AMD Ryzen 9 PRO 9965X3D vs Intel Xeon 6745P Comparison

The Intel Xeon 6745P and AMD Ryzen 9 PRO 9965X3D represent two fundamentally different philosophies for high-performance computing, and the benchmark data starkly illustrates the divide. The Xeon 6745P, with its massive 32-core count and server-oriented design, dominates nearly every multi-threaded workload, while the Ryzen 9 PRO 9965X3D counterpunches with a decisive victory in single-thread performance. The data shows a clear split: 9 benchmark wins for the Intel part versus 2 for the AMD part, but the magnitude and nature of those wins tell a more nuanced story than the raw tally suggests.

Head-to-Head Benchmarks

The most lopsided victory in the entire comparison belongs to the Intel Xeon 6745P in floating-point math. Here, the Intel part scores 267,438 against the AMD's 157,123, a massive 70.2% advantage. This is not a marginal edge; it is a categorical difference that will manifest in any workload relying heavily on scientific computation, financial modeling, or simulation. The gap is so wide that it suggests architectural efficiency differences rather than just core-count scaling, as the Xeon’s 32 cores versus 16 cores alone cannot fully account for such a disparity.

Similarly, extended instructions show a 53% lead for the Intel Xeon 6745P, with scores of 108,326 versus 70,797. This benchmark often reflects encoding, cryptography, and vectorized workloads, and the Intel part’s substantial advantage here reinforces its server-grade positioning. Data encryption follows a similar pattern, with the Xeon 6745P scoring 66,665 versus 43,905, a 51.8% delta. Encryption is a critical server workload, and a lead of this size suggests that the Xeon 6745P is not just faster but disproportionately better suited for secure data processing tasks.

Data compression, another bread-and-butter server workload, shows a 51.1% lead for the Intel part, scoring 1,352,801 versus 895,563. This is one of the largest absolute score differentials in the comparison, and it points to the Xeon’s superior memory bandwidth (409.6 GB/s versus 89.6 GB/s) playing a pivotal role. The 43.8% advantage in random string sorting (133,528 versus 92,886) corroborates this theory, as sorting algorithms are heavily memory-latency and bandwidth sensitive.

The multi-threaded PassMark score gives the Xeon 6745P a 23.4% win, with 84,210 versus 68,225. While this is a significant victory, it is smaller than the core-count ratio of 2:1 might suggest. This implies that the AMD Ryzen 9 PRO 9965X3D’s higher base clock of 4.30 GHz and boost clock of 5.50 GHz are helping to close the gap in some concurrent workloads. The physics benchmark shows a 28% lead for Intel, scoring 6,144 versus 4,801, which again points to the Xeon’s raw compute throughput in structured workloads.

Integer math provides a 41.6% Intel advantage (336,926 versus 237,955), and even find prime numbers, where the Xeon wins 681 versus 599, shows a narrower 13.7% margin. This smaller delta in prime number finding, a workload that often scales with single-thread speed and cache efficiency, hints at the AMD part’s strengths. The single-thread benchmark is where the tables turn completely: the AMD Ryzen 9 PRO 9965X3D scores 4,614 versus the Intel’s 3,450, a 25.2% advantage for AMD. This is a massive single-thread lead, and it is the only category where the AMD part wins decisively, with the single-thread and singlethread entries both reflecting identical scores.

Where Each One Wins

The Intel Xeon 6745P is the undisputed king of throughput-oriented workloads. Every multi-threaded benchmark in the comparison falls to the Intel part, and the margins are often crushing. Data compression, encryption, extended instructions, floating-point math, integer math, multithread, physics, random string sorting, and prime number finding all favor the Xeon. This makes it the clear choice for database servers, virtualization hosts, content delivery networks, and any environment where many parallel processes must run simultaneously without bottlenecks. The 32 cores and 64 threads provide a level of parallelism that the 16-core, 32-thread AMD part cannot match, and the 336 MB of shared L3 cache (compared to 128 MB) further accelerates data-heavy server workloads.

The AMD Ryzen 9 PRO 9965X3D wins where responsiveness and single-thread speed matter most. Its 25.2% lead in single-thread performance is not a trivial advantage; it means that applications reliant on a single fast core—such as legacy software, certain database queries, or lightly threaded engineering tools—will feel noticeably snappier on the AMD platform. The higher clock speeds (4.30 GHz base, 5.50 GHz boost versus 3.10 GHz and 4.30 GHz) are the obvious drivers here. However, it is crucial to note that the AMD part does not win a single multi-threaded benchmark, so its utility is strictly limited to single-thread-dominant scenarios or workloads where its 128 MB L3 cache provides a special advantage not captured by these specific tests.

Architecture Differences

The two processors are built on different manufacturing nodes and foundries, which explains some of their performance characteristics. The Intel Xeon 6745P uses a 5 nm process from Intel’s own fabs, while the AMD Ryzen 9 PRO 9965X3D uses a 4 nm process from TSMC. The AMD part’s smaller node likely contributes to its higher clock speeds and better single-thread efficiency, while the Intel part’s larger node is paired with a much larger die size of 2x 598 mm² compared to AMD’s 2x 70.6 mm². This is a staggering difference in silicon area, with the Intel chip being roughly 8.5 times larger per die.

The core architectures themselves are from different generations. The Intel Xeon 6745P is based on Granite Rapids, part of the Xeon 6 family, while the AMD Ryzen 9 PRO 9965X3D uses the Zen 5 architecture (Granite Ridge) from the Ryzen 9000 series. The Intel part has 32 cores and 64 threads, while the AMD has exactly half of each: 16 cores and 32 threads. Cache hierarchies also diverge significantly. The Intel part has 112 KB of L1 cache per core and 2 MB of L2 per core, while the AMD has 80 KB of L1 and 1 MB of L2 per core. The L3 cache shows the most dramatic difference: Intel offers 336 MB shared, while AMD offers 128 MB. Despite having half the cores, the AMD’s 128 MB L3 is still substantial, but the Intel’s 336 MB is in a different league for server workloads.

Memory architecture is another fundamental split. The Intel Xeon 6745P supports eight-channel DDR5 memory with a theoretical bandwidth of 409.6 GB/s, whereas the AMD Ryzen 9 PRO 9965X3D uses dual-channel DDR5 with only 89.6 GB/s of bandwidth. This 4.6x difference in memory bandwidth is almost certainly a primary reason for the Intel part’s dominance in data-heavy benchmarks like compression and sorting. Both support ECC memory, but the Xeon’s 88 PCIe Gen 5 lanes dwarf the AMD’s 24 lanes, making the Intel part vastly more expandable for storage and accelerators. The AMD part includes integrated Radeon Graphics, while the Intel has no integrated GPU. The sockets are completely different: Intel Socket 4710 versus AMD Socket AM5.

FAQ

Q: Why does the Intel Xeon 6745P win so many benchmarks despite the AMD having a higher clock speed?

A: The Intel part has 32 cores versus 16, and 64 threads versus 32. In multi-threaded tests, the sheer core count allows the Xeon to process more work simultaneously, outweighing the AMD’s clock advantage. Additionally, the Intel’s 336 MB L3 cache and 409.6 GB/s memory bandwidth provide a massive data throughput advantage in memory-intensive workloads.

Q: Is the AMD Ryzen 9 PRO 9965X3D better for any task at all?

A: Yes, the data shows a 25.2% lead in single-thread performance, scoring 4,614 versus 3,450. This makes it superior for software that relies on a single fast core, where the higher boost clock of 5.50 GHz provides a clear edge over the Intel’s 4.30 GHz.

Q: What is the significance of the 70.2% delta in floating-point math?

A: This is the largest performance gap in the comparison. The Intel score of 267,438 versus 157,123 indicates a fundamental architectural advantage in compute-heavy scientific or financial workloads, likely stemming from a combination of more cores, larger caches, and wider memory bandwidth.

Q: How do the memory systems compare between the two CPUs?

A: The Intel Xeon 6745P uses eight-channel DDR5 memory with 409.6 GB/s bandwidth, while the AMD uses dual-channel DDR5 with 89.6 GB/s. This is a 4.6x difference in theoretical bandwidth, which heavily influences the Xeon’s wins in data compression (51.1% lead) and random string sorting (43.8% lead).

Q: Which CPU has more PCIe lanes for expansion?

A: The Intel Xeon 6745P provides 88 PCIe Gen 5 lanes (CPU only), compared to the AMD’s 24 lanes. This makes the Intel part significantly more capable for servers needing many NVMe drives, GPUs, or network cards.

Q: Are the two CPUs from the same process generation?

A: No, the Intel uses a 5 nm process from Intel’s foundry, while the AMD uses a 4 nm process from TSMC. The AMD also has a much smaller die size (2x 70.6 mm²) compared to Intel (2x 598 mm²), which contributes to its higher clock speeds and lower power draw.

Specification Differences

• Cores: 32 (Intel) vs 16 (AMD)
• Threads: 64 (Intel) vs 32 (AMD)
• Base Clock: 3.10 GHz (Intel) vs 4.30 GHz (AMD)
• Boost Clock: 4.30 GHz (Intel) vs 5.50 GHz (AMD)
• TDP: 300 W (Intel) vs 170 W (AMD)
• Socket: Intel Socket 4710 vs AMD Socket AM5
• Process Node: 5 nm (Intel) vs 4 nm (AMD)
• Foundry: Intel vs TSMC
• Die Size: 2x 598 mm² (Intel) vs 2x 70.6 mm² (AMD)
• L1 Cache (per core): 112 KB (Intel) vs 80 KB (AMD)
• L2 Cache (per core): 2 MB (Intel) vs 1 MB (AMD)
• L3 Cache (shared): 336 MB (Intel) vs 128 MB (AMD)
• Memory Bus: Eight-channel (Intel) vs Dual-channel (AMD)
• Memory Bandwidth: 409.6 GB/s (Intel) vs 89.6 GB/s (AMD)
• PCIe Lanes: 88 (Intel) vs 24 (AMD)
• Integrated Graphics: N/A (Intel) vs Radeon Graphics (AMD)
• Release Date: 2025-02-23 (Intel) vs 2026-06-29 (AMD)
• Part Number: SRWPAQ7L9 (Intel) vs 100-000001999 (AMD)

The Verdict

The Intel Xeon 6745P is the clear choice for server and workstation environments where multi-threaded throughput is paramount. Its 9 out of 11 benchmark wins, including dominant margins of 70.2% in floating-point math and 51.8% in encryption, make it the superior option for database servers, virtualization, scientific computing, and any workload that can utilize its 32 cores and 64 threads. The 409.6 GB/s memory bandwidth and 336 MB L3 cache are decisive advantages that no amount of clock speed on the AMD part can overcome in parallel workloads. The Xeon 6745P also offers far more expansion capability with its 88 PCIe lanes, making it a more versatile platform for high-density storage and accelerator configurations.

The AMD Ryzen 9 PRO 9965X3D is the pick for scenarios that demand the fastest possible single-thread performance. Its 25.2% lead in single-thread benchmarks, driven by 5.50 GHz boost clock, makes it ideal for lightly threaded applications where latency and responsiveness are critical. However, its dual-channel memory and 24 PCIe lanes limit its server utility, and its failure to win a single multi-threaded benchmark means it cannot compete in throughput-oriented roles. The data suggests that the AMD part is a niche product for users who prioritize single-core speed above all else, while the Intel Xeon 6745P is the all-around server powerhouse.