AMD EPYC 7413 vs AMD Ryzen Threadripper PRO 9945WX Comparison

Where Each One Wins

The AMD EPYC 7413 and AMD Ryzen Threadripper PRO 9945WX target different priorities within the server and workstation space, and the benchmark data reflects that split cleanly. The EPYC 7413 takes 5 of the 17 head-to-head benchmark wins, while the Threadripper PRO 9945WX claims 12. These are not close, evenly matched chips; they are two very different designs optimized for opposite ends of the workload spectrum.

The EPYC 7413 wins exclusively in PassMark workloads that reward raw throughput and data manipulation. It leads in data compression by 6.5%, data encryption by a massive 32.7%, floating point math by 6.6%, integer math by 16.3%, and prime number finding by 18.5%. This is a chip built around a 24-core, 48-thread Zen 3 design with a large shared L3 cache, and it shows in workloads that scale with core count and cache capacity. The encryption lead is particularly notable, as it is the largest single delta in either direction. For server-side tasks like database encryption, secure socket termination, or compression-heavy data pipelines, the EPYC 7413 is clearly the stronger option.

The Threadripper PRO 9945WX wins every Cinebench test, both single-core and multi-core, by a consistent 10.9% to 11% margin. It also dominates in PassMark's physics simulation, multithread, random string sorting, extended instructions, and single-thread tests. The single-thread win is not small: it is 47.5% ahead of the EPYC 7413 in PassMark's single-thread test, and 90.1% ahead in the Cinebench R23 single-core test when comparing raw scores (6822 versus 6076 is actually only 12.3% ahead, but the PassMark gap is the standout). This is a chip with a 4.70 GHz base clock and 5.40 GHz boost clock, built on Zen 5, and it clearly favors low-latency, high-frequency, and lightly threaded workloads.

In practical terms, the EPYC 7413 is the pick for server workloads that hammer memory and cache with parallel data streams, while the Threadripper PRO 9945WX is the pick for workstation tasks like CAD, simulation, rendering previews, and any application where a single thread's responsiveness matters. The Threadripper's Cinebench multi-core wins also suggest it holds its own in heavily threaded render workloads, despite having half the core count, because its per-core performance is so much higher.

FAQ

Q: Which CPU is faster in single-core performance?

A: The AMD Ryzen Threadripper PRO 9945WX wins every single-thread benchmark in the database. It leads by 10.9% in Cinebench R15 and R23 single-core, by 11% in Cinebench R20 single-core, and by a dominant 47.5% in PassMark's single-thread test.

Q: Does the EPYC 7413 beat the Threadripper PRO 9945WX in any multi-threaded test?

A: Yes, but only in specific PassMark workloads. The EPYC 7413 wins data compression (6.5% ahead), data encryption (32.7% ahead), floating point math (6.6% ahead), integer math (16.3% ahead), and prime number finding (18.5% ahead). However, it loses the Cinebench multi-core tests and PassMark's multithread test to the Threadripper PRO 9945WX.

Q: What is the core and thread count difference?

A: The AMD EPYC 7413 has 24 cores and 48 threads, while the AMD Ryzen Threadripper PRO 9945WX has 12 cores and 24 threads. Despite having half the cores, the Threadripper wins the multi-core Cinebench tests due to its higher clock speeds and newer architecture.

Q: Which CPU has a higher base clock speed?

A: The AMD Ryzen Threadripper PRO 9945WX has a base clock of 4.70 GHz, which is significantly higher than the EPYC 7413's 2.65 GHz base clock. Its boost clock is 5.40 GHz versus 3.60 GHz for the EPYC.

Q: Are both CPUs for the same socket?

A: No. The AMD EPYC 7413 uses AMD Socket SP3, while the AMD Ryzen Threadripper PRO 9945WX uses AMD Socket sTR5. They are not interchangeable between platforms.

Q: What is the memory bandwidth difference?

A: The Threadripper PRO 9945WX supports DDR5 memory with a total bandwidth of 409.6 GB/s, exactly double the EPYC 7413's 204.8 GB/s, which is limited to DDR4 memory.

Head-to-Head Benchmarks

The most striking result in the entire dataset is the PassMark single-thread test. The Threadripper PRO 9945WX scores 4573, while the EPYC 7413 scores 2400, a delta of 47.5% in favor of the Threadripper. This is not a small margin; it is a generational leap. The same pattern appears in the Cinebench single-core tests, with the Threadripper leading by 10.9% in R15 (687 versus 612) and R23 (6822 versus 6076), and by 11% in R20 (2865 versus 2551). The Threadripper's 5.40 GHz boost clock and Zen 5 architecture are clearly the drivers here.

The multi-core Cinebench results are also telling. The Threadripper wins Cinebench R15 multi-core with 4871 versus 4338, R20 multi-core with 20296 versus 18078, and R23 multi-core with 48325 versus 43044. Each of these wins is a 10.9% margin, which is surprisingly consistent. It means that despite having only 12 cores versus 24, the Threadripper's higher frequency and architectural efficiency more than compensate. The same margin appears in PassMark's multithread test, where the Threadripper scores 56854 versus 50641, again a 10.9% lead.

The EPYC 7413's wins are concentrated in the PassMark math and data workloads. The largest is data encryption, where the EPYC scores 48492 versus the Threadripper's 36540, a 32.7% lead. This is a workload that benefits heavily from the EPYC's 128 MB shared L3 cache and 24-core topology. Integer math follows with a 16.3% lead (215629 versus 185421), prime number finding with an 18.5% lead (397 versus 335), floating point math with a 6.6% lead (118881 versus 111566), and data compression with a 6.5% lead (715616 versus 671963). The EPYC also wins PassMark's physics test by a narrow margin? No, it does not: the Threadripper wins physics by 23% (6118 versus 4708), and random string sorting by 4% (84498 versus 81134).

The extended instructions test goes to the Threadripper by 16% (54406 versus 45696). Overall, the data shows a clear pattern: the Threadripper wins every latency-sensitive and frequency-bound test, while the EPYC wins only the throughput-heavy, cache-hungry data workloads.

Specification Differences

The two CPUs differ in nearly every core specification. The EPYC 7413 has 24 cores and 48 threads, while the Threadripper PRO 9945WX has 12 cores and 24 threads, exactly half. The base clock is 2.65 GHz for the EPYC versus 4.70 GHz for the Threadripper, and the boost clock is 3.60 GHz versus 5.40 GHz. The EPYC has a TDP of 180 watts, while the Threadripper has a TDP of 350 watts. The sockets are different: SP3 for the EPYC, sTR5 for the Threadripper.

Memory support differs significantly. The EPYC uses DDR4 with an eight-channel bus and 204.8 GB/s of bandwidth. The Threadripper uses DDR5 with an eight-channel bus and 409.6 GB/s of bandwidth. Both support ECC memory. PCIe connectivity also differs: the EPYC provides Gen 4 with 128 lanes (CPU only), while the Threadripper provides Gen 5 with 128 lanes (CPU only). The Threadripper has an unlocked multiplier, while the EPYC does not. The EPYC has a launch MSRP of $1825, while the Threadripper's launch MSRP is not recorded. The EPYC was released on 2021-03-14, and the Threadripper on 2025-06-30.

Architecture Differences

The architectural gap between these two is substantial. The EPYC 7413 is built on Zen 3 architecture, codenamed Milan, and belongs to the EPYC 7003 series. It uses a 7 nm process node from TSMC, with 16,600 million transistors spread across a die size of 4x 81 mm². The Threadripper PRO 9945WX is built on Zen 5 architecture, codenamed Shimada Peak, and belongs to the 9000 series. It uses a 4 nm process node from TSMC, with 16,630 million transistors across a die size of 2x 70.6 mm². The transistor count is nearly identical, but the Threadripper packs them into a smaller total die area, which partially explains its higher clock speeds.

Cache configurations differ as well. Both have 64 KB of L1 cache per core. The L2 cache is 512 KB per core on the EPYC, but 1 MB per core on the Threadripper, double the amount. The L3 cache is 128 MB (shared) on the EPYC, versus 64 MB on the Threadripper. This is a key reason why the EPYC wins the data-heavy PassMark workloads: it has twice the L3 capacity, which helps with large working sets in encryption and compression. The Threadripper's smaller but faster cache, combined with its much higher clocks, favors lower-latency and single-threaded tasks.

The integrated graphics are listed as N/A for the Threadripper, and not present for the EPYC. The EPYC's production status is Active, as is the Threadripper's. The EPYC's part number is 100-000000323100-100000323WOF, while the Threadripper's is 100-000000726. The Threadripper's market segment is Server/Workstation, matching the EPYC. Both sit at the 95th percentile against all CPUs in the database, but their average benchmark scores differ: the EPYC averages 80041, while the Threadripper averages 76513. The nearest rivals for the EPYC include the Intel Core Ultra 9 290HX Plus (0.6% ahead), Intel Xeon w5-2565X (0.8% behind), Intel Core i9-14900KF (0.8% ahead), and Intel Xeon 638 (0.8% behind). The Threadripper's nearest rivals include the AMD EPYC Embedded 8224P (0% delta), AMD Ryzen 9 8945HX (0.4% behind), Intel Core Ultra 9 285HX (0.5% behind), and AMD Ryzen 9 9950X3D (1% behind). In short, the architecture difference is a classic trade-off: more cores and more cache on older Zen 3 versus fewer cores, much higher frequency, and newer Zen 5 on the Threadripper.