AMD EPYC 9734 vs Intel Xeon 6747P Comparison

Head-to-Head Benchmarks

The recorded data shows a clear split between the AMD EPYC 9734 and the Intel Xeon 6747P, with the AMD part winning 10 of the 14 direct comparisons. The most decisive victories for the EPYC 9734 come in workloads that scale with core count and memory bandwidth. In PassMark integer math, the AMD processor scores 823150 against 468518 for the Intel part, a 75.7% advantage. Data encryption shows an even larger gap in relative terms: 179390 versus 90789, a 97.6% delta. Random string sorting is the single biggest win for AMD, with a 98.3% lead (357638 versus 180382). Data compression follows at 58.2% ahead (2900008 versus 1833378), and floating point math sits at 50.1% ahead (549045 versus 365904). Extended instructions also favor the EPYC 9734, with a 44.5% margin (205925 versus 142557).

The Intel Xeon 6747P wins four comparisons, and they are concentrated in areas that reward higher clock speeds and per-core efficiency. The most striking result is PassMark physics, where Intel scores 13398 against 6747, a 49.6% advantage. Single-thread performance also goes to Intel: 3236 versus 2310, a 28.6% lead. The same margin appears in the duplicate single-thread test. Prime number finding is the fourth Intel win, with 1151 versus 829, a 28% edge. These are not trivial margins; they show that the Intel part has a meaningful per-core advantage in latency-sensitive workloads.

The multi-core rendering tests are remarkably close. In Cinebench R15 multi-core, the EPYC 9734 scores 8763 versus 8712, a 0.6% win. Cinebench R20 multi-core shows 36516 versus 36301, again 0.6% in favor of AMD. Cinebench R23 multi-core repeats the pattern: 86943 versus 86432, a 0.6% lead. PassMark multithread also lands at 0.6% in favor of AMD, with 102286 versus 101685. That consistency across four different multi-threaded tests suggests the two processors deliver nearly identical aggregate throughput in rendering-style workloads, despite very different core counts and clock speeds.

The average benchmark scores tell a different story than the head-to-head deltas. The EPYC 9734 holds an average benchmark score of 310619, while the Xeon 6747P sits at 238263. The AMD part also ranks in the 99th percentile against all CPUs, matching the Intel part's 99th percentile placement. However, the nearest rivals for each CPU show where they sit in the broader landscape. The EPYC 9734 is nearly even with the AMD EPYC 9575F (0.4% behind), trails the Intel Xeon 6781P by 1.6%, and the AMD Ryzen Threadripper PRO 9985WX by 3.2%, while leading the Intel Xeon 6774P by 3.4%. The Xeon 6747P is 2.5% behind the AMD EPYC 9634 and 5.3% behind the Intel Xeon 6980P, while leading the AMD EPYC 9455P by 9.4% and the Intel Xeon w9-3595X by 13.5%. These figures indicate that the EPYC 9734 competes in a higher performance tier overall, while the Xeon 6747P sits in a lower tier despite its strong per-core results.

The Verdict

The data points to different buyers for each processor. The AMD EPYC 9734 is the choice for workloads that can use all available cores and threads. It wins the overwhelming majority of comparisons, and its victories in integer math, encryption, compression, and floating point are not marginal; they are substantial, ranging from roughly 44% to 98%. The near-tie in Cinebench multi-core scores is notable, but the EPYC 9734 still wins those tests by 0.6% each. Its average benchmark score of 310619 is far above the Xeon 6747P's 238263, and it sits close to much more expensive competition like the EPYC 9575F and the Xeon 6781P.

The Intel Xeon 6747P is the pick for workloads that are single-threaded or latency-bound. Its 28.6% lead in single-thread performance and 49.6% lead in physics are decisive. The prime number test also favors Intel by 28%. These are exactly the kinds of workloads where clock speed matters more than core count, and the Xeon 6747P's 3.90 GHz boost clock versus 3.00 GHz on the EPYC 9734 explains the pattern. The Intel part also uses less power on paper, with a 330 W TDP against 340 W for AMD, though the database does not include measured power consumption.

For a mixed workload environment, the tie-breaking data comes from the multithreaded tests. The EPYC 9734 wins all four of them, even if only by 0.6%. That means the AMD part does not sacrifice multi-threaded throughput to achieve its large wins elsewhere. The Intel part cannot say the same; it wins single-threaded tests but loses every multi-threaded comparison. Buyers who cannot predict their workload mix should favor the EPYC 9734, because it wins the broad set of benchmarks and loses only in narrowly defined single-threaded scenarios.

Where Each One Wins

The AMD EPYC 9734 dominates in data-heavy and parallel workloads. PassMark data compression shows a 58.2% lead, data encryption a 97.6% lead, and random string sorting a 98.3% lead. Integer math is 75.7% ahead, floating point math is 50.1% ahead, and extended instructions are 44.5% ahead. These are the workloads that benefit from the EPYC 9734's 112 cores and 224 threads, along with its twelve-channel memory bus and 460.8 GB/s of memory bandwidth. The Cinebench multi-core tests are also AMD wins, though by the narrow 0.6% margin, meaning rendering workloads are essentially a toss-up with a slight AMD tilt. PassMark multithread follows the same pattern.

The Intel Xeon 6747P wins in single-threaded and low-latency scenarios. The PassMark physics test is its best result, with a 49.6% margin. The prime number test shows a 28% lead. Single-thread performance is 28.6% ahead. These wins align with the Intel part's higher base clock of 2.70 GHz and boost clock of 3.90 GHz, compared to 2.20 GHz and 3.00 GHz for the AMD chip. The Intel part also has a larger L3 cache at 288 MB versus 256 MB, though the per-core L1 and L2 caches are also larger on Intel: 112 KB and 2 MB per core, versus 64 KB and 1 MB per core on AMD. Those cache differences may contribute to the Intel wins in latency-sensitive tests.

There is no workload category where both processors are weak. The AMD part is not competitive in single-threaded tests, and the Intel part is not competitive in throughput-heavy tests. The decision comes down to which category matters more for the intended deployment.

FAQ

Q: Which processor has more cores and threads?

A: The AMD EPYC 9734 has 112 cores and 224 threads. The Intel Xeon 6747P has 48 cores and 96 threads.

Q: How large is the single-thread performance gap?

A: The Intel Xeon 6747P leads by 28.6% in PassMark single-thread, scoring 3236 versus 2310 for the AMD EPYC 9734.

Q: Which processor is faster in Cinebench R23 multi-core?

A: The AMD EPYC 9734 wins with a score of 86943 versus 86432 for the Intel Xeon 6747P, a 0.6% margin.

Q: What is the biggest single benchmark win for either processor?

A: The AMD EPYC 9734 wins PassMark random string sorting by 98.3%, scoring 357638 versus 180382 for the Intel Xeon 6747P.

Q: Do both processors support DDR5 memory?

A: Yes, both support DDR5. The AMD EPYC 9734 uses a twelve-channel memory bus with 460.8 GB/s bandwidth, while the Intel Xeon 6747P uses an eight-channel bus with 409.6 GB/s bandwidth.

Q: How do their average benchmark scores compare?

A: The AMD EPYC 9734 has an average benchmark score of 310619, while the Intel Xeon 6747P has an average of 238263. Both rank in the 99th percentile against all CPUs.

Architecture Differences

The two processors come from different design philosophies. The AMD EPYC 9734 uses the Zen 4c architecture under the Bergamo codename, built on a 5 nm process at TSMC. It packs 71,000 million transistors across 8 dies, each measuring 73 mm². The Intel Xeon 6747P uses the Granite Rapids architecture, also on a 5 nm process but fabricated at Intel, with a die size of 2x 598 mm². The transistor count for the Intel part is not recorded in the database.

Cache layouts differ substantially. The AMD part provides 64 KB of L1 per core and 1 MB of L2 per core, with 256 MB of shared L3. The Intel part provides 112 KB of L1 per core and 2 MB of L2 per core, with 288 MB of shared L3. The larger per-core caches on Intel likely contribute to its single-threaded wins, while the larger shared L3 may help in workloads with large working sets.

Memory architecture also differs. The AMD EPYC 9734 uses a twelve-channel memory bus with 460.8 GB/s of bandwidth. The Intel Xeon 6747P uses an eight-channel bus with 409.6 GB/s. Both support DDR5 and ECC memory. The AMD part offers more PCIe lanes: 128 Gen 5 lanes (CPU only) versus 88 Gen 5 lanes (CPU only) on Intel.

The sockets are incompatible. The AMD part fits AMD Socket SP5, while the Intel part fits Intel Socket 4710. The Intel part has no integrated graphics and lists "N/A" for that field, while the AMD part does not record an integrated graphics option either. Both are active production parts for the server and workstation market segment.

Specification Differences

The core and thread counts are the largest specification gap. The AMD EPYC 9734 offers 112 cores and 224 threads, while the Intel Xeon 6747P offers 48 cores and 96 threads. Clock speeds favor Intel: the Xeon 6747P has a base clock of 2.70 GHz and a boost clock of 3.90 GHz, while the EPYC 9734 has a base clock of 2.20 GHz and a boost clock of 3.00 GHz. TDP is close but slightly lower on Intel: 330 W versus 340 W.

Memory bandwidth favors AMD at 460.8 GB/s over Intel's 409.6 GB/s, and the channel count differs at twelve versus eight. PCIe lane counts favor AMD at 128 Gen 5 lanes versus 88 Gen 5 lanes. The L3 cache favors Intel at 288 MB versus 256 MB. The release dates differ, with the AMD part launched on 2023-06-12 and the Intel part on 2025-02-23. The launch MSRP for the AMD EPYC 9734 is $9600, and for the Intel Xeon 6747P it is $6497. The part numbers are 100-000001235 for AMD and SRVEZ for Intel. Neither processor has an unlocked multiplier.