AMD Ryzen Threadripper 9980X vs Intel Xeon 6747P Comparison

FAQ

Q: How does the AMD Ryzen Threadripper 9980X compare to the Intel Xeon 6747P in overall average benchmark score?

A: The Threadripper 9980X records an average benchmark score of 321,753, while the Xeon 6747P scores 238,263. The AMD processor sits 35% higher in aggregate performance, placing it in the 99th percentile of all CPUs, the same percentile as the Intel part.

Q: Which processor wins the most head-to-head benchmark comparisons?

A: The AMD Ryzen Threadripper 9980X wins 12 of the 14 recorded head-to-head tests. The Intel Xeon 6747P claims only two victories, specifically in passmark_find_prime_numbers and passmark_physics.

Q: What are the core and thread counts for each processor?

A: The Threadripper 9980X provides 64 cores and 128 threads, while the Xeon 6747P provides 48 cores and 96 threads. Both support ECC memory, but the AMD part has an unlocked multiplier whereas the Intel part is locked.

Q: How large is the single-core performance gap between the two?

A: In passmark_single_thread, the Threadripper 9980X scores 4,537 versus 3,236 for the Xeon 6747P, a 40.2% advantage for AMD. In Cinebench R23 single-core, the AMD part scores 18,427 compared to 18,427 for Intel, showing a similar single-thread edge.

Q: What memory bandwidth do these platforms offer?

A: The Threadripper 9980X uses quad-channel DDR5 with 204.8 GB/s of bandwidth. The Xeon 6747P uses eight-channel DDR5 with 409.6 GB/s of bandwidth, giving Intel a 2x theoretical memory bandwidth advantage despite losing most compute benchmarks.

Q: What are the launch MSRP values for these processors?

A: The AMD Ryzen Threadripper 9980X has a launch MSRP of $4999, and the Intel Xeon 6747P has a launch MSRP of $6497.

Where Each One Wins

The head-to-head record is decisively lopsided, but the two wins for the Intel Xeon 6747P point to specific workloads where its architecture excels. In passmark_find_prime_numbers, the Xeon scores 1,151 against 769 for the Threadripper, a 33.2% advantage for Intel. Prime number finding is heavily dependent on integer division and modular arithmetic, and the Granite Rapids core design handles this pattern more efficiently. Similarly, passmark_physics shows the Xeon at 13,398 versus 8,001 for AMD, a 40.3% margin. Physics simulation workloads often feature irregular memory access and branch-heavy code, areas where the Xeon's larger per-core L1 cache (112 KB per core versus 64 KB per core) and 2 MB per-core L2 cache can help.

Everywhere else, the AMD Ryzen Threadripper 9980X dominates. The largest win comes in passmark_integer_math, where AMD scores 872,071 against 468,518 for Intel, an 86.1% advantage. This is the clearest signal of the Threadripper's raw compute throughput advantage in scalar integer work. Data compression also favors AMD heavily: 2,974,534 versus 1,833,378, a 62.2% delta. Encryption workloads show a 73.1% gap, and random string sorting shows a 61.9% gap. Floating-point math, extended instructions, and multithread tests all land in AMD's favor with margins between 39.3% and 52.8%.

The use-case split is straightforward. The Xeon 6747P suits workloads that stress prime-number generation, physics simulation, and any pattern resembling those access structures. The Threadripper 9980X is the better choice for encryption, compression, integer-heavy compute, and general multithreaded throughput. For buyers weighing the two, the 12-to-2 head-to-head record strongly favors AMD across the broad benchmark suite, with Intel's wins confined to niche computational patterns.

Architecture Differences

The two processors come from fundamentally different design philosophies. The AMD Ryzen Threadripper 9980X uses the Zen 5 architecture under the codename Shimada Peak, built on a 4 nm process at TSMC. The Intel Xeon 6747P uses Granite Rapids architecture, built on a 5 nm process at Intel. AMD's transistor count is listed at 66,520 million spread across 8 dies of 70.6 mm² each. Intel's die size is listed as 2x 598 mm², indicating a dual-die design with significantly larger individual dies.

Cache hierarchies differ in structure. The Threadripper 9980X has 64 KB of L1 per core and 1 MB of L2 per core, with a shared 256 MB L3. The Xeon 6747P has 112 KB of L1 per core and 2 MB of L2 per core, with a shared 288 MB L3. Intel dedicates more cache per core, while AMD's total L3 is 256 MB versus Intel's 288 MB shared pool.

Memory architecture diverges sharply. AMD offers quad-channel DDR5 with 204.8 GB/s of bandwidth. Intel offers eight-channel DDR5 with 409.6 GB/s of bandwidth. The Xeon doubles AMD's memory bandwidth on paper, which can matter for workloads that stream large datasets. PCIe connectivity also differs: AMD provides Gen 5 with 80 lanes from the CPU, while Intel provides Gen 5 with 88 lanes.

The market segments reflect intended use. AMD positions the Threadripper 9980X as a desktop part, while Intel positions the Xeon 6747P as a server or workstation processor. Both have active production status. AMD's release date is recorded as 2025-07-29, and Intel's as 2025-02-23, making the Xeon the earlier launch by several months.

Specification Differences

The most consequential specification gap is core count: 64 cores and 128 threads on the AMD side versus 48 cores and 96 threads on the Intel side. Clock speeds also favor AMD, with a base clock of 3.20 GHz and boost clock of 5.40 GHz for the Threadripper, versus 2.70 GHz base and 3.90 GHz boost for the Xeon. The 1.5 GHz boost clock gap is substantial.

Thermal design power is close: 350 W for AMD and 330 W for Intel. Socket compatibility is entirely separate, with AMD using Socket sTR5 and Intel using Socket 4710. The AMD part has an unlocked multiplier; the Intel part does not. Part numbers are recorded as 100-000001593 for AMD and SRVEZ for Intel.

Process node and foundry differ, with AMD on TSMC's 4 nm and Intel on its own 5 nm. Memory channels differ, with quad-channel on AMD and eight-channel on Intel. PCIe lane counts differ slightly, 80 for AMD and 88 for Intel. Both support DDR5 and ECC memory, and neither includes integrated graphics.

Both processors are in the 99th percentile of all CPUs in the database, but the average benchmark score gap is 321,753 versus 238,263. The nearest rival to the Threadripper is the AMD Ryzen Threadripper PRO 9985WX at 320,749 (0.3% behind), followed by the Intel Xeon 6781P at 315,524 (2% behind). The nearest rival to the Xeon 6747P is the AMD EPYC 9634 at 244,274 (2.5% ahead of the Intel part), followed by the Intel Xeon 6980P at 251,516 (5.3% ahead).

Head-to-Head Benchmarks

The Cinebench suite shows a consistent 51% advantage for AMD across all three versions. In Cinebench R15 multicore, the Threadripper scores 13,157 versus 8,712 for the Xeon. R20 shows 54,822 versus 36,301. R23 shows 130,529 versus 86,432. This consistency across render versions indicates the margin is structural, driven by core count and clock speed rather than workload-specific quirks.

The largest single margin in the entire comparison is passmark_integer_math at 86.1%, with AMD scoring 872,071 and Intel scoring 468,518. This is more than double the Cinebench margin and suggests the Zen 5 integer execution pipeline delivers exceptional throughput per clock. Data encryption shows a 73.1% gap (157,137 versus 90,789), and data compression shows a 62.2% gap (2,974,534 versus 1,833,378). Extended instructions follow at 60.6% (228,959 versus 142,557), and random string sorting at 61.9% (292,083 versus 180,382).

Floating-point math shows a 52.8% margin (559,003 versus 365,904). Multithread performance shows a 39.3% margin (141,641 versus 101,685), which is the smallest AMD win among the PassMark tests. Single-thread performance shows a 40.2% margin (4,537 versus 3,236), confirming that AMD's clock speed advantage translates directly to single-core throughput.

The Intel wins are isolated but notable. Passmark_find_prime_numbers shows the Xeon at 1,151 versus 769 for AMD, a 33.2% lead. Passmark_physics shows 13,398 versus 8,001, a 40.3% lead. These two tests reward the Xeon's larger per-core caches and different execution resource allocation. The physics result is particularly interesting because it is the largest margin in Intel's favor, suggesting the Granite Rapids core handles the physics test's instruction mix more efficiently despite lower clocks.

Overall, the benchmark data paints a clear picture. The Threadripper 9980X wins the overwhelming majority of tests with margins that often exceed 50%, while the Xeon 6747P claims narrow-to-moderate wins in two specialized workloads. The average benchmark score difference of 83,490 points (321,753 versus 238,263) encapsulates the overall performance hierarchy.