Intel Core Ultra 5 235H vs Intel Core Ultra 9 285 Comparison

Intel Core Ultra 5 235H and Intel Core Ultra 9 285 are both part of the Core Ultra Series 2, but they target very different segments of the market. The 235H is a mobile processor designed for laptops, while the 285 is a desktop part built for high-performance systems. The benchmark data in the database shows a clear performance hierarchy, with the 285 winning all 17 recorded head-to-head tests. However, the margins vary significantly by workload, which matters for real-world purchasing decisions.

Where Each One Wins

The Intel Core Ultra 9 285 wins every single benchmark in the recorded comparison. There are zero wins for the Intel Core Ultra 5 235H across all 17 tests. This is not a close contest, but the scale of the victory depends heavily on the type of workload.

The largest advantage for the 285 appears in integer math, where it scores 164,869 against 74,247 for the 235H, a 55 percent deficit for the mobile chip. Floating point math shows a similar pattern, with the 285 at 194,988 versus 93,509, a 52 percent gap. These are compute-heavy tasks that scale well with core count and clock speed, and the 285 has both advantages.

The smallest margin is in single-thread performance. The 285 scores 4,881 in PassMark single-thread tests, while the 235H scores 4,359, a difference of only 10.7 percent. This suggests that per-core efficiency is relatively close between the two, and the massive multicore gap comes primarily from having more cores and higher boost clocks.

For memory compression and encryption, the 285 delivers roughly double the throughput. Data compression scores 602,121 versus 301,979, and data encryption scores 46,949 versus 23,121. These workloads benefit from the 285's larger L3 cache and additional cores.

The 235H does have one unique advantage: it is a mobile chip with a 28 watt TDP, versus 65 watts for the 285. This means it can fit into thin and light laptops, which the desktop 285 cannot. In terms of pure performance, the 285 is the definitive winner, but the 235H wins on platform flexibility.

Architecture Differences

Both processors use the Arrow Lake architecture and are built on a 3 nm process node by TSMC. The core design is identical at the microarchitecture level, with 192 KB of L1 cache per core and 3 MB of L2 cache per core. The differences come from the physical configuration and target platform.

The 235H is codenamed Arrow Lake-H and has 14 cores with 14 threads. It uses the Intel BGA 2049 socket, which is a soldered mobile package. Its L3 cache is 18 MB shared. The 285 is codenamed Arrow Lake-S, has 24 cores with 24 threads, and uses the Intel Socket 1851 for desktop. Its L3 cache is 36 MB shared, exactly double the mobile part.

The 285 also carries a discrete die with 17,800 million transistors on a 243 mm² die size. The 235H does not have this data recorded in the database, but the core count difference suggests a smaller physical footprint.

Clock speeds differ substantially. The 235H has a base clock of 2.40 GHz and a boost clock of 5.00 GHz. The 285 starts at 2.50 GHz base and boosts to 5.60 GHz. Even at base clocks, the 285 runs faster, and the boost advantage of 0.60 GHz helps in single-thread workloads.

Memory support diverges. The 235H supports DDR5 and LPDDR5X, while the 285 supports only DDR5. Both run dual-channel with 102.4 GB/s bandwidth. The 235H does not support ECC memory; the 285 does support ECC.

PCIe connectivity is another split. The 235H provides Gen 5 with 8 lanes from the CPU, while the 285 provides Gen 5 with 20 lanes. This gives the desktop chip more room for expansion cards and storage.

Integrated graphics differ as well. The 235H uses Arc Graphics 140T, while the 285 uses Arc Xe-LPG Graphics with 64 execution units. The mobile part's integrated graphics may be more relevant since laptops rarely have discrete GPUs, whereas desktop systems often pair the 285 with a dedicated graphics card.

The 285 has a launch MSRP of $579, but the database does not list a launch price for the 235H. Both processors are currently active in production. The 285 released on 2024-12-31, and the 235H released on 2025-01-12, less than two weeks later.

Head-to-Head Benchmarks

The Cinebench suite shows consistent 47.7 percent deficits for the 235H across all six tests. In R15 multicore, the 285 scores 4,933 versus 2,580 for the 235H. R15 single-core shows 696 versus 364. R20 multicore: 20,556 versus 10,751. R20 single-core: 2,901 versus 1,517. R23 multicore: 48,945 versus 25,598. R23 single-core: 6,909 versus 3,613.

The uniform 47.7 percent delta across both single and multicore Cinebench tests is notable. It suggests that the clock speed advantage of the 285, combined with its higher core count, translates into a proportional improvement in both threaded and unthreaded rendering workloads. The 235H's 14 cores versus 24 cores for the 285 explains part of this, but the single-core tests also show the same gap, indicating that the 285's 5.60 GHz boost clock provides a meaningful per-core edge.

PassMark integer math shows the largest gap at 55 percent. The 285 scores 164,869, and the 235H scores 74,247. This workload is highly parallel and benefits from both core count and memory bandwidth. The 285's larger L3 cache likely reduces memory latency stalls during integer operations.

Floating point math shows a 52 percent gap, with the 285 at 194,988 and the 235H at 93,509. The 285's higher boost clock and additional cores deliver substantial throughput here.

Data compression and encryption both show roughly 50 percent gaps. Compression: 602,121 versus 301,979, a 49.8 percent delta. Encryption: 46,949 versus 23,121, a 50.8 percent delta. These workloads are memory and cache sensitive, and the 285's 36 MB of L3 cache versus 18 MB gives it a clear advantage.

Physics simulation shows a 44.8 percent gap, the smallest among multicore tests. The 285 scores 3,598, and the 235H scores 1,985. This workload likely relies on specific instruction patterns where the core count difference matters less.

Prime number finding shows a 47.9 percent gap, with 459 versus 239. Extended instructions show a 48.5 percent gap, with 45,357 versus 23,354. Random string sorting shows a 50.8 percent gap, with 73,651 versus 36,208.

The smallest gap is in single-thread performance, at 10.7 percent. The 285 scores 4,881, and the 235H scores 4,359. This is the only test where the 285's advantage drops below 40 percent. It indicates that the 235H's 5.00 GHz boost clock is nearly competitive with the 285's 5.60 GHz, and the underlying core architecture performs similarly at equivalent clock speeds.

FAQ

Q: How much faster is the Intel Core Ultra 9 285 in multicore workloads?

A: The 285 leads by 47.7 percent across all Cinebench multicore tests. In PassMark multithread, the 285 scores 56,602 versus 30,091 for the 235H, a 46.8 percent difference. Float point math shows a 52 percent gap, and integer math shows a 55 percent gap.

Q: Is the single-thread performance gap smaller than the multicore gap?

A: Yes. PassMark single-thread shows the 285 at 4,881 and the 235H at 4,359, a 10.7 percent difference. Cinebench R23 single-core shows a 47.7 percent gap, but that is because the 285 scores 6,909 versus 3,613. The PassMark single-thread test reveals that per-core performance is much closer than the Cinebench scores suggest.

Q: Which processor supports ECC memory?

A: The Intel Core Ultra 9 285 supports ECC memory. The Intel Core Ultra 5 235H does not support ECC.

Q: What are the socket requirements for each processor?

A: The 235H uses Intel BGA 2049, a soldered mobile socket. The 285 uses Intel Socket 1851, a desktop LGA socket. They are not interchangeable.

Q: How does the cache configuration differ?

A: Both have 192 KB of L1 per core and 3 MB of L2 per core. The 235H has 18 MB of shared L3 cache, while the 285 has 36 MB of shared L3 cache.

Q: Which processor has more PCIe lanes?

A: The 285 provides Gen 5 with 20 lanes from the CPU. The 235H provides Gen 5 with 8 lanes from the CPU.

The Verdict

The data is unambiguous: the Intel Core Ultra 9 285 outperforms the Intel Core Ultra 5 235H in every recorded benchmark. The 285 wins 17 out of 17 head-to-head tests, and its average benchmark score is 75,488 versus 37,522 for the 235H. The 285 also sits at the 95th percentile of all CPUs in the database, compared to the 85th percentile for the 235H.

For desktop builders who need maximum compute throughput, the 285 is the clear choice. Its 24 cores, 36 MB of L3 cache, and 5.60 GHz boost clock deliver roughly double the performance of the 235H in most multicore workloads. The 47.7 percent lead in Cinebench tests and the 55 percent lead in integer math confirm that this is a substantial upgrade for rendering, compilation, and scientific computing.

The 235H is not a poor processor; it sits at the 85th percentile of all CPUs, and its nearest rivals include the Intel Core i9-13900HK with a 0.3 percent difference and the Intel Core i5-13600K with a 0.4 percent difference. It delivers strong performance for a 28 watt mobile part. However, in direct comparison to the 285, it loses by a wide margin. The 10.7 percent gap in PassMark single-thread shows that the 235H holds its own in lightly threaded tasks, so laptop users who prioritize single-core responsiveness will not be disappointed.

The choice comes down to platform. The 235H is for laptops, and the 285 is for desktops. No one will realistically choose between them for the same build. But the database shows that if you need desktop-level performance, the 285 delivers it decisively over the mobile part.

Specification Differences

| Specification | Intel Core Ultra 5 235H | Intel Core Ultra 9 285 |

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

| Cores | 14 | 24 |

| Threads | 14 | 24 |

| Base Clock | 2.40 GHz | 2.50 GHz |

| Boost Clock | 5.00 GHz | 5.60 GHz |

| TDP | 28 W | 65 W |

| Socket | Intel BGA 2049 | Intel Socket 1851 |

| Architecture | Arrow Lake | Arrow Lake |

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

| Process Node | 3 nm | 3 nm |

| Foundry | TSMC | TSMC |

| Transistors | Not recorded | 17,800 million |

| Die Size | Not recorded | 243 mm² |

| L1 Cache | 192 KB (per core) | 192 KB (per core) |

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

| L3 Cache | 18 MB (shared) | 36 MB (shared) |

| Memory Support | DDR5, LPDDR5X | DDR5 |

| Memory Bus | Dual-channel | Dual-channel |

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

| ECC Memory | No | Yes |

| PCIe | Gen 5, 8 Lanes (CPU only) | Gen 5, 20 Lanes (CPU only) |

| Integrated Graphics | Arc Graphics 140T | Arc Xe-LPG Graphics 64EU |

| Market Segment | Mobile | Desktop |

| Production Status | Active | Active |

| Release Date | 2025-01-12 | 2024-12-31 |

| Launch MSRP | Not available | $579 |

| Multiplier Unlocked | No | No |

| Part Number | SRQAP | SRQD4 |

| Avg Benchmark Score | 37,522 | 75,488 |

| Percentile vs All CPUs | 85 | 95 |