Intel Core Ultra 7 265KF vs Intel Core Ultra X9 378H Comparison

FAQ

Q: Which processor has more cores and threads?

A: The Intel Core Ultra 7 265KF has 20 cores and 20 threads, while the Intel Core Ultra X9 378H has 16 cores and 16 threads. Neither processor uses simultaneous multithreading, so thread counts equal core counts for both.

Q: What is the performance gap in single-core workloads?

A: The 265KF leads by 53% in Cinebench R15 single-core (707 vs 462) and by 52.8% in Cinebench R23 single-core (7021 vs 4595). The PassMark single-thread test shows a smaller 10.7% advantage (4928 vs 4453).

Q: Which chip has the larger L3 cache?

A: The 265KF offers 30 MB of shared L3 cache, compared to 18 MB on the 378H. Per-core L2 cache also differs: 3 MB per core on the 265KF versus 2.5 MB per core on the 378H.

Q: What are the memory bandwidth specifications for each?

A: The 265KF supports DDR5 with 102.4 GB/s bandwidth on a dual-channel bus. The 378H supports LPDDR5X with 153.6 GB/s bandwidth, also dual-channel. The mobile chip delivers 50% more theoretical memory bandwidth.

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

A: The 265KF records an average benchmark score of 71910, placing it in the 94th percentile of all CPUs. The 378H averages 47468, in the 89th percentile. The desktop chip exceeds the mobile chip by approximately 51.5% in average score.

Q: Do both processors have integrated graphics?

A: No. The 265KF has no integrated graphics (N/A), while the 378H includes Arc B390 integrated graphics. The desktop chip requires a discrete GPU, whereas the mobile chip has built-in display output capability.

Architecture Differences

The Intel Core Ultra 7 265KF and Intel Core Ultra X9 378H represent different branches of Intel's Core Ultra lineup. The 265KF belongs to Core Ultra Series 2, uses the Arrow Lake architecture with the Arrow Lake-S codename, and is built on a 3 nm process at TSMC. The 378H belongs to Core Ultra Series 3, uses the Panther Lake architecture with the Panther Lake-H codename, and is also built on a 3 nm process but at Intel's own foundry. This foundry split is one of the most significant architectural distinctions between the two.

The 265KF contains 17,800 million transistors on a 243 mm² die. The 378H has no transistor count or die size recorded in the database. The desktop part is a socketed design for Intel Socket 1851, while the mobile part uses Intel BGA 2540, a ball-grid array package. The 265KF has an unlocked multiplier, enabling overclocking; the 378H is locked. The desktop chip's part number is SRQCU, while the 378H's part number is listed as unknown.

Cache hierarchies differ noticeably. Both processors use 192 KB of L1 cache per core. The 265KF provides 3 MB of L2 per core and 30 MB of shared L3. The 378H provides 2.5 MB of L2 per core and 18 MB of shared L3. The larger L3 on the 265KF gives it a meaningful advantage in workloads that benefit from shared cached data across cores.

The memory controllers are distinct. The 265KF supports DDR5 memory with a dual-channel bus and 102.4 GB/s bandwidth. The 378H supports LPDDR5X with 153.6 GB/s bandwidth on a dual-channel bus. Despite the 378H having higher theoretical bandwidth, the 265KF's higher core clocks and greater core count deliver superior measured performance in the benchmark suite.

PCIe lane configurations also differ substantially. The 265KF provides Gen 5 with 20 lanes (CPU only), while the 378H provides Gen 5 with 4 lanes (CPU only). This makes the desktop processor far more suitable for multiple high-bandwidth expansion devices. The 378H compensates with integrated Arc B390 graphics, which the 265KF lacks entirely.

The production status for both is active. The 265KF released on 2024-10-23, and the 378H released on 2026-04-03. The 265KF has a launch MSRP of $379, while the 378H has no recorded MSRP.

Where Each One Wins

The data shows a clean sweep: the Intel Core Ultra 7 265KF wins all 17 head-to-head benchmark comparisons, with the 378H recording zero wins. The margin varies considerably by workload type, which reveals where each processor's design strengths lie.

The 265KF dominates in heavily parallel, multi-core workloads. In Cinebench R23 multi-core, it scores 49736 versus 32553, a 52.8% advantage. The PassMark multi-thread test shows a similar 52.8% gap (58518 vs 38298). These results align with the 265KF's 20-core count, 3.90 GHz base clock, and 5.50 GHz boost clock, versus the 378H's 16 cores, 2.00 GHz base, and 5.00 GHz boost.

The largest single delta appears in PassMark random string sorting, where the 265KF leads by 78.6% (79735 vs 44648). This workload is sensitive to memory latency and cache capacity, areas where the 265KF's 30 MB L3 and higher clock speeds provide an advantage. PassMark extended instructions shows a 74.1% gap (54513 vs 31315), and data compression shows 72.4% (666589 vs 386591). These results indicate that the 265KF's architecture handles instruction-heavy and data-movement tasks with significantly more throughput.

The 378H is relatively closer in single-threaded PassMark tests. The single-thread score of 4453 trails the 265KF's 4928 by only 10.7%. PassMark physics shows the smallest gap at 6.7% (3633 vs 3404). These narrower margins suggest that the 378H's Panther Lake cores are efficient per clock, but the 265KF's higher boost clock and larger cache still keep it ahead.

The 378H's advantages lie outside the benchmark suite. It integrates Arc B390 graphics, supports LPDDR5X with 153.6 GB/s bandwidth, and operates in a 25 W TDP envelope. The 265KF has a 125 W TDP and requires a discrete GPU. For mobile or power-constrained systems, the 378H is the only viable choice. For raw CPU performance, the 265KF wins every recorded test.

Specification Differences

| Specification | Intel Core Ultra 7 265KF | Intel Core Ultra X9 378H |

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

| Series | Core Ultra Series 2 | Core Ultra Series 3 |

| Cores | 20 | 16 |

| Threads | 20 | 16 |

| Base Clock | 3.90 GHz | 2.00 GHz |

| Boost Clock | 5.50 GHz | 5.00 GHz |

| TDP | 125 W | 25 W |

| Socket | Intel Socket 1851 | Intel BGA 2540 |

| Codename | Arrow Lake-S | Panther Lake-H |

| Foundry | TSMC | Intel |

| Process Node | 3 nm | 3 nm |

| Transistors | 17,800 million | Not recorded |

| Die Size | 243 mm² | Not recorded |

| L2 Cache | 3 MB (per core) | 2.5 MB (per core) |

| L3 Cache | 30 MB (shared) | 18 MB (shared) |

| Memory Support | DDR5 | LPDDR5X |

| Memory Bandwidth | 102.4 GB/s | 153.6 GB/s |

| PCIe | Gen 5, 20 Lanes (CPU only) | Gen 5, 4 Lanes (CPU only) |

| Integrated Graphics | N/A | Arc B390 |

| Market Segment | Desktop | Mobile |

| Release Date | 2024-10-23 | 2026-04-03 |

| Launch MSRP | $379 | Not recorded |

| Multiplier Unlocked | Yes | No |

| Part Number | SRQCU | Unknown |

The two processors differ in nearly every major specification category. The 265KF has more cores, higher clocks, a larger L3 cache, more PCIe lanes, and an unlocked multiplier. The 378H has higher memory bandwidth, integrated graphics, a much lower TDP, and a newer release date. Both use 3 nm processes, dual-channel memory buses, and 192 KB L1 per core, and neither supports ECC memory.

Head-to-Head Benchmarks

All 17 recorded head-to-head benchmarks result in a win for the Intel Core Ultra 7 265KF. The margins range from 6.7% to 78.6%, and the pattern of deltas corresponds closely to workload characteristics.

The Cinebench suite shows a consistent 52.8% to 53% advantage for the 265KF across all six tests. Cinebench R15 multi-core scores 5013 versus 3281, a 52.8% delta. Cinebench R15 single-core scores 707 versus 462, a 53% delta. Cinebench R20 multi-core scores 20889 versus 13672, a 52.8% delta. Cinebench R20 single-core scores 2948 versus 1929, also 52.8%. Cinebench R23 multi-core scores 49736 versus 32553, and single-core scores 7021 versus 4595, both 52.8%. The uniformity of these deltas suggests that the 265KF's clock and core advantages translate almost linearly into rendering performance.

PassMark data compression shows a 72.4% gap (666589 vs 386591). This workload is heavily dependent on memory bandwidth and cache efficiency. Despite the 378H's higher theoretical memory bandwidth of 153.6 GB/s, the 265KF's larger L3 cache and higher clocks produce a much faster result. The 378H's LPDDR5X bandwidth does not compensate for its lower core clocks and smaller cache.

PassMark data encryption shows a 61.5% delta (48198 vs 29840). Encryption workloads often scale with core count and clock speed, and the 265KF's 20 cores at 5.50 GHz boost clearly outperform the 378H's 16 cores at 5.00 GHz.

PassMark extended instructions shows a 74.1% gap (54513 vs 31315). This test exercises SIMD and specialized instruction paths. The 265KF's Arrow Lake architecture delivers substantially more throughput in this area.

PassMark find prime numbers shows a 36.1% delta (486 vs 357). This is the smallest gap among the compute-heavy PassMark tests, but the 265KF still holds a decisive lead.

PassMark floating point math shows a 65.4% margin (189431 vs 114500). Floating-point throughput is a core strength of the 265KF's higher-clocked, more numerous cores.

PassMark integer math shows a 54.8% delta (143351 vs 92603). This is close to the Cinebench margins, again reflecting the core and clock advantage.

PassMark multi-thread shows a 52.8% gap (58518 vs 38298), consistent with the Cinebench multi-core results.

PassMark physics shows the narrowest margin at 6.7% (3633 vs 3404). Physics simulations can be sensitive to memory latency, and the 378H's LPDDR5X interface appears to narrow the gap considerably. Still, the 265KF wins.

PassMark random string sorting shows the largest margin at 78.6% (79735 vs 44648). This workload is particularly sensitive to cache capacity and memory access patterns, where the 265KF's 30 MB L3 and desktop memory subsystem excel.

PassMark single-thread and singlethread tests both show a 10.7% delta (4928 vs 4453). The 265KF's 5.50 GHz boost clock provides a modest but consistent single-thread advantage.

The overall picture is unambiguous. The 265KF holds a 52.8% or larger advantage in most multi-core and throughput-oriented tests, with smaller but still positive margins in single-threaded and physics workloads. The 378H's strengths lie in its mobile form factor, integrated graphics, and power efficiency, none of which are captured in these CPU benchmark scores.