Intel Xeon 6517P vs Intel Xeon 6724P Comparison

The Intel Xeon 6724P and Intel Xeon 6517P are two Granite Rapids server processors that share the same core count and memory architecture, but their benchmark results reveal distinct strengths. The 6724P wins 14 of 17 head-to-head comparisons, while the 6517P takes 3, primarily in single-threaded PassMark tests and data compression. Despite the 6724P’s overall dominance, the average benchmark scores are nearly identical — 72,396 versus 72,350, a delta of just 0.1% — meaning the choice between them depends heavily on workload, not raw aggregate performance.

Where Each One Wins

The Intel Xeon 6724P is the clear winner in compute-heavy and multi-threaded workloads. Its advantages are most pronounced in Cinebench suites: it leads by 3% or 3.1% across R15, R20, and R23, both single-core and multi-core. In PassMark integer math, it scores 172,216 against 162,671, a 5.9% edge, and in floating-point math it leads by 6.2% (135,404 vs. 127,497). The largest margin for the 6724P comes in PassMark physics, where it scores 5,004 against 4,452 — a 12.4% advantage. It also leads in prime number finding by 11% (372 vs. 335) and data encryption by 6% (34,332 vs. 32,385). Extended instruction workloads favor the 6724P by 4.3%, and multithreaded PassMark shows a 3.1% lead.

The Intel Xeon 6517P wins in a narrower set of scenarios. Its most notable victory is in PassMark data compression, scoring 653,338 against 627,185 — a 4% advantage. It also edges out the 6724P in single-threaded PassMark tests, scoring 3,311 versus 3,279, a 1% difference. This single-thread win is consistent across both the “single_thread” and “singlethread” entries, which are identical. The 6517P also wins in random string sorting by 1.5% (67,480 vs. 68,477 is not a win; the actual delta is -1.5% for the 6724P, so the 6517P wins with the higher score of 67,480 versus 68,477? Wait, the data shows the 6517P scored 67,480 and the 6724P scored 68,477, so the 6724P wins that test. The 6517P’s wins are data compression and the two single-thread PassMark entries, totaling 3 wins.)

For workloads that prioritize compression throughput, the 6517P is the better pick. For anything involving physics simulation, prime number calculation, encryption, or general math operations, the 6724P provides a measurable advantage. The 6517P’s single-thread PassMark lead is small but present, suggesting it may handle lightly threaded tasks with slightly better efficiency.

Architecture Differences

Both processors are built on the same Granite Rapids architecture, using Intel’s 5 nm process node and the Intel Socket 4710. They are identical in core count, with 16 cores and 32 threads each, and share the same cache hierarchy: 112 KB of L1 per core, 2 MB of L2 per core, and 72 MB of shared L3. Memory support is also the same — DDR5 with an eight-channel bus and 409.6 GB/s of bandwidth, plus ECC support. PCIe connectivity matches at Gen 5 with 88 lanes (CPU only). Neither has integrated graphics, and both are unlocked multipliers.

The key architectural differences are clock speeds and power. The 6724P has a base clock of 3.60 GHz and a boost clock of 4.30 GHz, while the 6517P runs at 3.20 GHz base and 4.20 GHz boost. This 0.40 GHz base and 0.10 GHz boost advantage for the 6724P explains most of its benchmark lead. The 6724P also has a higher TDP of 210 watts versus 190 watts for the 6517P, allowing it to sustain higher clocks under load. The L3 cache is identical at 72 MB, so the performance gap is not due to cache size.

The 6724P has a higher launch MSRP of $3,622, while the 6517P is listed at $1,195. Both were released on the same date and are marked as active production. The part numbers differ (SRVUA for the 6724P, SRVU4 for the 6517P), but there are no other architectural distinctions in the data.

FAQ

Q: Which processor has a higher base clock speed?

A: The Intel Xeon 6724P has a base clock of 3.60 GHz, while the Intel Xeon 6517P has a base clock of 3.20 GHz.

Q: Are the cache sizes identical between the two?

A: Yes, both have 112 KB of L1 per core, 2 MB of L2 per core, and 72 MB of shared L3 cache.

Q: Which processor wins in Cinebench R23 multi-core?

A: The Intel Xeon 6724P scores 43,643, which is 3% ahead of the 6517P’s 42,352.

Q: Does the 6517P win any benchmarks?

A: Yes, it wins PassMark data compression (653,338 vs. 627,185) and PassMark single-thread tests (3,311 vs. 3,279).

Q: What is the TDP difference?

A: The 6724P has a TDP of 210 watts, while the 6517P has a TDP of 190 watts.

Q: Do both processors support the same memory?

A: Yes, both support DDR5 with an eight-channel bus and 409.6 GB/s bandwidth, plus ECC memory.

Specification Differences

The two processors differ in the following fields: base clock (3.60 GHz vs. 3.20 GHz), boost clock (4.30 GHz vs. 4.20 GHz), TDP (210 vs. 190 watts), launch MSRP ($3,622 vs. $1,195), and part number (SRVUA vs. SRVU4). All other specifications — cores, threads, cache, memory, PCIe, socket, architecture, process node, foundry, integrated graphics, market segment, production status, release date, multiplier unlock status, and ECC support — are identical.

Head-to-Head Benchmarks

The biggest win for the Intel Xeon 6724P is in PassMark physics, where it scores 5,004 against 4,452 for the 6517P, a 12.4% advantage. This is followed by PassMark find prime numbers, where the 6724P leads by 11% (372 vs. 335). In PassMark floating-point math, the 6724P wins by 6.2% (135,404 vs. 127,497), and in data encryption, it wins by 6% (34,332 vs. 32,385). Integer math shows a 5.9% lead for the 6724P (172,216 vs. 162,671), and extended instructions favor the 6724P by 4.3% (54,101 vs. 51,891). The Cinebench R15, R20, and R23 multi-core and single-core tests all show the 6724P ahead by either 3% or 3.1%, with the multi-core scores being 4,399 vs. 4,268 (R15), 18,330 vs. 17,787 (R20), and 43,643 vs. 42,352 (R23). PassMark multithread also goes to the 6724P by 3.1% (51,345 vs. 49,786), and random string sorting is a narrow 1.5% win for the 6724P (68,477 vs. 67,480).

The Intel Xeon 6517P’s largest win is PassMark data compression, where it scores 653,338 against 627,185, a 4% advantage. Its other two wins are the PassMark single-thread tests, where it scores 3,311 versus 3,279 for the 6724P, a 1% margin. These are the only benchmarks where the 6517P comes out ahead, and all three wins are in PassMark tests rather than Cinebench.

The Verdict

The data clearly shows that the Intel Xeon 6724P is the stronger processor for most compute tasks. Its higher base and boost clocks (3.60 GHz and 4.30 GHz versus 3.20 GHz and 4.20 GHz) translate into consistent 3% to 12.4% wins across Cinebench and PassMark math, physics, and encryption workloads. The 12.4% physics advantage and 11% prime number advantage are particularly notable, suggesting the 6724P is better suited for simulation and scientific computing. Its 210-watt TDP, compared to 190 watts for the 6517P, indicates it is designed to sustain higher performance under sustained load.

However, the Intel Xeon 6517P is not without merit. Its 4% lead in data compression makes it the better choice for storage and archival workloads where compression throughput is critical. Its 1% single-thread PassMark lead, while small, suggests it may handle lightly threaded tasks with slightly better efficiency, despite its lower clock speeds. The near-identical average benchmark scores (72,350 vs. 72,396) reflect that the 6517P’s compression win and single-thread edge nearly offset the 6724P’s multi-thread dominance in aggregate.

For users prioritizing raw compute performance — physics, math, encryption, or multi-threaded rendering — the Intel Xeon 6724P is the clear pick, and the 3% Cinebench leads are consistent across all versions. For users focused on data compression or lightly threaded workloads, the Intel Xeon 6517P offers a specific advantage, and its lower launch MSRP of $1,195 versus $3,622 makes it a distinct option in that niche. The 6724P wins 14 of 17 benchmarks, but the 6517P’s wins are meaningful in their respective domains.