Intel Core i5-13600K vs Intel Core Ultra 9 285H Comparison

The Intel Core Ultra 9 285H and the Intel Core i5-13600K are two very different takes on similar performance. One is a mobile part from Core Ultra Series 2, built on Arrow Lake-H and soldered to a BGA package; the other is a 13th Gen Raptor Lake-S desktop chip on Socket 1700. The database places both in the 86th percentile against all recorded CPUs, with average benchmark scores of 38312 for the Ultra 9 285H and 37680 for the i5-13600K, a gap small enough that the head-to-head record (12 wins for the i5-13600K against 7 for the Ultra 9 285H) is the more telling statistic.

Head-to-Head Benchmarks

The multi-core rendering suites favor the desktop chip consistently. In Cinebench R23 multi-core the i5-13600K scores 24221 against 20781.5 for the Ultra 9 285H, a 14.2 percent advantage, and the pattern repeats across Cinebench R15 multi-core (3642 vs 3177.5, the i5 ahead by 12.8 percent) and Cinebench R20 multi-core (13373 vs 12201, an 8.8 percent gap). Geekbench multi-core also goes to the i5-13600K, 15575 to 14743, a 5.3 percent win, as does PassMark multithread, 37680 to 34171, 9.3 percent apart. The i5-13600K reaches these numbers with 14 cores and 20 threads versus the 16 cores and 16 threads of the Ultra 9 285H, so hyperthreading on the Raptor Lake part is doing real work in heavily threaded loads.

Single-core results split by test suite. Geekbench and Cinebench R20 single-core go to the i5-13600K (2297 vs 2178, and 1887 vs 1722, gaps of 5.2 and 8.7 percent), while the Ultra 9 285H takes Cinebench R15 single-core (313 vs 287.5, 8.9 percent ahead) and Cinebench R23 single-core (2129.5 vs 2000.5, 6.4 percent ahead). PassMark single thread also goes to the mobile chip, 4415 to 4124, a 7.1 percent lead. These splits suggest the Arrow Lake-H cores hold higher per-thread throughput in sustained scalar work, aided by a 5.40 GHz boost clock against 5.10 GHz for the i5-13600K, even though the i5 counters with a higher 3.50 GHz base versus 2.90 GHz.

The PassMark subtests produce the most lopsided results in either direction. The largest win for the i5-13600K is integer math: 122481 versus 85922, a 29.8 percent margin. Data compression is nearly as dramatic, 476394 versus 335859, the i5 ahead by 29.5 percent, and random string sorting goes to the i5 by 19.9 percent (51073 vs 40931). The Ultra 9 285H answers with an even bigger single result in find prime numbers: 330 versus 155, a 112.9 percent lead, more than double the desktop chip's score. It also wins floating point math by 20.6 percent (109190 vs 90538), PassMark physics by 11.3 percent (2513 vs 2258), extended instructions by 7 percent margin in the i5's favor at 28805 vs 26794 being the exception here, and data encryption narrowly to the i5 at 27075 versus 26140, a 3.5 percent gap. Correcting the sequence: extended instructions belongs to the i5-13600K column, while physics and floating point math belong to the Ultra 9 285H.

The 3DMark CPU thread-scaling suite exists only for the i5-13600K in this dataset, from 1085 in single thread up to 10157 at maximum threads, with intermediate scores of 2168, 4283, 7074 and 9368 at 2, 4, 8 and 16 threads. There are no comparable entries for the Ultra 9 285H, so no cross-chip conclusion can be drawn from that suite.

FAQ

Q: Which CPU is faster overall?

A: The i5-13600K wins 12 of 19 head-to-head tests, but the average benchmark scores are nearly tied: 38312 for the Ultra 9 285H versus 37680 for the i5-13600K. The i5 dominates multi-core throughput; the Ultra 9 285H wins several single-thread and math-focused tests.

Q: Which one has better single-core performance?

A: It depends on the test. The Ultra 9 285H leads in Cinebench R15 single-core (313 vs 287.5), Cinebench R23 single-core (2129.5 vs 2000.5) and PassMark single thread (4415 vs 4124). The i5-13600K leads in Cinebench R20 single-core (1887 vs 1722) and Geekbench single-core (2297 vs 2178).

Q: How power-hungry are these chips?

A: Their TDP ratings differ enormously: 45 W for the Ultra 9 285H and 125 W for the i5-13600K. That gap explains why the mobile part trails in multi-core tests despite competitive per-thread results.

Q: Which has the better integrated graphics?

A: The Ultra 9 285H ships with Arc Graphics 140T, the i5-13600K with UHD Graphics 770. The database records the names but no graphics benchmark scores for this comparison, so only the distinction in class of integrated GPU can be stated.

Q: Can these chips be overclocked?

A: Only the i5-13600K. Its multiplier is unlocked; the Ultra 9 285H's is not.

Q: What memory does each support?

A: The Ultra 9 285H supports DDR5 and LPDDR5X with a recorded bandwidth of 102.4 GB/s on a dual-channel bus. The i5-13600K supports DDR4 and DDR5, also dual-channel, with no bandwidth figure recorded.

Architecture Differences

These are two Intel architectures separated by a full platform generation. The Ultra 9 285H is Arrow Lake-H, the top tier of Core Ultra Series 2, manufactured by TSMC on a 3 nm process. The i5-13600K is Raptor Lake-S from the 13th Gen Core family, manufactured by Intel on a 10 nm process, with a recorded die size of 257 mm² (no die size is recorded for the Arrow Lake part).

Core topology differs in philosophy. The Ultra 9 285H has 16 cores and 16 threads, one thread per core. The i5-13600K has 14 cores and 20 threads, using simultaneous multithreading to add logical cores. Cache per core is where the mobile chip pulls ahead: 192 KB of L1 and 3 MB of L2 per core, against 80 KB of L1 and 2 MB of L2 per core on the i5. Both share an identical 24 MB of L3 cache.

Platform integration also diverges. The Ultra 9 285H is a mobile BGA 2049 part, soldered and non-removable, with Gen 5 PCIe limited to 8 CPU lanes. The i5-13600K uses the socketed Intel Socket 1700 and offers 16 Gen 5 CPU lanes, double the direct-attach connectivity, which matters for multi-device storage and GPU configurations. Both support ECC memory.

Specification Differences

The fields where these two CPUs differ, based on the recorded data:

• Cores and threads: 16C/16T (Ultra 9 285H) vs 14C/20T (i5-13600K)
• Base clock: 2.90 GHz vs 3.50 GHz
• Boost clock: 5.40 GHz vs 5.10 GHz
• TDP: 45 W vs 125 W
• Socket: Intel BGA 2049 vs Intel Socket 1700
• Architecture: Arrow Lake (Arrow Lake-H) vs Raptor Lake (Raptor Lake-S)
• Series: Core Ultra Series 2 vs Core 13th Gen
• Process node: 3 nm (TSMC) vs 10 nm (Intel)
• L1 cache: 192 KB per core vs 80 KB per core
• L2 cache: 3 MB per core vs 2 MB per core
• Memory: DDR5 and LPDDR5X vs DDR4 and DDR5; the mobile part records 102.4 GB/s bandwidth
• PCIe: Gen 5, 8 CPU lanes vs Gen 5, 16 CPU lanes
• Integrated graphics: Arc Graphics 140T vs UHD Graphics 770
• Market segment: Mobile vs Desktop
• Release date: January 2025 vs September 2022
• Launch MSRP: $651 vs $319
• Multiplier: locked vs unlocked
• Part number: SRQAL vs SRMBD

The shared specifications are worth noting: 24 MB of shared L3, dual-channel memory buses, ECC support, active production status, and the same 86th percentile ranking against all CPUs in the database.

Where Each One Wins

The i5-13600K wins wherever sustained all-core throughput dominates. All four multi-core suite scores go its way, Cinebench R15, R20, R23 and Geekbench, by margins from 5.3 to 14.2 percent, alongside PassMark multithread. Its strongest subtests are integer math (29.8 percent ahead), data compression (29.5 percent ahead) and string sorting (19.9 percent ahead), which points to compilation, archive handling and general throughput workloads. The unlocked multiplier on a socketed platform further reinforces its position as the tuner's choice.

The Ultra 9 285H wins in three distinct areas. First, several single-thread tests: Cinebench R15 and R23 single-core plus PassMark single thread, by 6.4 to 8.9 percent. Second, raw compute: floating point math by 20.6 percent and find prime numbers by 112.9 percent, results that favor scientific and mathematical workloads. Third, PassMark physics by 11.3 percent, relevant to simulation-heavy tasks. It delivers this while drawing a rated 45 W rather than 125 W, and its nearest rivals, the Intel Core 9 270H at an average score of 38335, the Intel Core i5-13600HX at 38261, the Intel Xeon w3-2525 at 38392 and the AMD Ryzen 7 250 at 38221, all sit within a fraction of a percent of its 38312 average, confirming it competes at the top of the mobile tier.

The Verdict

The data supports a clear split. For a desktop workstation where multi-core rendering, integer-heavy processing and compression matter most, the i5-13600K is the stronger performer, winning the threaded suites by 5.3 to 14.2 percent and the integer and compression subtests by roughly 30 percent. Its socketed platform, doubled PCIe lane count, DDR4 compatibility option and unlocked multiplier add flexibility the BGA-packaged mobile part cannot match.

The Ultra 9 285H is the pick where per-thread responsiveness, floating point work and efficiency converge, which is the defining character of a premium mobile platform: a 45 W TDP, LPDDR5X support, 3 MB of L2 per core and wins in PassMark single thread, physics, floating point math and prime number computation. Its average database score of 38312 actually exceeds the i5-13600K's 37680 despite the threaded deficits, and both rank in the 86th percentile. Choose the i5-13600K for maximum sustained throughput on the desktop; choose the Ultra 9 285H for near-equal overall performance in a power-efficient mobile form factor.