Intel Core Ultra 7 265T vs Intel Core Ultra 9 285T Comparison

The Intel Core Ultra 7 265T and Intel Core Ultra 9 285T are two 35-watt desktop processors built on the same Arrow Lake-S architecture, yet benchmark results reveal a clear split in workload specialization. The Ultra 9 285T wins 13 of 17 head-to-head tests, but the Ultra 7 265T takes decisive victories in specific math and instruction-heavy tasks, making the choice between them dependent on the application rather than a simple performance hierarchy.

Where Each One Wins

The Ultra 9 285T dominates the Cinebench suite, taking all six multi-core and single-core tests with narrow margins. In Cinebench R23 multi-core, it scores 31,399 versus the Ultra 7’s 31,312, a 0.3% lead. The pattern repeats in R20 multi-core (13,187 vs 13,151) and R15 multi-core (3,164 vs 3,156). Single-core results are similarly tight, with the Ultra 9 leading by 0.2-0.3% across R15, R20, and R23. These are incremental gains, but they are consistent across every render test.

The Ultra 9 also wins the PassMark multi-thread test (36,940 vs 36,838), integer math (135,890 vs 123,852, an 8.9% advantage), and physics (2,875 vs 2,498, a 13.1% lead). Data compression and encryption also favor the Ultra 9: 328,003 vs 316,131 (3.6% lead) and 28,813 vs 27,943 (3% lead), respectively. Single-thread performance goes to the Ultra 9 as well, with 4,775 vs 4,624 in PassMark’s single-thread test, a 3.2% margin.

The Ultra 7 265T wins four tests, and these are not close calls. Floating-point math shows an 18.7% advantage (153,188 vs 129,032), while find prime numbers is 12.5% faster (359 vs 319). Extended instructions favor the Ultra 7 by 11.4% (24,451 vs 21,956). Random string sorting is effectively a tie, with the Ultra 7 edging out a 0% delta at 41,106 vs 41,091. The Ultra 7’s wins are concentrated in specialized computational workloads, whereas the Ultra 9’s wins span general-purpose and multi-threaded tasks.

Architecture Differences

Both processors share the same fundamental design: Arrow Lake-S architecture on a 3 nm TSMC process, with 17,800 million transistors on a 243 mm² die. They use the Intel Socket 1851, support DDR5 memory over a dual-channel bus with 102.4 GB/s bandwidth, and integrate Arc Xe-LPG Graphics with 64 execution units. PCIe connectivity is identical at Gen 5 with 20 CPU lanes. Neither chip has an unlocked multiplier.

The core counts differ: the Ultra 9 285T has 24 cores and 24 threads, while the Ultra 7 265T has 20 cores and 20 threads. This four-core gap explains the Ultra 9’s consistent, if modest, wins in render tests and multi-threaded workloads. The Ultra 9’s base clock is 1,400 MHz compared to the Ultra 7’s 1,500 MHz, but the Ultra 9 boosts higher to 5.40 GHz versus 5.30 GHz. The Ultra 7’s higher base clock may contribute to its lead in floating-point math and extended instructions.

Cache allocation also diverges. Both have 192 KB of L1 per core and 3 MB of L2 per core, but the shared L3 differs: the Ultra 9 has 36 MB, while the Ultra 7 has 30 MB. The 6 MB L3 advantage for the Ultra 9 likely supports its integer math and data compression wins. ECC memory support is another differentiator: the Ultra 9 supports ECC, while the Ultra 7 does not. Both chips launch with the same release date and remain active in production, with launch MSRPs of $549 for the Ultra 9 and $384 for the Ultra 7.

The Verdict

The data supports the Ultra 9 285T as the more versatile processor for most users. Its 13 wins include every Cinebench test, all major PassMark multi-threaded benchmarks, and the single-thread test. The margins are small in render workloads—never exceeding 0.3%—but the Ultra 9 also delivers larger gains in physics (13.1%) and integer math (8.9%), which are meaningful for simulation and general computation. The 24-core configuration and 36 MB L3 cache give it a structural advantage that the Ultra 7 cannot offset in typical productivity tasks.

The Ultra 7 265T is the pick for specific scientific or numerical workloads. Its 18.7% lead in floating-point math and 12.5% lead in find prime numbers suggest an architectural efficiency that favors certain algorithms. The 11.4% advantage in extended instructions points to better handling of specialized instruction sets. For users whose software relies heavily on these operations, the Ultra 7 delivers superior performance despite having fewer cores and a smaller L3 cache.

Average benchmark scores are nearly identical: the Ultra 7 averages 46,468 versus the Ultra 9’s 46,409, a 0.1% difference that favors the Ultra 7. Both chips sit at the 92nd percentile among all CPUs. The Ultra 7’s nearest rivals include the Ultra 9 285T (0.1% ahead), the AMD Ryzen AI 9 HX 375 (0.3% ahead), and the Intel Core i9-13900HX (0.3% behind). The Ultra 9’s rivals include the Ultra 7 (0.1% behind), the same AMD chip (0.2% ahead), and the i9-13900HX (0.4% behind). The overall scores reveal that these two processors are effectively peers in aggregate performance, with workload-specific differences determining the better choice.

FAQ

Q: Which processor has more cores?

A: The Intel Core Ultra 9 285T has 24 cores and 24 threads, while the Intel Core Ultra 7 265T has 20 cores and 20 threads.

Q: How much faster is the Ultra 7 in floating-point math?

A: The Ultra 7 scores 153,188 in PassMark floating-point math versus the Ultra 9’s 129,032, an 18.7% advantage.

Q: Does the Ultra 9 support ECC memory?

A: Yes, the Ultra 9 285T has ECC memory support, while the Ultra 7 265T does not.

Q: What is the L3 cache difference between the two?

A: The Ultra 9 has 36 MB of shared L3 cache, while the Ultra 7 has 30 MB, a 6 MB difference.

Q: Which chip wins the PassMark single-thread test?

A: The Ultra 9 wins with a score of 4,775 versus the Ultra 7’s 4,624, a 3.2% lead.

Q: Are both processors on the same manufacturing process?

A: Yes, both use a 3 nm process from TSMC, with 17,800 million transistors on a 243 mm² die.

Head-to-Head Benchmarks

The Cinebench suite shows a consistent pattern: the Ultra 9 wins every test by 0.2-0.3%. In R15 multi-core, the Ultra 9 scores 3,164 against 3,156, a 0.3% lead. R20 multi-core is 13,187 vs 13,151 (0.3%), and R23 multi-core is 31,399 vs 31,312 (0.3%). Single-core results are equally tight: R15 at 446 vs 445 (0.2%), R20 at 1,861 vs 1,856 (0.3%), and R23 at 4,432 vs 4,420 (0.3%). These margins are within run-to-run variance for many systems, but the Ultra 9’s consistency across all six tests indicates a real advantage.

PassMark’s multi-thread test gives the Ultra 9 a 0.3% win (36,940 vs 36,838), mirroring the Cinebench pattern. Integer math is a more substantial win: the Ultra 9 scores 135,890 versus 123,852, an 8.9% advantage. Physics shows the largest gap in the Ultra 9’s favor at 13.1% (2,875 vs 2,498). Data compression (328,003 vs 316,131, 3.6%) and encryption (28,813 vs 27,943, 3%) round out the Ultra 9’s wins. Single-thread performance is also in the Ultra 9’s column: 4,775 vs 4,624, a 3.2% margin, recorded identically in both PassMark single-thread and singlethread entries.

The Ultra 7’s wins are concentrated and dramatic. Floating-point math is its largest victory: 153,188 vs 129,032, an 18.7% lead. Find prime numbers follows at 12.5% (359 vs 319). Extended instructions show an 11.4% advantage (24,451 vs 21,956). Random string sorting is a statistical tie: 41,106 vs 41,091, with a deltaPct of 0. These results suggest the Ultra 7’s 20-core configuration with a 1,500 MHz base clock executes certain numerical workloads more efficiently than the 24-core Ultra 9, which has a lower 1,400 MHz base clock. The Ultra 9’s higher boost clock of 5.40 GHz does not translate into wins in these specific tests, indicating that the Ultra 7’s advantage comes from architectural scheduling or clock behavior under load rather than raw core count.