AMD EPYC 9454 vs Intel Core Ultra 5 236V Comparison

Head-to-Head Benchmarks

The recorded data shows a complete sweep for the AMD EPYC 9454 across all six shared Cinebench tests. The Intel Core Ultra 5 236V does not post a single winning score in any head-to-head comparison, with the EPYC 9454 taking all six victories. The margin is uniform at 78.7% in every test, which indicates a consistent and overwhelming performance gap rather than a workload-specific advantage.

In Cinebench R23 multi-core, the EPYC 9454 scores 73,375 against the Core Ultra 5 236V's 15,628. That is a 78.7% deficit for the Intel part, meaning the AMD processor delivers roughly 4.7 times the multi-threaded rendering throughput. The single-core R23 result tells a similar story, with the EPYC 9454 at 10,358 versus 2,206 for the Core Ultra 5 236V, again a 78.7% difference. What stands out here is that the EPYC 9454 wins by the same proportional margin in single-threaded tests as it does in multi-threaded tests, which is unusual for a high-core-count server chip.

Cinebench R20 multi-core shows the EPYC 9454 at 30,817 and the Core Ultra 5 236V at 6,563. The R20 single-core test has the EPYC 9454 at 4,350 and the Intel part at 926. In Cinebench R15, the EPYC 9454 scores 7,396 multi-core and 1,044 single-core, while the Core Ultra 5 236V manages 1,575 and 222 respectively. Every delta is identical at 78.7%, which suggests the benchmark suite is measuring a fundamental per-thread advantage for the AMD architecture, not just a core-count scaling effect.

The average benchmark score in the database places the Core Ultra 5 236V at 21,952, against 21,223 for the EPYC 9454. This is a narrow 0.2% difference in favor of the Intel chip, despite the EPYC 9454 dominating every Cinebench test. The explanation lies in the benchmark mix: the Core Ultra 5 236V has additional PassMark scores for data compression, encryption, floating-point math, integer math, and physics, while the EPYC 9454 only has Cinebench entries. The Intel part's PassMark single-thread score of 3,893 and multi-thread score of 18,375 contribute to its higher average, but those tests do not appear in the head-to-head comparison.

The nearest rivals in the database place the Core Ultra 5 236V within 0.2% of the AMD EPYC 9534, which scores 21,900. The EPYC 9454 sits within 0.2% of the Intel Core Ultra 7 155U, which scores 21,174. This means the two processors occupy nearly identical positions in the overall percentile ranking, both at the 75th percentile of all CPUs. The percentile alignment is notable given the massive Cinebench discrepancy, reinforcing that the average score metric is heavily influenced by the test suite composition.

The Verdict

The data presents two processors with opposite design philosophies. The Core Ultra 5 236V is a mobile part with 8 cores and 8 threads, a 17 W TDP, and integrated Arc 130V graphics. The EPYC 9454 is a server/workstation processor with 48 cores and 96 threads, a 290 W TDP, and no integrated graphics. The choice between them depends entirely on the workload and platform constraints.

For single-threaded responsiveness, the EPYC 9454 wins every Cinebench single-core test by 78.7%. This is counterintuitive for a server chip, but the data is unambiguous: the EPYC 9454's base clock of 2.75 GHz and boost clock of 3.80 GHz, combined with Zen 4 architecture, produce higher single-thread scores than the Core Ultra 5 236V's 2.10 GHz base and 4.70 GHz boost. The Intel part's higher boost clock does not translate into a single-thread win in any recorded test.

For multi-threaded workloads, the EPYC 9454's advantage is equally decisive. With 48 cores and 96 threads against 8 cores and 8 threads, the AMD processor delivers roughly 4.7 times the Cinebench R23 multi-core score. The 256 MB shared L3 cache versus 8 MB shared L3 cache, and the twelve-channel memory bus with 460.8 GB/s bandwidth versus dual-channel memory, support this scaling in memory-intensive parallel tasks.

The Core Ultra 5 236V does have a few qualitative advantages. It supports integrated graphics, which the EPYC 9454 lacks. It also has a much lower TDP at 17 W versus 290 W, making it suitable for fanless or low-power mobile designs. The Intel part uses a 3 nm process node from TSMC, while the EPYC 9454 uses a 5 nm process node from the same foundry. However, neither of these advantages appears in the benchmark scores.

The database places both processors at the 75th percentile, which means the average score metric does not differentiate them meaningfully. The EPYC 9454 is the clear choice for server workloads, rendering farms, or any application that uses more than 8 threads. The Core Ultra 5 236V is the only option for mobile platforms, given its BGA 2833 socket and integrated graphics. There is no scenario in the recorded data where the Core Ultra 5 236V outperforms the EPYC 9454 on a shared benchmark.

Architecture Differences

The two processors come from entirely different architectural lineages. The Core Ultra 5 236V uses Lunar Lake, the second generation of Intel's Core Ultra Series 2, built on a 3 nm process node at TSMC. The EPYC 9454 uses Zen 4 under the Genoa codename, built on a 5 nm process node at the same foundry. Both use TSMC, but the Intel part is a mobile-oriented design while the AMD part targets server and workstation deployments.

The Core Ultra 5 236V has 8 cores and 8 threads, with no hyperthreading. The EPYC 9454 has 48 cores and 96 threads, meaning each core supports two threads. The core count difference is 6x, and the thread count difference is 12x. This alone explains the multi-threaded benchmark gap, but the single-threaded gap requires a different explanation.

Cache hierarchies differ substantially. The Core Ultra 5 236V has 192 KB of L1 cache per core, 2.5 MB of L2 cache per core, and 8 MB of shared L3 cache. The EPYC 9454 has 64 KB of L1 cache per core, 1 MB of L2 cache per core, and 256 MB of shared L3 cache. The EPYC's L3 cache is 32 times larger, which benefits workloads with large working sets that exceed the Intel part's 8 MB capacity.

The EPYC 9454 has 52,560 million transistors across 8 dies, each measuring 72 mm². The Core Ultra 5 236V does not have recorded transistor or die size data in the database. The EPYC's multi-die design with eight chiplets is typical of server processors, while the Core Ultra 5 236V uses a monolithic design for mobile power efficiency.

Memory support differs at a fundamental level. The Core Ultra 5 236V's memory support is listed as dependent on the motherboard, with a dual-channel memory bus. The EPYC 9454 supports DDR5 natively, uses a twelve-channel memory bus, and has a recorded memory bandwidth of 460.8 GB/s. The EPYC also supports ECC memory, while the Core Ultra 5 236V does not.

PCIe connectivity is another differentiator. The Core Ultra 5 236V provides Gen 5 with 4 lanes from the CPU. The EPYC 9454 provides Gen 5 with 128 lanes from the CPU, a 32x difference in available lanes. This makes the EPYC suitable for large GPU arrays, NVMe storage farms, and high-speed networking, none of which the Core Ultra 5 236V can support at scale.

Specification Differences

The two processors differ on nearly every specification in the database. The Core Ultra 5 236V has 8 cores and 8 threads, while the EPYC 9454 has 48 cores and 96 threads. The base clock is 2.10 GHz for the Intel part and 2.75 GHz for the AMD part. The boost clock is 4.70 GHz for the Intel part and 3.80 GHz for the AMD part. The Intel part boosts higher but starts lower, while the AMD part has a higher floor and a lower ceiling.

TDP is 17 W for the Core Ultra 5 236V and 290 W for the EPYC 9454. The socket is Intel BGA 2833 for the Intel part and AMD Socket SP5 for the AMD part. The process node is 3 nm for the Intel part and 5 nm for the AMD part, both from TSMC. The EPYC 9454 has recorded transistor count of 52,560 million and a die size of 8x 72 mm², while the Core Ultra 5 236V has no recorded values for either.

L1 cache is 192 KB per core for the Intel part and 64 KB per core for the AMD part. L2 cache is 2.5 MB per core for the Intel part and 1 MB per core for the AMD part. L3 cache is 8 MB shared for the Intel part and 256 MB shared for the AMD part. The memory bus is dual-channel for the Intel part and twelve-channel for the AMD part. Memory bandwidth is not recorded for the Intel part but is 460.8 GB/s for the AMD part.

ECC memory support is false for the Core Ultra 5 236V and true for the EPYC 9454. PCIe is Gen 5 with 4 lanes for the Intel part and Gen 5 with 128 lanes for the AMD part. The Core Ultra 5 236V has integrated Arc 130V graphics, while the EPYC 9454 has no integrated graphics. The market segment is Mobile for the Intel part and Server/Workstation for the AMD part. Release dates differ, with the Core Ultra 5 236V released on 2024-09-23 and the EPYC 9454 on 2022-11-09. The EPYC 9454 has a launch MSRP of $5225, while the Core Ultra 5 236V has no recorded launch MSRP.

FAQ

Q: Which processor has more cores and threads?

A: The AMD EPYC 9454 has 48 cores and 96 threads, while the Intel Core Ultra 5 236V has 8 cores and 8 threads.

Q: Does the Intel Core Ultra 5 236V win any head-to-head benchmark against the AMD EPYC 9454?

A: No. The EPYC 9454 wins all six shared Cinebench tests, with a 78.7% margin in every case.

Q: What is the single-threaded performance difference in Cinebench R23?

A: The EPYC 9454 scores 10,358 in Cinebench R23 single-core, while the Core Ultra 5 236V scores 2,206, a 78.7% advantage for the AMD part.

Q: Which processor supports ECC memory?

A: The AMD EPYC 9454 supports ECC memory. The Intel Core Ultra 5 236V does not support ECC memory.

Q: What is the memory bandwidth of each processor?

A: The EPYC 9454 has a recorded memory bandwidth of 460.8 GB/s. The Core Ultra 5 236V has no recorded memory bandwidth in the database.

Q: How do the average benchmark scores compare?

A: The Core Ultra 5 236V has an average benchmark score of 21,952, while the EPYC 9454 has an average of 21,223, a difference of 0.2% in favor of the Intel part. Both are at the 75th percentile of all CPUs.