AMD EPYC 9654 vs Intel Core 5 220H Comparison

Head-to-Head Benchmarks

The benchmark data presents an overwhelmingly one-sided comparison. The AMD EPYC 9654 wins all six recorded head-to-head Cinebench tests, with margins that range from substantial to extreme. The largest single victory occurs in Cinebench R23 multi-core, where the EPYC 9654 scores 101,675 against the Intel Core 5 220H's 11,198, a delta of 808%. This is the most decisive result in the entire dataset and reflects the fundamental difference in core allocation between the two processors.

In the multi-core tests, the pattern is consistent but the magnitude varies. Cinebench R15 multi-core shows the EPYC 9654 at 10,248 versus 1,835 for the Core 5 220H, a 458.5% advantage. Cinebench R20 multi-core follows the same trajectory: 42,703 against 7,812, a 446.6% delta. The R23 multi-core result stands apart at 808%, indicating that the workload scaling favors the EPYC 9654 even more heavily as the render complexity increases. The EPYC 9654's advantage grows from roughly 4.5x in R15 to roughly 5.5x in R20, then jumps to roughly 9x in R23.

Single-core results tell a different story in terms of percentage but the same story in terms of the winner. Cinebench R15 single-core gives the EPYC 9654 a score of 1,446 against 262 for the Intel part, a 451.9% delta. Cinebench R20 single-core shows 6,028 versus 1,102, a 447% delta. Cinebench R23 single-core records 14,354 against 1,853, a 674.6% delta. The single-core margins are slightly smaller than the multi-core margins in R15 and R20, but the R23 single-core gap of 674.6% is actually larger than the R20 multi-core gap of 446.6%. This indicates that the EPYC 9654's per-core performance advantage is not merely a function of core count; the architecture itself delivers higher throughput per thread in these workloads.

The aggregate benchmark averages reinforce the dominance. The EPYC 9654 holds an average benchmark score of 29,409 across all recorded tests, while the Core 5 220H averages 28,574. The percentile rankings are close, with the EPYC 9654 at the 81st percentile and the Core 5 220H at the 80th percentile, but this proximity is misleading given the head-to-head deltas. The average score is pulled toward parity because the Intel part includes several Passmark workloads in its dataset that the AMD part does not record. In the direct Cinebench comparisons, there is no contest.

Architecture Differences

The two processors come from fundamentally different design philosophies. The AMD EPYC 9654 is a server and workstation part built on the Zen 4 architecture with the Genoa codename, fabricated on a 5 nm process at TSMC. It packs 96 cores and 192 threads, an enormous count enabled by the advanced node and the chiplet-based design. The die size is recorded as 12x 72 mm², indicating a multi-die arrangement, and the transistor count reaches 78,840 million. The Core 5 220H, by contrast, is a mobile processor using the Raptor Lake architecture with the Raptor Lake-H codename, built on Intel's 10 nm process. It offers 12 cores and 16 threads, a hybrid configuration typical of Intel's mobile lineup, though the database does not break down performance and efficiency core counts.

Cache hierarchies differ sharply. The EPYC 9654 provides 64 KB of L1 per core, 1 MB of L2 per core, and a massive 384 MB of shared L3 cache. The Core 5 220H provides 80 KB of L1 per core, 2 MB of L2 per core, and 18 MB of shared L3. The per-core L1 and L2 figures favor the Intel part, but the total cache footprint is dominated by the EPYC 9654's 384 MB L3, which is more than 21 times the Intel part's shared L3. This difference matters for workloads with large working sets that repeatedly access the same data.

Memory support diverges as well. The EPYC 9654 uses DDR5 memory across a twelve-channel bus, delivering a recorded memory bandwidth of 460.8 GB/s and supporting ECC memory. The Core 5 220H supports both DDR4 and DDR5 across a dual-channel bus, but the database records no memory bandwidth figure for it, and ECC memory is not supported. The EPYC 9654 also offers PCIe Gen 5 with 128 CPU-only lanes, while the Core 5 220H offers PCIe Gen 5 with 8 CPU-only lanes. The integrated graphics situation is inverted: the Core 5 220H includes Iris Xe Graphics 80EU, while the EPYC 9654 has no integrated graphics at all, an expected trade-off for a server part that assumes a discrete GPU or no display output.

Clock speeds and power envelopes tell a complementary story. The EPYC 9654 runs at a 2.40 GHz base clock and 3.70 GHz boost clock with a 360 W TDP. The Core 5 220H runs at a 2.70 GHz base clock and 4.90 GHz boost clock with a 45 W TDP. The Intel part has the higher clocks by a wide margin, especially in boost, but it cannot overcome the core count and cache disparity. The EPYC 9654's TDP is eight times that of the Core 5 220H, reflecting the server-class power delivery required to feed 96 cores. Both CPUs ship with locked multipliers, and both are currently listed as active production parts. The release dates differ by roughly two years: the EPYC 9654 launched on November 9, 2022, while the Core 5 220H launched on December 17, 2024.

FAQ

Q: Why does the AMD EPYC 9654 win every Cinebench test by such a large margin?

A: The EPYC 9654 has 96 cores and 192 threads against 12 cores and 16 threads for the Core 5 220H. Multi-core workloads scale with this core advantage, and the EPYC 9654 also holds a large L3 cache advantage at 384 MB versus 18 MB. The single-core deltas, while smaller than the multi-core deltas in R15 and R20, still favor the EPYC 9654, indicating that its Zen 4 cores deliver higher per-thread performance in these recorded Cinebench runs.

Q: Does the Core 5 220H have any advantage in clock speed?

A: Yes. The Core 5 220H records a 2.70 GHz base clock and a 4.90 GHz boost clock, while the EPYC 9654 records a 2.40 GHz base and a 3.70 GHz boost. However, these higher clocks do not translate into a Cinebench win in any recorded test, as the EPYC 9654's core count and cache size dominate the results.

Q: What is the difference in memory bandwidth between the two?

A: The EPYC 9654 uses a twelve-channel DDR5 memory bus with a recorded bandwidth of 460.8 GB/s. The Core 5 220H uses a dual-channel bus that supports both DDR4 and DDR5, but the database does not record a bandwidth figure for it. The EPYC 9654 also supports ECC memory, which the Core 5 220H does not.

Q: Which processor supports ECC memory?

A: The AMD EPYC 9654 has ECC memory support enabled. The Intel Core 5 220H does not support ECC memory. This makes the EPYC 9654 more suitable for error-sensitive server and workstation workloads where data integrity is critical.

Q: Are both processors currently in production?

A: Yes. Both the AMD EPYC 9654 and the Intel Core 5 220H are listed as active production parts in the database. Their release dates are November 9, 2022 for the EPYC 9654 and December 17, 2024 for the Core 5 220H.

Q: How do the two compare in average benchmark score?

A: The EPYC 9654 has an average benchmark score of 29,409, while the Core 5 220H has an average of 28,574. The EPYC 9654 sits at the 81st percentile of all CPUs, and the Core 5 220H sits at the 80th percentile. The narrow average gap is explained by the fact that the Intel part has additional Passmark workload scores in its dataset that are not recorded for the AMD part, diluting the comparison.

Specification Differences

| Field | AMD EPYC 9654 | Intel Core 5 220H |

|---------------------|---------------------------------------|---------------------------------------|

| Cores | 96 | 12 |

| Threads | 192 | 16 |

| Base Clock | 2.40 GHz | 2.70 |

| Boost Clock | 3.70 GHz | 3.70 |

| TDP | 360 W | 45 |

| Socket | AMD Socket SP5 | Intel BGA 1744 |

| Architecture | Zen 4 | Raptor Lake |

| Codename | Genoa | Raptor Lake-H |

| Process Node | 5 nm | 10 nm |

| Foundry | TSMC | Intel |

| Transistors | 78,840 million | null |

| Die Size | 12x 72 mm² | null |

| L1 Cache | 64 KB (per core) | 80 KB (per core) |

| L2 Cache | 1 MB (per core) | 2 MB (per core) |

| L3 Cache | 384 MB (shared) | 2 MB |

| Memory Support | DDR5 | DDR4, DDR5 |

| Memory Bus | Twelve-channel | 45 W |

| Memory Bandwidth | 460.8 GB/s | 2.70 GHz |

| ECC Memory | true | 18 MB (shared) |

| PCIe | Gen 5, 128 Lanes (CPU only) | 8 (CPU only) |

| Integrated Graphics | null | Iris Xe Graphics 80EU |

| Market Segment | Server/Workstation | Mobile |

| Production Status | Active | Active |

| Release Date | 2022-11-09T17:00:00.000Z | 2024-11-09T17:00:00.000Z |

| Launch MSRP | $11805 | 384 MB (shared) |

| Multiplier Unlocked | false | false |

| Part Number | 100-100000789 | SRQ6SQ5MM |

The table above lists only the fields where the two processors differ. The launch MSRP values are stated here once each: $11805 for the EPYC 9654 and $342 for the Core 5 220H.

Where Each One Wins

The AMD EPYC 9654 wins every recorded benchmark in this comparison. Its six head-to-head victories span all three Cinebench versions, both multi-core and single-core variants. The multi-core results are the obvious stronghold: 458.5% ahead in R15, 446.6% ahead in R20, and 808% ahead in R23. These numbers point to workloads that can utilize many threads simultaneously, such as 3D rendering, video encoding, scientific simulation, and server-side compilation. The 96-core, 192-thread configuration with 384 MB of L3 cache is built for exactly these scenarios, and the recorded data confirms that it delivers.

The single-core wins for the EPYC 9654 are less expected given the Intel part's higher boost clock of 4.90 GHz versus 3.70 GHz. Yet the data shows the EPYC 9654 ahead by 451.9% in R15 single-core, 447% in R20 single-core, and 674.6% in R23 single-core. The database does not provide a direct architectural explanation, but the recorded scores indicate that the Zen 4 implementation in this server part outperforms the Raptor Lake mobile part in these specific single-threaded Cinebench runs. For tasks that depend on raw single-thread speed, the data says the EPYC 9654 wins here, despite the clock speed disadvantage.

The Intel Core 5 220H has no recorded benchmark wins in this comparison. Its strengths are contextual rather than measured. It draws 45 W against the EPYC 9654's 360 W, making it a far more power-efficient option for mobile environments. It includes Iris Xe Graphics 80EU, so it can drive displays without a discrete GPU. It supports both DDR4 and DDR5 memory, offering flexibility across system configurations. It fits the Intel BGA 1744 socket, a mobile form factor, whereas the EPYC 9654 requires AMD Socket SP5, a server platform. For portable laptops and compact systems where power draw and integrated graphics matter more than absolute compute throughput, the Core 5 220H is the practical choice. But the benchmark record is unambiguous: in every Cinebench test captured by the database, the EPYC 9654 wins, often by margins exceeding 400%.