vs 95476, a 98.6% advantage. Data encryption also favors Intel by 29.1% (48246 vs 37376). The Intel part wins data compression by 8.2% (665554 vs 614873), extended instructions by 24.7% (54333 vs 43562), and Passmark multithread by 22.9% (58594 vs 47665).

The AMD part wins where its massive cache and memory bandwidth matter. It leads Cinebench R23 single-core by 64.7% (5719 vs 2020) and multi-core by 11.5% (40515 vs 35850). Passmark physics shows a 44.1% AMD advantage (6674 vs 3731), and integer math favors AMD by 9% (157483 vs 143242). Random string sorting is nearly tied, with AMD ahead by 0.2% (79913 vs 79752).

So the use-case split is: Intel for general desktop productivity, single-thread responsiveness, and floating-point or encryption-heavy tasks; AMD for server workloads, physics simulation, and integer math.

Architecture Differences

The two processors come from different design philosophies. The Intel Core Ultra 7 265K uses Arrow Lake-S, built on a 3 nm process at TSMC, with 17,800 million transistors on a 243 mm² die. It has 20 cores and 20 threads, meaning no simultaneous multithreading. Each core has 192 KB of L1 and 3 MB of L2, with 30 MB of shared L3. Memory support is DDR5 over a dual-channel bus with 102.4 GB/s of bandwidth. PCIe is Gen 5 with 20 CPU lanes. It includes integrated Arc Xe-LPG Graphics with 64 execution units. The multiplier is unlocked, and the socket is Intel Socket 1851.

The AMD EPYC 9184X uses Zen 4 Genoa-X, built on a 5 nm process at TSMC, with 90,160 million transistors across 8x 72 mm² dies. It has 16 cores and 32 threads, using simultaneous multithreading. Each core has 64 KB of L1 and 1 MB of L2, with a massive 768 MB of shared L3. Memory support is DDR5 over a twelve-channel bus with 460.8 GB/s of bandwidth. PCIe is Gen 5 with 128 CPU lanes. There is no integrated graphics. The multiplier is locked, and the socket is AMD Socket SP5.

The most striking architectural difference is cache and memory. The AMD part carries 768 MB of shared L3 versus 30 MB on the Intel part, and its twelve-channel memory bus delivers 460.8 GB/s versus 102.4 GB/s. The Intel part counters with a higher boost clock of 5.50 GHz versus 4.20 GHz, and a smaller 3 nm process node versus 5 nm. The transistor counts also differ sharply: 17,800 million on Intel versus 90,160 million on AMD, reflecting the AMD part's multiple dies and large cache.

Head-to-Head Benchmarks

The largest Intel win is Passmark floating-point math, where the Core Ultra 7 265K scores 189629 against 95476, a 98.6% advantage. This is nearly double the AMD result. The next biggest Intel win is Passmark single-thread, 4928 vs 2822, a 74.6% lead. Data encryption follows at 29.1% (48246 vs 37376), then extended instructions at 24.7% (54333 vs 43562). The Cinebench R15 and R20 results are consistent: Intel leads multi-core by 22.9% (5020 vs 4083 in R15, 20918 vs 17016 in R20) and single-core by 22.9% to 23% (708 vs 576 in R15, 2953 vs 2401 in R20). Passmark multithread also shows a 22.9% Intel lead (58594 vs 47665). Data compression favors Intel by 8.2% (665554 vs 614873), and find prime numbers by 5.6% (491 vs 465).

The AMD part takes its biggest win in Cinebench R23 single-core, scoring 5719 against 2020, a 64.7% lead. This is a sharp reversal from the R15 and R20 single-core results, where Intel leads by roughly 23%. Passmark physics is the second-largest AMD win at 44.1% (6674 vs 3731). Cinebench R23 multi-core goes to AMD by 11.5% (40515 vs 35850), again reversing the R15 and R20 multi-core results. Integer math favors AMD by 9% (157483 vs 143242), and random string sorting is essentially tied, with AMD ahead by 0.2% (79913 vs 79752).

The pattern is notable: the two Cinebench R23 results contradict the R15 and R20 results. In R15 and R20, Intel wins both single and multi-core by about 23%. In R23, AMD wins single-core by 64.7% and multi-core by 11.5%. The database records both sets of scores, so the verdict depends on which benchmark generation is trusted.

Specification Differences

The two parts differ in nearly every specification field. The Intel Core Ultra 7 265K has 20 cores and 20 threads; the AMD EPYC 9184X has 16 cores and 32 threads. Base clocks are 3.90 GHz for Intel and 3.55 GHz for AMD; boost clocks are 5.50 GHz and 4.20 GHz. TDP is 125 W for Intel and 320 W for AMD. Sockets are Intel Socket 1851 and AMD Socket SP5. Process nodes are 3 nm and 5 nm, both TSMC. Transistor counts are 17,800 million and 90,160 million. Die size is 243 mm² for Intel and 8x 72 mm² for AMD.

Cache differs per core and in total. Intel has 192 KB L1 per core, 3 MB L2 per core, and 30 MB shared L3. AMD has 64 KB L1 per core, 1 MB L2 per core, and 768 MB shared L3. Memory buses are dual-channel for Intel and twelve-channel for AMD, with bandwidth of 102.4 GB/s and 460.8 GB/s. PCIe lanes are 20 for Intel and 128 for AMD, both Gen 5. Intel includes integrated Arc Xe-LPG Graphics with 64 execution units; AMD has none. Intel's multiplier is unlocked; AMD's is locked. Market segments are Desktop and Server/Workstation. Release dates are 2024-10-23 for Intel and 2023-06-12 for AMD. Launch MSRP is $394 for Intel and $4928 for AMD. The Intel part number is SRQCW; the AMD part number is not recorded.

FAQ

Q: Which processor has more cores?

A: The Intel Core Ultra 7 265K has 20 cores, while the AMD EPYC 9184X has 16 cores. However, the AMD part has 32 threads versus 20, because it supports simultaneous multithreading.

Q: Which processor is faster in single-threaded workloads?

A: It depends on the benchmark. The Intel part wins Passmark single-thread by 74.6% (4928 vs 2822) and Cinebench R15 and R20 single-core by about 23%. The AMD part wins Cinebench R23 single-core by 64.7% (5719 vs 2020).

Q: Which processor has more cache?

A: The AMD EPYC 9184X has 768 MB of shared L3 cache, compared to 30 MB on the Intel Core Ultra 7 265K. The Intel part has larger per-core L1 and L2 caches: 192 KB and 3 MB per core versus 64 KB and 1 MB per core.

Q: Which processor supports more memory bandwidth?

A: The AMD EPYC 9184X supports 460.8 GB/s over a twelve-channel DDR5 bus. The Intel Core Ultra 7 265K supports 102.4 GB/s over a dual-channel DDR5 bus.

Q: Does either processor include integrated graphics?

A: Yes, the Intel Core Ultra 7 265K includes Arc Xe-LPG Graphics with 64 execution units. The AMD EPYC 9184X has no integrated graphics.

Q: Which processor is newer?

A: The Intel Core Ultra 7 265K was released on 2024-10-23. The AMD EPYC 9184X was released on 2023-06-12.

The Verdict

The data points to a clear recommendation for each use case. For a desktop build, the Intel Core Ultra 7 265K is the stronger choice. It wins 12 of 17 head-to-head benchmarks, leads by 74.6% in Passmark single-thread, by 98.6% in floating-point math, and by 29.1% in data encryption. Its 125 W TDP and unlocked multiplier suit a desktop environment, and it includes integrated graphics. The launch MSRP of $394 reflects its consumer positioning.

For server or workstation workloads, the AMD EPYC 9184X is the better fit. It wins Cinebench R23 multi-core by 11.5% and single-core by 64.7%, leads Passmark physics by 44.1%, and integer math by 9%. Its 768 MB L3 cache and 460.8 GB/s memory bandwidth are designed for data-heavy server tasks. The 320 W TDP and locked multiplier are typical of a server part, and the launch MSRP of $4928 places it in the enterprise segment.

Both parts sit at the 94th percentile of all CPUs in the database. The Intel part has an average benchmark score of 70879, slightly above its nearest rival, the Intel Core i7-14700KF, by 1%, and slightly below the Intel Xeon 6511P by 0.2%. The AMD part has an average benchmark score of 68202, slightly above its nearest rival, the AMD EPYC 7352, by 0.1%. The choice is not about overall speed but about workload: Intel for desktop responsiveness and floating-point throughput, AMD for cache capacity, memory bandwidth, and server-class physics and integer performance.