CPU Comparison

The AMD EPYC 7351P and AMD Ryzen Threadripper 1920 are both 14 nm Zen-based AMD processors, but they target different segments: the EPYC 7351P is a server/workstation part on the SP3 platform, while the Threadripper 1920 is a desktop enthusiast part on the SP3r2 platform. The benchmark data shows a consistent performance advantage for the EPYC 7351P across every Cinebench test, with the EPYC winning all six head-to-head comparisons. The EPYC 7351P holds a 0.8% higher average benchmark score (5469 vs 5425) and a 61st percentile ranking versus the Threadripper’s 60th, placing them in the same performance tier despite their architectural and platform differences.

FAQ

Q: Which processor has more cores and threads?

A: The AMD EPYC 7351P has 16 cores and 32 threads, while the AMD Ryzen Threadripper 1920 has 12 cores and 24 threads. This gives the EPYC a 33% core and thread count advantage.

Q: How do their average benchmark scores compare?

A: The EPYC 7351P has an average benchmark score of 5469, which is 0.8% higher than the Threadripper 1920’s 5425. The EPYC also ranks at the 61st percentile of all CPUs, one point above the Threadripper’s 60th percentile.

Q: What is the memory configuration difference?

A: The EPYC 7351P supports eight-channel DDR4 memory with a bandwidth of 170.6 GB/s and ECC support. The Threadripper 1920 uses quad-channel DDR4 with 85.3 GB/s bandwidth and no ECC support, giving the EPYC a 100% memory bandwidth advantage.

Q: Are these processors unlocked for overclocking?

A: Yes, both the EPYC 7351P and the Threadripper 1920 have an unlocked multiplier, as indicated by the multiplierUnlocked field. However, the EPYC is listed as Active in production, while the Threadripper is End-of-life.

Q: Which chip has a larger L3 cache?

A: The EPYC 7351P has 64 MB of shared L3 cache, while the Threadripper 1920 has 32 MB. Both have identical per-core L1 (96 KB) and L2 (512 KB) caches.

Q: What is the transistor count difference?

A: The Threadripper 1920 uses 9,600 million transistors across two 213 mm² dies, while the EPYC 7351P uses 4,800 million transistors on a single 213 mm² die. This reflects the Threadripper’s dual-die design.

Where Each One Wins

The EPYC 7351P wins in every benchmark category represented in the data, making it the clear performance leader in this comparison. Its wins are driven by a combination of higher core count (16 vs 12), larger L3 cache (64 MB vs 32 MB), and significantly higher memory bandwidth (170.6 GB/s vs 85.3 GB/s). The EPYC’s advantage is consistent across all Cinebench tests, with a uniform 18% to 18.1% delta in both single-core and multi-core workloads.

The Threadripper 1920, despite losing all head-to-head matchups, still holds relevance in its intended desktop segment. It offers a lower TDP of 140 watts versus the EPYC’s 170 watts, and its dual-die design with 9,600 million transistors suggests a different packaging approach for enthusiast platforms. The Threadripper’s higher base clock (3.20 GHz vs 2.40 GHz) and boost clock (3.80 GHz vs 2.90 GHz) indicate it is tuned for frequency-sensitive desktop tasks, even though the benchmark data does not show a win in any tested workload.

In practical terms, the EPYC 7351P is the better choice for multi-threaded server and workstation workloads where its eight-channel memory and larger cache provide a structural advantage. The Threadripper 1920, with its higher clocks and lower TDP, may be more suitable for desktop workflows that do not scale with core count, but the data shows no benchmark where it outperforms the EPYC. The EPYC’s 18% lead in multi-core tests like Cinebench R23 (22135 vs 18756) underscores its dominance in heavily threaded applications.

Architecture Differences

Both processors are built on the Zen architecture using a 14 nm process from GlobalFoundries, but they diverge in implementation. The EPYC 7351P uses the Naples codename and is part of the EPYC 7001 series, featuring a monolithic 213 mm² die with 4,800 million transistors. The Threadripper 1920, codenamed Zen (Whitehaven), uses two 213 mm² dies with a combined 9,600 million transistors, effectively doubling the transistor count through a dual-die configuration.

The memory architecture is a key differentiator. The EPYC 7351P supports eight-channel DDR4 memory with a bandwidth of 170.6 GB/s and ECC memory, making it suitable for data integrity-critical server environments. The Threadripper 1920 supports quad-channel DDR4 with 85.3 GB/s bandwidth and lacks ECC support, aligning it with desktop use cases. The EPYC also has twice the L3 cache (64 MB vs 32 MB), which helps with large working sets in server workloads.

The sockets differ as well: the EPYC uses AMD Socket SP3, while the Threadripper uses AMD Socket SP3r2. Both are unlocked for overclocking, but the EPYC is listed as Active in production status, whereas the Threadripper is End-of-life. The EPYC is classified as a Server/Workstation part, while the Threadripper is Desktop, reflecting their intended platforms. The Threadripper’s PCIe configuration is listed as Gen 3 with 60 lanes (CPU only), while the EPYC does not specify lane count in the data.

Specification Differences

The core and thread counts differ significantly: the EPYC 7351P has 16 cores and 32 threads, while the Threadripper 1920 has 12 cores and 24 threads. Base and boost clocks also differ, with the Threadripper running higher at 3.20 GHz base and 3.80 GHz boost versus the EPYC’s 2.40 GHz base and 2.90 GHz boost. The TDP is higher for the EPYC at 170 watts compared to the Threadripper’s 140 watts.

Cache hierarchies are different in L3 size: the EPYC has 64 MB shared L3, while the Threadripper has 32 MB. Both have identical L1 (96 KB per core) and L2 (512 KB per core) caches. The memory bus is eight-channel for the EPYC versus quad-channel for the Threadripper, with corresponding bandwidth figures of 170.6 GB/s and 85.3 GB/s. ECC memory support is present on the EPYC but absent on the Threadripper.

The socket types differ (SP3 vs SP3r2), and the transistor counts are 4,800 million for the EPYC versus 9,600 million for the Threadripper. The die size is 213 mm² for the EPYC and 2x 213 mm² for the Threadripper. The market segments are Server/Workstation for the EPYC and Desktop for the Threadripper, with production statuses of Active and End-of-life, respectively. Both have unlocked multipliers and share the same 14 nm process node, foundry, and Zen architecture.

Head-to-Head Benchmarks

The EPYC 7351P wins all six head-to-head benchmark comparisons, with a consistent 18% to 18.1% delta in every test. In Cinebench R15 multi-core, the EPYC scores 2231 against the Threadripper’s 1890, an 18% lead. The single-core R15 test shows the same 18% delta, with scores of 314 and 266 respectively. This consistency suggests a fundamental architectural advantage rather than a workload-specific edge.

In Cinebench R20, the multi-core results are 9296 for the EPYC versus 7877 for the Threadripper, again an 18% difference. The single-core R20 test shows 1312 versus 1111, a slightly larger 18.1% delta. The pattern continues in Cinebench R23, where the EPYC scores 22135 in multi-core versus 18756 for the Threadripper (18% delta), and 3125 versus 2647 in single-core (18.1% delta). The EPYC’s advantage is remarkably uniform across all three Cinebench generations, indicating that the 16-core configuration and larger cache provide a consistent performance uplift.

The single-core results are particularly telling, as they show the EPYC’s advantage is not solely due to core count. Despite the Threadripper’s higher base and boost clocks (3.20/3.80 GHz vs 2.40/2.90 GHz), the EPYC still leads by 18% in single-threaded tests. This suggests that the EPYC’s larger L3 cache and eight-channel memory architecture contribute significantly to per-core performance, possibly through better memory latency characteristics. The Geekbench results, while not part of the head-to-head list, reinforce this: the EPYC scores 4607 multi-core and 733 single-core, though the Threadripper does not have Geekbench scores in the data.

The average benchmark scores confirm the head-to-head results, with the EPYC at 5469 and the Threadripper at 5425. The nearest rival data places the EPYC 0.5% ahead of the Intel Xeon W-1290P and 0.8% ahead of the Threadripper, while the Threadripper is 0.3% behind the Xeon and 0.8% behind the EPYC. This places both chips in a tight performance cluster, but the EPYC’s consistent 18% lead in Cinebench tests is the standout finding. The EPYC’s wins across all benchmark categories give it a clean 6-0 record, with no workload in the data where the Threadripper can claim an advantage.