AMD EPYC 9335 vs AMD Ryzen Threadripper PRO 9965WX Comparison

Both processors are active Zen 5 parts built on TSMC's 4 nm node, but they are aimed at different corners of the professional market. The EPYC 9335 is a 32-core server chip for Socket SP5, while the Ryzen Threadripper PRO 9965WX is a 24-core workstation part for Socket sTR5. The benchmark data shows a clear split: the Threadripper wins every single head-to-head test recorded, yet the EPYC holds a higher average benchmark score. This apparent contradiction is one of the most interesting findings in the database, and it suggests that the two chips are optimized for different workloads despite sharing a common architectural foundation.

Where Each One Wins

The head-to-head results are entirely one-sided. The Ryzen Threadripper PRO 9965WX wins all 11 recorded comparisons, which means it takes every category from single-thread performance to multi-thread throughput. The largest margin is in the physics test, where the Threadripper scores 7529 against the EPYC's 1905, a delta of -74.7% from the EPYC's perspective. This is not a small gap; it is a dominant performance in a test that typically reflects real-time simulation and interactive workloads.

The EPYC 9335, despite losing every direct comparison, still occupies a distinct position. Its average benchmark score is 194228, while the Threadripper's average is 147009. This anomaly is explained by the fact that the two chips are tested in different benchmark suites. The EPYC's score is bolstered by high marks in data compression (1203096) and integer math (346291), while the Threadripper's average is dragged down by its lower scores in tests like find prime numbers (752) and single-thread (4551), even though those individual scores are higher than the EPYC's.

So, where does each one win? In the pure head-to-head data, the Threadripper wins everywhere. But in the broader database context, the EPYC's higher average score suggests it excels in sustained server workloads that are not captured in the direct comparison set. The data implies the EPYC is built for throughput in dense, multi-tenant environments, while the Threadripper is built for responsiveness and raw speed in a single-user workstation.

Architecture Differences

Both chips share the same Zen 5 architecture, the same 4 nm process node from TSMC, and the same transistor count of 33,260 million. They also share the same die configuration of 4x 70.6 mm². However, the similarities end there. The EPYC 9335 has 32 cores and 64 threads, while the Threadripper PRO 9965WX has 24 cores and 48 threads. The EPYC's base clock is 3.00 GHz with a boost of 4.40 GHz, whereas the Threadripper starts higher at 4.20 GHz and boosts to 5.40 GHz.

Cache configurations differ notably. The EPYC has 80 KB of L1 per core, while the Threadripper has 64 KB per core. Both have 1 MB of L2 per core, but the EPYC's L3 is 128 MB shared, and the Threadripper also has 128 MB of L3. The L1 difference is curious, as it suggests a different internal design for the server part, possibly to handle more parallel requests.

Memory support is another major divergence. The EPYC uses twelve-channel DDR5 with a bandwidth of 576.0 GB/s, while the Threadripper uses eight-channel DDR5 with 409.6 GB/s. Both support ECC memory. PCIe is identical at Gen 5 with 128 lanes from the CPU. The EPYC is not multiplier unlocked, while the Threadripper is, which allows for overclocking. The EPYC was released on 2024-10-09, and the Threadripper followed much later on 2025-07-22.

FAQ

Q: Which processor has more cores?

A: The AMD EPYC 9335 has 32 cores and 64 threads, while the AMD Ryzen Threadripper PRO 9965WX has 24 cores and 48 threads. The EPYC offers 8 more cores and 16 more threads for parallel workloads.

Q: Is the Threadripper faster in single-threaded tasks?

A: Yes, the data shows the Threadripper PRO 9965WX scores 4551 in the PassMark single-thread test, while the EPYC 9335 scores 2732. This is a -40% delta from the EPYC's perspective, meaning the Threadripper is significantly faster in single-threaded performance.

Q: What is the memory bandwidth difference?

A: The EPYC 9335 has a twelve-channel memory bus with a bandwidth of 576.0 GB/s, while the Threadripper PRO 9965WX has an eight-channel bus with 409.6 GB/s. The EPYC provides 166.4 GB/s more memory bandwidth.

Q: Are both processors on the same manufacturing process?

A: Yes, both are built on TSMC's 4 nm process node. They also have the same transistor count of 33,260 million and the same die size of 4x 70.6 mm².

Q: Can either processor be overclocked?

A: The Ryzen Threadripper PRO 9965WX has an unlocked multiplier, so it can be overclocked. The EPYC 9335 does not have an unlocked multiplier, so its clock speeds are fixed.

Q: Which chip has a higher boost clock?

A: The Threadripper PRO 9965WX boosts to 5.40 GHz, while the EPYC 9335 boosts to 4.40 GHz. The Threadripper's boost is 1.0 GHz higher.

Specification Differences

The two processors differ across several key specifications. The EPYC 9335 has 32 cores versus the Threadripper's 24, and 64 threads versus 48. Base clocks are 3.00 GHz for the EPYC and 4.20 GHz for the Threadripper. Boost clocks are 4.40 GHz and 5.40 GHz, respectively. The TDP is 210 watts for the EPYC and 350 watts for the Threadripper, a substantial difference in power envelope.

The sockets are different: the EPYC uses AMD Socket SP5, while the Threadripper uses AMD Socket sTR5. The codenames also differ: the EPYC is "Turin" and the Threadripper is "Shimada Peak." The L1 cache is 80 KB per core for the EPYC and 64 KB per core for the Threadripper. Memory channels are twelve for the EPYC and eight for the Threadripper, with corresponding bandwidths of 576.0 GB/s and 409.6 GB/s.

The release dates are different, with the EPYC launching on 2024-10-09 and the Threadripper on 2025-07-22. The multiplier is locked on the EPYC and unlocked on the Threadripper. The part numbers are 100-000001149 for the EPYC and 100-000000724 for the Threadripper.

Head-to-Head Benchmarks

The direct comparison set contains 11 tests, and the Threadripper PRO 9965WX wins all of them. The most striking difference is in the physics test, where the Threadripper scores 7529 versus the EPYC's 1905, a delta of -74.7%. This indicates a massive advantage in workloads that rely on physics simulation, likely due to the higher boost clock of 5.40 GHz and the unlocked multiplier.

In the find prime numbers test, the Threadripper scores 752 against the EPYC's 340, a -54.8% delta. This is another large gap, suggesting the Threadripper's higher single-thread performance is critical for such integer-heavy tasks. The single-thread test itself shows a -40% delta, with the Threadripper at 4551 and the EPYC at 2732.

The multithread test shows the Threadripper at 92604 versus the EPYC's 65811, a -28.9% delta. This is a significant win, despite the EPYC having more cores. The Threadripper's higher clocks and per-core performance overcome the core count disadvantage. Random string sorting shows a -22.1% delta, with the Threadripper at 149617 and the EPYC at 116608.

The smaller margins are in floating point math (-0.7%), integer math (-0.8%), and extended instructions (-2.8%). These are close calls, but the Threadripper still edges ahead. Data compression shows a -10.6% delta, and data encryption shows a -4.5% delta. The data implies the Threadripper is consistently faster across the board, but the gap narrows in pure compute tasks where the EPYC's additional cores can partially compensate.

The Verdict

The data clearly points to the Ryzen Threadripper PRO 9965WX as the faster processor in every direct comparison. If the task is single-threaded performance, the Threadripper wins with a 40% advantage. If the task is multi-threaded, it still wins, despite having 8 fewer cores. The physics test is not even close, with the Threadripper outperforming the EPYC by nearly four times. For any workstation use case where the CPU is the bottleneck, the Threadripper is the better choice.

The EPYC 9335, however, is not without merit. Its higher average benchmark score of 194228 versus the Threadripper's 147009 indicates that in the database's broader suite, the EPYC excels in specific server-oriented workloads. Its twelve-channel memory bus provides 576.0 GB/s of bandwidth, which is 166.4 GB/s more than the Threadripper. This makes it suitable for memory-intensive server tasks that are not in the head-to-head set. Additionally, its lower TDP of 210 watts versus 350 watts means it can be deployed in denser configurations with less cooling overhead.

The decision comes down to the use case. For a single user running interactive simulations, rendering, or compiling, the Threadripper PRO 9965WX is the obvious pick based on its benchmark dominance. For a server environment where memory bandwidth and core count matter more than single-thread speed, the EPYC 9335 is the data-backed choice. The Threadripper's unlocked multiplier also offers flexibility for those who want to push performance further, while the EPYC's locked multiplier ensures stability in enterprise settings. The data does not show one as universally better; it shows two tools for different jobs.