AMD Ryzen Threadripper 9970X vs Intel Xeon 6747P Comparison

The Verdict

The Intel Xeon 6747P and AMD Ryzen Threadripper 9970X occupy different positions in the workstation and server spectrum, and the benchmark data reflects that split clearly. The Xeon 6747P wins 7 of 11 head-to-head tests, while the Threadripper 9970X takes 4. However, the magnitude and nature of those wins matter more than the raw count. The Xeon dominates compute-heavy integer and floating-point workloads, often by double-digit margins, while the Threadripper counterattacks with a decisive single-thread advantage and wins in multithreaded throughput and string sorting.

For users whose workloads are physics simulations, prime-number searches, or floating-point math, the Xeon 6747P is the clear choice—its 96% lead in physics and 87.2% lead in prime-number tests are not close. For mixed desktop workstation use where single-thread responsiveness and per-core speed matter, the Threadripper 9970X offers a 28.6% single-thread advantage and a 5.3% multithread win. The Xeon also leads in overall average benchmark score categories relative to its nearest rivals, but the Threadripper's average score of 279778 sits higher than the Xeon's 238263, indicating broader strength across the full benchmark suite despite losing most individual head-to-head tests. Both parts sit in the 99th percentile of all CPUs, so neither is a weak choice; the decision hinges on workload character.

Architecture Differences

The two processors come from fundamentally different design philosophies. The Intel Xeon 6747P is built on Granite Rapids architecture, fabricated on a 5 nm process at Intel, with a die size of 2x 598 mm². It packs 48 cores and 96 threads, with base and boost clocks of 2.70 GHz and 3.90 GHz respectively. Its cache hierarchy is generous: 112 KB of L1 per core, 2 MB of L2 per core, and 288 MB of shared L3. Memory support spans DDR5 across an eight-channel bus, delivering 409.6 GB/s of bandwidth, with ECC support standard. PCIe comes in at Gen 5 with 88 lanes from the CPU. It uses the Intel Socket 4710 and has a TDP of 330 W. The multiplier is locked, and the launch MSRP is $6497.

The AMD Ryzen Threadripper 9970X, part of the 9000 series, uses Zen 5 architecture on a 4 nm TSMC process, with a die size of 4x 70.6 mm² and 33,260 million transistors. It has 32 cores and 64 threads, with base and boost clocks of 4.00 GHz and 5.40 GHz—both substantially higher than the Xeon. Cache is smaller in aggregate: 64 KB L1 per core, 1 MB L2 per core, and 128 MB L3. Memory is DDR5 over a quad-channel bus, yielding 204.8 GB/s of bandwidth—exactly half the Xeon's figure—but ECC support is present. PCIe is Gen 5 with 80 lanes. It uses AMD Socket sTR5, has a TDP of 350 W, and the multiplier is unlocked. The launch MSRP is $2499. The Threadripper is classified as a desktop part, while the Xeon is server/workstation, which explains the divergent memory channels and core counts.

These architectural differences map directly onto benchmark outcomes. The Xeon's 16 additional cores and 288 MB of L3 drive its wins in parallel compute, while the Threadripper's higher clocks and smaller 4 nm process drive its single-thread dominance. The Xeon's eight-channel memory provides 409.6 GB/s versus 204.8 GB/s, yet the Threadripper still wins in random string sorting and multithread tests—suggesting that memory bandwidth alone does not determine those outcomes.

Where Each One Wins

Intel Xeon 6747P wins: The data shows the Xeon dominates in compute-bound, highly parallel workloads. In floating-point math, it scores 365904 against the Threadripper's 309719, an 18.1% lead. Integer math is closer but still favors Intel: 468518 versus 465378, a 0.7% margin. The largest gaps appear in physics (13398 versus 6835, a 96% lead) and prime-number finding (1151 versus 615, an 87.2% lead). Data compression (1833378 versus 1757998, 4.3% ahead) and data encryption (90789 versus 86765, 4.6% ahead) also favor the Xeon. Extended instructions are essentially a tie at 142557 versus 142342, just 0.2% apart. The Xeon's wins total 7 out of 11 head-to-head benchmarks.

AMD Ryzen Threadripper 9970X wins: The Threadripper excels where clock speed and per-core efficiency matter. Single-thread performance is its strongest category: 4530 versus 3236, a 28.6% advantage. Multithread throughput also favors AMD at 107399 versus 101685, a 5.3% lead. Random string sorting goes to the Threadripper at 191445 versus 180382, a 5.8% edge. These four wins are concentrated in memory-latency-sensitive or latency-sensitive tasks, where the higher 5.40 GHz boost clock and 4.00 GHz base clock provide an advantage over the Xeon's 3.90 GHz boost and 2.70 GHz base. The Threadripper's average benchmark score of 279778 also exceeds the Xeon's 238263, indicating that across the broader benchmark suite, the AMD part holds up well despite losing most direct comparisons.

The workload split is clear: if the task is number-crunching with massive parallelism, choose Intel. If the task involves mixed workloads with significant single-thread components or memory-latency-sensitive operations, choose AMD.

FAQ

Q: Which processor has more cores and threads?

A: The Intel Xeon 6747P has 48 cores and 96 threads, while the AMD Ryzen Threadripper 9970X has 32 cores and 64 threads.

Q: How do their single-thread performances compare?

A: The Threadripper 9970X scores 4530 in PassMark single-thread tests, which is 28.6% higher than the Xeon 6747P's score of 3236.

Q: Which CPU wins in floating-point math workloads?

A: The Xeon 6747P scores 365904 in floating-point math, which is 18.1% higher than the Threadripper's 309719.

Q: What is the memory bandwidth difference?

A: The Xeon 6747P supports eight-channel DDR5 memory with 409.6 GB/s bandwidth, while the Threadripper 9970X supports quad-channel DDR5 with 204.8 GB/s bandwidth—exactly half.

Q: Are both CPUs in the top percentile of all processors?

A: Yes, both the Xeon 6747P and the Threadripper 9970X are in the 99th percentile of all CPUs according to their benchmark scores.

Q: Which CPU has the higher launch MSRP?

A: The Intel Xeon 6747P has a launch MSRP of $6497, while the AMD Ryzen Threadripper 9970X has a launch MSRP of $2499.

Head-to-Head Benchmarks

The biggest win for the Intel Xeon 6747P comes in PassMark physics, where it scores 13398 against the Threadripper's 6835—a 96% advantage. This is nearly double the performance, and it reflects the Xeon's 48-core design and 288 MB of L3 cache handling physics simulation workloads far more efficiently than the Threadripper's 32 cores and 128 MB L3. The second-largest margin is in prime-number finding: 1151 versus 615, an 87.2% lead. These two tests alone establish the Xeon as the superior choice for scientific and engineering computation that relies on integer-heavy loops and physics modeling.

Floating-point math shows an 18.1% lead for the Xeon (365904 versus 309719), and data encryption a 4.6% edge (90789 versus 86765). Data compression goes to the Xeon by 4.3% (1833378 versus 1757998). Integer math is nearly a tie at 0.7% (468518 versus 465378), and extended instructions are within 0.2% (142557 versus 142342). These results indicate that the Xeon's advantage grows as the workload becomes more computationally intensive and parallel, while it narrows to parity in simpler instruction streams.

The AMD Ryzen Threadripper 9970X's largest win is single-thread performance: 4530 versus 3236, a 28.6% advantage. This is the direct result of the 5.40 GHz boost clock versus 3.90 GHz, and it matters for applications that cannot parallelize across many cores. The Threadripper also wins in multithread throughput at 107399 versus 101685, a 5.3% margin, which is notable given that the Xeon has 50% more cores—the Threadripper's higher per-core efficiency compensates for the core deficit. Random string sorting goes to the Threadripper at 191445 versus 180382, a 5.8% lead, suggesting better memory latency characteristics in pointer-chasing workloads despite the Xeon's higher raw memory bandwidth.

The overall head-to-head record stands at 7 wins for the Xeon and 4 for the Threadripper. Yet the Threadripper's average benchmark score of 279778 exceeds the Xeon's 238263 by about 17.4%, and its nearest rivals include the Intel Xeon 6780E at 280438 (just 0.2% higher) and the AMD EPYC 9565 at 285471 (2% higher). The Xeon's nearest rivals include the AMD EPYC 9634 at 244274 (2.5% higher) and the Intel Xeon 6980P at 251516 (5.3% higher). These proximity values show that both parts are tightly grouped with their respective performance classes, but the Threadripper sits in a faster overall company.

Benchmark results indicate that the Xeon 6747P is the stronger raw compute engine in most specific tests, while the Threadripper 9970X offers superior single-thread speed and competitive multithread performance despite fewer cores. The choice between them depends on whether the workload prioritizes massive parallel throughput or per-thread responsiveness.