Intel Core 9 273PQE vs Intel Core Ultra 5 235A Comparison
Intel Core 9 273PQE
Core Ultra 5 235A
PERFORMANCE BENCHMARKS
Analysis: Intel Core 9 273PQE vs Intel Core Ultra 5 235A
Head-to-Head Benchmarks
The benchmark data shows a decisive overall victory for the Intel Core 9 273PQE, which wins 15 of the 17 recorded comparisons. The margins, however, vary dramatically by workload, ranging from a near-tie in single-threaded PassMark tests to an 85.8% blowout in integer math.
In the Cinebench suite, the Core 9 273PQE is consistently 20.1% ahead of the Core Ultra 5 235A across every version and workload. The R15 multicore score of 3950 versus 3289, the R20 multicore score of 16459 versus 13705, and the R23 multicore score of 39190 versus 32633 all follow the same pattern. Single-core Cinebench results tell the same story: the Core 9 leads by 20% in R15 (557 versus 464), R20 (2323 versus 1934), and R23 (5532 versus 4607). This consistency across all three Cinebench versions indicates a fundamental performance advantage that scales with the workload rather than a benchmark-specific quirk.
The PassMark multithread test confirms the trend with another 20.1% lead (46107 versus 38392). The largest single delta appears in PassMark integer math, where the Core 9 273PQE scores 164629 against 88626 for the Core Ultra 5 235A, a staggering 85.8% advantage. Data compression also favors the Core 9 heavily at 585752 versus 393800, a 48.7% gap. Extended instructions show a 22.5% lead (38743 versus 31625), while floating-point math is closer at 5.7% (125546 versus 118778). Physics (2754 versus 2437, a 13% lead) and random string sorting (53167 versus 49489, a 7.4% lead) both go to the Core 9 as well.
The Core Ultra 5 235A secures two wins, both in specialized PassMark tests. Data encryption shows a slim 1.7% advantage (30136 versus 29636). The more substantial victory comes in the find prime numbers test, where the Core Ultra 5 scores 392 against the Core 9's 198, a 49.5% lead. That result is notable because it is the only test where the Core 9 falls below 50% of the opponent's score. The single-thread PassMark test is practically a tie: 4573 versus 4557, a 0.4% edge for the Core 9.
Looking at overall averages, the Core 9 273PQE records an average benchmark score of 66099, placing it in the 93rd percentile of all CPUs in the database. Its nearest rival is the Intel Core Ultra 5 250KF Plus at 66159, a 0.1% difference. The Core Ultra 5 235A averages 48201, in the 90th percentile, with its nearest rival the AMD Ryzen AI Max PRO 380 at 48171, also a 0.1% gap. The 66099 versus 48201 average puts the Core 9 roughly 37% ahead in aggregate, though the individual deltas show this is not a uniform advantage across all instruction types.
The Verdict
The data supports a clear split. The Intel Core 9 273PQE is the stronger processor for threaded workloads, integer-heavy computation, compression, and any task that benefits from its 24 threads and high boost clock. Its 20.1% lead across every Cinebench test and the 85.8% integer math advantage make it the obvious choice for rendering, compilation, and general productivity where parallel execution dominates.
The Intel Core Ultra 5 235A wins in only two recorded tests: prime number calculation by 49.5% and data encryption by 1.7%. The prime number result is substantial, and the encryption win, while small, is still a measurable edge. Neither of these victories suggests an overall performance win, but they do indicate that the Ultra 5 has specific strengths in certain algorithmic patterns.
For users whose workloads resemble Cinebench, PassMark multithread, data compression, or extended instructions, the recorded data points squarely to the Core 9 273PQE. The Core Ultra 5 235A remains competitive in single-thread PassMark tests, where the 0.4% delta is effectively a tie, but in every other category except the two noted, the Core 9 holds a meaningful lead. The 90th percentile ranking of the Ultra 5 versus the 93rd percentile of the Core 9 also reflects this hierarchy in the broader database.
Architecture Differences
The two processors come from different Intel design generations. The Core 9 273PQE uses the Bartlett Lake codename and is built on Intel's 10 nm process at Intel's own foundry. The Core Ultra 5 235A belongs to the Core Ultra Series 2, uses the Arrow Lake architecture with the Arrow Lake-S codename, and is manufactured by TSMC on a 3 nm node. The Ultra 5 also has a listed transistor count of 17,800 million and a die size of 243 mm², while the Core 9 has no transistor or die size figures recorded in the database.
Threading philosophy differs sharply. The Core 9 273PQE has 12 cores and 24 threads, indicating simultaneous multithreading. The Core Ultra 5 235A has 14 cores but only 14 threads, meaning it does not use multithreading per core. This explains why the Core 9, despite having fewer physical cores, wins the multithreaded tests by roughly 20%: its 24 threads outperform the Ultra 5's 14 threads in parallel workloads.
Cache hierarchies also diverge. The Core 9 provides 80 KB of L1 per core, 2 MB of L2 per core, and 36 MB of shared L3. The Ultra 5 offers 192 KB of L1 per core, 3 MB of L2 per core, but only 24 MB of shared L3. The Ultra 5 has more per-core L1 and L2, while the Core 9 has 50% more shared L3 capacity.
Memory support differs as well. The Core 9 supports both DDR4 and DDR5, while the Ultra 5 supports DDR5 only. Both use dual-channel memory buses, but the recorded bandwidth is 89.6 GB/s for the Core 9 and 102.4 GB/s for the Ultra 5. ECC memory is supported on the Core 9 but not on the Ultra 5. PCIe connectivity also varies: the Core 9 provides Gen 5 with 16 CPU lanes, while the Ultra 5 provides Gen 5 with 20 CPU lanes. Integrated graphics are different as well, with the Core 9 using UHD Graphics 770 and the Ultra 5 using Arc Xe-LPG Graphics 24EU.
Specification Differences
The socket is a primary divider: the Core 9 273PQE uses Intel Socket 1700, while the Core Ultra 5 235A uses Intel Socket 1851. Base clocks are identical at 3.40 GHz for both. Boost clocks differ, with the Core 9 reaching 5.90 GHz against 5.00 GHz for the Ultra 5. Thermal design power is 125 W for the Core 9 and 65 W for the Ultra 5. Neither processor has an unlocked multiplier.
The Core 9 has 12 cores and 24 threads; the Ultra 5 has 14 cores and 14 threads. L1 is 80 KB per core versus 192 KB per core, L2 is 2 MB per core versus 3 MB per core, and L3 is 36 MB shared versus 24 MB shared. The Core 9 supports DDR4 and DDR5, the Ultra 5 only DDR5. Memory bandwidth is 89.6 GB/s against 102.4 GB/s. ECC is present on the Core 9, absent on the Ultra 5. PCIe lanes are 16 versus 20, both Gen 5. Integrated graphics are UHD Graphics 770 versus Arc Xe-LPG Graphics 24EU. The Core 9 has a recorded launch MSRP of $589; the Ultra 5 has a recorded launch MSRP of $269. Release dates are March 2026 for the Core 9 and July 2025 for the Ultra 5. Both are active production parts for the desktop market segment.
FAQ
Q: Which processor has more cores?
A: The Intel Core Ultra 5 235A has 14 cores, while the Intel Core 9 273PQE has 12 cores. However, the Core 9 has 24 threads to the Ultra 5's 14 threads.
Q: Why does the Core 9 win most multithreaded tests despite having fewer cores?
A: The Core 9 273PQE supports simultaneous multithreading, giving it 24 threads from 12 cores. The Core Ultra 5 235A has 14 threads from 14 cores. In the Cinebench R23 multicore test, the Core 9 scores 39190 against 32633, a 20.1% advantage.
Q: In which tests does the Core Ultra 5 235A beat the Core 9 273PQE?
A: The Core Ultra 5 235A wins the PassMark find prime numbers test (392 versus 198, a 49.5% lead) and the PassMark data encryption test (30136 versus 29636, a 1.7% lead).
Q: How close are the two processors in single-threaded performance?
A: In the PassMark single-thread test, the Core 9 scores 4573 and the Ultra 5 scores 4557, a 0.4% difference. In Cinebench R23 single-core, the Core 9 leads by 20.1% with 5532 versus 4607.
Q: What is the average benchmark score difference?
A: The Core 9 273PQE has an average benchmark score of 66099, and the Core Ultra 5 235A has an average of 48201. The Core 9 sits in the 93rd percentile of all CPUs, the Ultra 5 in the 90th.
Q: Do both processors support ECC memory?
A: No. The Core 9 273PQE supports ECC memory. The Core Ultra 5 235A does not.
Where Each One Wins
The Intel Core 9 273PQE dominates in rendering and threaded productivity. Every Cinebench test, from R15 to R23, single-core and multi-core, shows a 20% or 20.1% lead. PassMark multithread confirms the pattern at 46107 versus 38392. Data compression is a major win at 585752 versus 393800, a 48.7% gap that suggests strong performance for archiving and file-handling workloads. Integer math is the standout category at 85.8% ahead, making the Core 9 well suited to compute-heavy tasks that rely on integer operations. Extended instructions (22.5% lead), floating-point math (5.7% lead), physics (13% lead), and random string sorting (7.4% lead) all add to the Core 9's profile as the more capable all-around processor in this comparison.
The Intel Core Ultra 5 235A wins in two specific categories. The prime number test, where it scores 392 versus 198, indicates an advantage in workloads that involve primality testing or similar algorithmic patterns. The data encryption result, 30136 versus 29636, is narrow but consistent. The Ultra 5 also holds advantages that are not reflected in benchmark wins: a lower 65 W TDP against 125 W, higher memory bandwidth at 102.4 GB/s against 89.6 GB/s, more per-core L1 and L2 cache, and 20 PCIe Gen 5 lanes against 16. It also uses a 3 nm TSMC process, which suggests a more modern fabrication node than the 10 nm Intel process used by the Core 9, though the benchmark data does not translate that into a performance win.
For single-thread PassMark workloads, the two are effectively tied at 4573 versus 4557. The Core 9's 0.4% edge there is negligible. The Core 9's larger 36 MB L3 cache versus 24 MB, its higher 5.90 GHz boost clock, and its 24 threads give it the structural advantages that show up across most of the recorded tests. The Ultra 5's advantages in per-core cache and bandwidth do not overcome the thread count disparity in the measured workloads.