AMD Ryzen Threadripper 9970X vs Intel Xeon 6774P Comparison
AMD Ryzen Threadripper 9970X
Xeon 6774P
PERFORMANCE BENCHMARKS
Analysis: AMD Ryzen Threadripper 9970X vs Intel Xeon 6774P
The Verdict
The benchmark data presents a clear split between these two processors, with the Intel Xeon 6774P claiming a decisive 9-to-2 victory in direct head-to-head tests. The Xeon 6774P dominates in every multi-threaded workload measured, often by substantial margins, while the AMD Ryzen Threadripper 9970X wins only the single-threaded tests. The Intel part’s average benchmark score of 300372 places it 7.4% ahead of the Threadripper’s 279778, and both sit in the 99th percentile of all CPUs.
For workloads that scale across many cores, the Xeon 6774P is the clear choice. Its 64 cores and 128 threads, combined with a massive 336 MB of shared L3 cache, deliver crushing performance in parallel tasks. The data shows the Xeon leading by 105.5% in prime number finding, 134.4% in physics calculations, and 50.2% in floating-point math. These are not marginal advantages; they represent fundamentally different performance tiers in throughput-oriented applications.
The Ryzen Threadripper 9970X, however, is the pick for single-threaded sensitivity. Its 32 cores based on Zen 5 architecture run at a 4.00 GHz base and 5.40 GHz boost clock, producing a 4530 PassMark single-thread score that beats the Xeon’s 3047 by 32.7%. That advantage matters for lightly threaded applications, interactive workloads, or any scenario where per-core latency dominates over raw core count. The Threadripper also offers an unlocked multiplier, enabling overclocking, whereas the Xeon is locked. For a desktop workstation where single-core responsiveness and frequency flexibility are priorities, the Threadripper 9970X is the logical selection.
Where Each One Wins
The Xeon 6774P wins in every multi-threaded category listed in the head-to-head data. Its largest victory comes in PassMark physics, where it scores 16023 against the Threadripper’s 6835, a 134.4% advantage. Prime number finding shows a 105.5% lead, with scores of 1264 versus 615. Floating-point math favors the Xeon by 50.2%, with scores of 465314 versus 309719. Data encryption shows a 32.6% lead (115025 versus 86765), and data compression follows at 31.4% (2309868 versus 1757998). Integer math, extended instructions, random string sorting, and the overall multithread score all go to the Xeon, with leads ranging from 5% to 37.1%. The smallest multi-threaded margin is in the PassMark multithread test itself, where the Xeon scores 112749 versus 107399, a 5% edge that reflects the Threadripper’s relatively strong scaling despite having half the cores.
The Threadripper 9970X wins solely in single-threaded performance. Its PassMark single-thread score of 4530 beats the Xeon’s 3047 by 32.7%. This is the only category where the AMD part shows superiority, and it is a meaningful one for workloads that cannot utilize many cores. The Threadripper also carries a higher base clock of 4.00 GHz versus 2.50 GHz, and a higher boost clock of 5.40 GHz versus 3.90 GHz, which explains the single-thread gap. For desktop users who run a mix of lightly and heavily threaded applications, the Threadripper’s strong single-core performance may compensate for its multi-core deficit in real-world usage patterns.
Architecture Differences
The two processors come from different design philosophies. The Intel Xeon 6774P uses Granite Rapids architecture on a 5 nm process from Intel’s own foundry, with a die size of 2x 598 mm². It packs 64 cores and 128 threads, with 112 KB of L1 cache per core, 2 MB of L2 per core, and a shared 336 MB L3 cache. Memory support is DDR5 over an eight-channel bus, delivering 409.6 GB/s of bandwidth. PCIe connectivity is Gen 5 with 136 lanes from the CPU. The chip targets the server and workstation market, fits Intel Socket 4710, and has a TDP of 350 W.
The AMD Ryzen Threadripper 9970X uses Zen 5 architecture under the codename Shimada Peak, fabricated by TSMC on a 4 nm process. It has 32 cores and 64 threads, with 64 KB of L1 per core, 1 MB of L2 per core, and 128 MB of L3 cache. The transistor count is 33,260 million across a die size of 4x 70.6 mm². Memory is DDR5 on a quad-channel bus, providing 204.8 GB/s — exactly half the Xeon’s bandwidth. PCIe is Gen 5 with 80 lanes. This is a desktop part using AMD Socket sTR5, with a 350 W TDP and an unlocked multiplier for overclocking.
The core count difference is stark: the Xeon has exactly twice the cores and threads. The L3 cache difference is also large, with the Xeon offering 336 MB versus 128 MB. However, the Threadripper counters with a 4 nm process versus 5 nm, higher clocks, and a smaller die. The Xeon’s eight-channel memory bus gives it double the theoretical memory bandwidth, which is critical for data-intensive server workloads. The Threadripper’s quad-channel bus is half that, but its higher clocks and newer process node help it remain competitive in single-threaded tasks. Both support ECC memory, and neither has integrated graphics.
FAQ
Q: Which processor has more cores?
A: The Intel Xeon 6774P has 64 cores and 128 threads, while the AMD Ryzen Threadripper 9970X has 32 cores and 64 threads. The Xeon has exactly double the core count.
Q: How does single-threaded performance compare?
A: The Threadripper 9970X wins the PassMark single-thread test with a score of 4530, which is 32.7% higher than the Xeon 6774P’s 3047. The Threadripper’s higher base clock of 4.00 GHz and boost clock of 5.40 GHz contribute to this lead.
Q: What is the biggest benchmark margin between the two?
A: The largest difference is in PassMark physics, where the Xeon scores 16023 versus the Threadripper’s 6835, a 134.4% advantage for Intel. Prime number finding is second, with the Xeon leading by 105.5%.
Q: Which processor has higher memory bandwidth?
A: The Xeon 6774P offers 409.6 GB/s over an eight-channel DDR5 bus, which is double the Threadripper’s 204.8 GB/s over a quad-channel bus. This gives Intel a significant advantage in memory-intensive workloads.
Q: Can the Threadripper be overclocked?
A: Yes, the Ryzen Threadripper 9970X has an unlocked multiplier. The Intel Xeon 6774P does not, so it cannot be overclocked.
Q: What is the market segment for each processor?
A: The Xeon 6774P is classified as a server and workstation part, while the Threadripper 9970X is a desktop processor. Both are currently in active production.
Head-to-Head Benchmarks
The PassMark suite reveals a consistent pattern: the Xeon 6774P wins every multi-threaded test, often by wide margins, while the Threadripper 9970X takes only the single-threaded tests. The most lopsided result is in physics, where the Xeon scores 16023 against 6835, a 134.4% lead. This suggests the Xeon’s core count and cache hierarchy provide exceptional throughput in physics simulations. Prime number finding shows a similar story, with the Xeon at 1264 versus 615, a 105.5% advantage that reflects its ability to handle large numbers of concurrent integer operations.
Floating-point math is another area of clear Intel superiority. The Xeon scores 465314, which is 50.2% ahead of the Threadripper’s 309719. Integer math shows a 27.5% lead for Intel, with scores of 593440 versus 465378. Data encryption favors the Xeon by 32.6%, scoring 115025 versus 86765. Data compression shows a 31.4% edge, with the Xeon at 2309868 versus 1757998. Extended instructions go to Intel by 24.5%, with scores of 177273 versus 142342. Random string sorting is 37.1% higher on the Xeon, scoring 262417 versus 191445.
The closest multi-threaded contest is the PassMark multithread test itself. The Xeon scores 112749, only 5% above the Threadripper’s 107399. This narrow margin is surprising given the Xeon’s 2x core count and 2.6x L3 cache, but it indicates that the Threadripper’s higher clocks and Zen 5 architecture allow it to scale efficiently across its 32 cores. Despite this, the Xeon still wins, and the overall average benchmark score reflects that: 300372 for Intel versus 279778 for AMD, a 7.4% gap.
Single-threaded performance is the one bright spot for AMD. The Threadripper’s 4530 score beats the Xeon’s 3047 by 32.7%. This is consistent with the clock speed difference, as the Threadripper boosts to 5.40 GHz versus the Xeon’s 3.90 GHz. For workloads that are latency-bound or run on one or two threads, the Threadripper will feel noticeably faster. The data also shows the Threadripper’s nearest rivals are Intel Xeon 6780E, AMD EPYC 9565, Intel Xeon 696X, and AMD EPYC 9555P, with delta percentages ranging from -0.2% to -2.5%. The Xeon 6774P, meanwhile, sits within 3.3% to 5% of AMD EPYC 9734, AMD EPYC 9575F, AMD EPYC 9555P, and Intel Xeon 696X, indicating that both processors are tightly clustered among top-tier workstation and server parts.
In summary, the Xeon 6774P is the throughput king, winning 9 of 11 head-to-head benchmarks with margins as high as 134.4%. The Threadripper 9970X is the single-thread specialist, offering a 32.7% boost in that category and an unlocked multiplier. The choice between them depends entirely on whether the workload is dominated by parallel computation or single-threaded responsiveness.