Intel Xeon 6741P vs Intel Xeon 6745P Comparison
Intel Xeon 6741P
Xeon 6745P
PERFORMANCE BENCHMARKS
Analysis: Intel Xeon 6741P vs Intel Xeon 6745P
The Intel Xeon 6741P and the Intel Xeon 6745P are both Granite Rapids server processors on Intel Socket 4710, built on the same 5 nm process with a 300 W TDP. They share the same eight-channel DDR5 memory interface with 409.6 GB/s bandwidth and ECC support. The recorded benchmark data shows a clear split: the 6741P wins 14 of the 16 head-to-head tests, while the 6745P takes the remaining two. The results point to the 6741P as the multi-threaded leader and the 6745P as the single-thread specialist.
The Verdict
The data is unambiguous. The Intel Xeon 6741P is the processor to choose for any workload that scales across many cores. Its 48 cores and 96 threads give it a 19.5% advantage in every Cinebench test recorded, from R15 multicore (8624 vs 7214) to R23 multicore (85561 vs 71578). The same edge appears in Passmark multithread (100660 vs 84210), also at 19.5%. For heavily parallel tasks, the 6741P is the stronger part.
The Intel Xeon 6745P, with 32 cores and 64 threads, counters with higher clock speeds: a 3.10 GHz base and 4.30 GHz boost versus the 6741P's 2.50 GHz and 3.80 GHz. That translates into a single-thread win. In Passmark single-thread and singlethread tests, the 6745P scores 3450 against the 6741P's 3195, a 7.4% advantage. For lightly threaded applications, the 6745P is the better pick.
The overall average benchmark scores reinforce this. The 6741P averages 194901 points, placing it in the 99th percentile of all CPUs. The 6745P averages 154858 points, in the 98th percentile. The 6741P also sits close to its nearest rivals: 0.3% ahead of the AMD EPYC 9335 and 0.7% ahead of the Intel Xeon 678X. The 6745P, in contrast, trails the Intel Xeon 676X by 2.3% and the AMD EPYC 9355P by 3.4%. So the 6741P is not only faster in most tests, it also ranks higher in the overall database.
FAQ
Q: Which processor has more cores?
A: The Intel Xeon 6741P has 48 cores and 96 threads, while the Intel Xeon 6745P has 32 cores and 64 threads.
Q: Which processor has the higher clock speed?
A: The Intel Xeon 6745P has a base clock of 3.10 GHz and a boost clock of 4.30 GHz. The Intel Xeon 6741P runs at 2.50 GHz base and 3.80 GHz boost.
Q: Which processor wins in multi-threaded benchmarks?
A: The 6741P wins all recorded multi-threaded Cinebench tests by 19.5% and Passmark multithread by the same margin.
Q: Which processor wins in single-threaded tests?
A: The 6745P wins both Passmark single-thread tests, scoring 3450 versus 3195, a 7.4% advantage.
Q: How does L3 cache size differ?
A: The 6745P has 336 MB of shared L3 cache, while the 6741P has 288 MB.
Q: Which processor has more PCIe lanes?
A: The 6741P provides 136 Gen 5 lanes (CPU only), while the 6745P provides 88 Gen 5 lanes.
Architecture Differences
Both processors share the same underlying architecture: Granite Rapids, built on Intel's 5 nm process with a die size of 2x 598 mm². They use the same socket (Intel Socket 4710) and belong to the same Xeon 6 (Granite Rapids-SP) generation. The L1 cache is 112 KB per core and the L2 cache is 2 MB per core on both parts. The key architectural divergence is in core count and the shared L3 cache: the 6741P has 48 cores and 288 MB of L3, while the 6745P has 32 cores and 336 MB of L3. The 6745P packs more L3 per core (10.5 MB per core versus 6 MB per core), but the 6741P has more total cores. The memory subsystem is identical: DDR5, eight-channel, 409.6 GB/s bandwidth, and ECC support. Both have no integrated graphics and are aimed at the server/workstation segment. Production status is active for both, with the same release date of 2025-02-23.
Specification Differences
The following specifications differ between the two processors:
- Core count: 48 (6741P) vs 32 (6745P)
- Thread count: 96 vs 64
- Base clock: 2.50 GHz vs 3.10 GHz
- Boost clock: 3.80 GHz vs 4.30 GHz
- L3 cache: 288 MB shared vs 336 MB shared
- PCIe lanes: 136 Gen 5 lanes (CPU only) vs 88 Gen 5 lanes
- Part number: SRVEY vs SRWPAQ7L9
- Launch MSRP: $4421 for the 6741P; $5250 for the 6745P
All other recorded specifications are identical: TDP 300 W, same socket, same memory bus, same bandwidth, same ECC support, same process node, and same die size. The 6745P carries a higher launch MSRP despite having fewer cores, likely reflecting its higher clock speeds and larger L3.
Head-to-Head Benchmarks
The 6741P dominates the head-to-head results. In Cinebench, the 6741P leads by 19.5% in every recorded test: R15 multicore (8624 vs 7214), R15 singlecore (1217 vs 1018), R20 multicore (35935 vs 30062), R20 singlecore (5073 vs 4244), and R23 multicore (85561 vs 71578). The same 19.5% margin appears in Passmark multithread (100660 vs 84210).
The gap widens in specific Passmark workloads. The 6741P is 34.3% ahead in data compression (1816408 vs 1352801), 34.6% ahead in data encryption (89746 vs 66665), and 31.7% ahead in extended instructions (142682 vs 108326). Integer math shows a 36% lead (458058 vs 336926), and floating-point math a 34% lead (358423 vs 267438). The largest single margin is in Passmark physics, where the 6741P scores 13890 against 6144, a 126.1% advantage. The find-prime-numbers test shows an 82.4% gap (1242 vs 681), and random string sorting is 32.8% ahead (177322 vs 133528).
The only two wins for the 6745P are the single-thread tests. In Passmark single-thread and Passmark singlethread, it scores 3450 versus 3195, a 7.4% improvement. These are the sole benchmarks where the 6745P's higher clock speeds translate into a measurable lead.
Where Each One Wins
The 6741P is the clear choice for multi-threaded and throughput-oriented workloads. Its 48 cores drive a 19.5% advantage across the Cinebench suite, which reflects sustained all-core rendering. The Passmark results show even larger leads in tasks that scale with core count: data compression, encryption, extended instructions, integer math, floating-point math, physics, and prime number calculation. The physics test, with a 126.1% advantage, is the most extreme example. For any application that can use more than a few cores, the 6741P is the better processor.
The 6745P wins only in single-threaded tests. Its higher base and boost clocks (3.10 GHz and 4.30 GHz) give it a 7.4% lead in Passmark single-thread and singlethread. For lightly threaded applications, such as legacy software or single-threaded database queries, the 6745P is the stronger performer. Its larger L3 cache (336 MB vs 288 MB) does not appear to translate into wins in any recorded benchmark, but it may still benefit workloads that are not represented in the database.
In summary, the benchmark data leaves no ambiguity: the 6741P is the multi-threaded champion, the 6745P is the single-thread specialist. The 6741P also ranks higher in overall average score and percentile, making it the more balanced choice for most server workloads. The 6745P should only be considered when the workload is strictly single-threaded and the higher clock speed is the deciding factor.