AMD EPYC 9115 vs Intel Core 9 273PQE Comparison

Head-to-Head Benchmarks

The recorded data presents a lopsided contest. The AMD EPYC 9115 claims 14 of the 17 head-to-head wins, while the Intel Core 9 273PQE manages only 3. Yet the margins tell a more nuanced story than the win count suggests.

Starting with the Cinebench suite, the EPYC 9115 maintains a consistent 6.7% advantage across all six tests. In Cinebench R23 multicore, the EPYC scores 42003 against the Intel part's 39190. The single-core R23 result follows the same pattern: 5929 for AMD versus 5532 for Intel. This uniformity across both single and multi-threaded workloads indicates a fundamental architectural edge rather than a workload-specific quirk.

The Passmark suite reveals where each processor flexes its muscles. The EPYC 9115 dominates in several specialized workloads. Its biggest victory comes in Passmark physics, scoring 4188 against the Intel's 2754, a 34.2% margin. Prime number finding shows a 31.5% gap (289 versus 198), and random string sorting favors AMD by 24.2% (70151 versus 53167). Extended instructions also lean heavily toward the EPYC, with a 14.8% lead (45477 versus 38743).

Data encryption shows an 11.5% advantage for AMD (33489 versus 29636), while integer math favors the EPYC by 9.4% (181807 versus 164629). Data compression is closer, with AMD ahead by just 2.4% (600099 versus 585752). Multithread performance in Passmark gives AMD a 5.8% edge (48936 versus 46107).

The Intel Core 9 273PQE, however, posts two notable victories. The most striking is Passmark single-thread, where it scores 4573 against the EPYC's 3360, a massive 36.1% lead. This is the largest delta in either direction across the entire benchmark set. Intel also wins floating-point math, scoring 125546 versus 113853, a 10.3% advantage.

The overall average benchmark scores reflect this split. The EPYC 9115 averages 69288 across all tests, while the Core 9 273PQE averages 66099. That places the AMD part in the 94th percentile of all CPUs, versus the 93rd percentile for Intel. The nearest rivals for each part underscore their positioning: the EPYC 9115 sits within 1% of the Intel Core i7-14700K and AMD Ryzen 9 7950X, while the Core 9 273PQE trades blows with the Intel Core Ultra 5 250KF Plus and AMD Ryzen 9 7950X3D.

The Verdict

The data paints a clear picture for most workloads: the AMD EPYC 9115 is the stronger processor. It wins the majority of tests, often by substantial margins, and its average benchmark score of 69288 exceeds the Intel's 66099 by roughly 4.8%. For users running physics simulations, prime number calculations, encryption workloads, or any task that scales with thread count, the EPYC 9115 is the obvious choice based on the recorded measurements.

The Intel Core 9 273PQE, however, holds a decisive advantage in single-threaded performance. Its 36.1% lead in Passmark single-thread is not a marginal difference; it is a categorical gap. Any application that relies heavily on single-core speed, such as certain legacy software or latency-sensitive workloads, would favor the Intel part. Additionally, the Intel processor wins floating-point math, which could matter for specific scientific or financial computations.

The market segments reinforce this split. The EPYC 9115 is a server and workstation part with 16 cores and 32 threads, while the Core 9 273PQE is a desktop processor with 12 cores and 24 threads. The EPYC's 4 nm process from TSMC, its 64 MB of shared L3 cache, and twelve-channel memory support position it for data-center duties. The Intel part's 10 nm process, 36 MB of L3 cache, and dual-channel memory support target desktop workloads where single-thread responsiveness matters.

Architecture Differences

The two processors come from fundamentally different design philosophies. The Intel Core 9 273PQE uses the Bartlett Lake codename and is built on Intel's 10 nm process node, manufactured in-house by Intel. The AMD EPYC 9115 belongs to the EPYC 9005 series, uses the Zen 5 architecture with the Turin codename, and is fabricated on TSMC's 4 nm process.

The transistor counts differ substantially. The EPYC 9115 packs 16,630 million transistors across two chiplets, each measuring 70.6 mm². The Intel part does not have recorded transistor or die size data in the database, but its monolithic approach on 10 nm suggests a different scaling strategy.

Cache hierarchies diverge significantly. Both parts allocate 80 KB of L1 cache per core, but the similarities end there. Intel provides 2 MB of L2 cache per core, while AMD provides 1 MB per core. The L3 cache tells a different story: Intel has 36 MB shared across all cores, while AMD offers 64 MB shared. This larger L3 pool on the EPYC likely contributes to its strong performance in data-heavy workloads like random string sorting and data compression, where it leads by 24.2% and 2.4% respectively.

Memory architecture differs at a fundamental level. The EPYC 9115 supports twelve-channel DDR5 memory with a staggering 576.0 GB/s of bandwidth. The Core 9 273PQE supports dual-channel DDR4 and DDR5, with a recorded bandwidth of 89.6 GB/s. This 486.4 GB/s difference in memory bandwidth is the largest architectural gap between the two parts and directly explains the EPYC's dominance in bandwidth-sensitive tasks.

PCIe connectivity also differs drastically. The EPYC 9115 provides Gen 5 with 128 lanes from the CPU, while the Intel part offers Gen 5 with only 16 lanes. For workloads that require high-throughput I/O, such as network processing or storage arrays, the EPYC's lane count is a decisive factor.

Specification Differences

The core and thread counts differ: the Intel Core 9 273PQE has 12 cores and 24 threads, while the AMD EPYC 9115 has 16 cores and 32 threads. This 33% advantage in core count for AMD aligns with its multicore benchmark wins.

Clock speeds present an interesting inversion. The Intel part has a base clock of 3.40 GHz and a boost clock of 5.90 GHz. The EPYC 9115 runs at 2.60 GHz base and 4.10 GHz boost. Despite the EPYC's lower clocks, it wins the majority of benchmarks, suggesting that its Zen 5 architecture and larger cache compensate for the clock deficit. The Intel part's 5.90 GHz boost clock likely drives its single-thread dominance.

Both processors have a 125 W TDP, but they use different sockets: Intel Socket 1700 versus AMD Socket SP5. The Intel part includes integrated graphics (UHD Graphics 770), while the EPYC 9115 has no integrated graphics, reflecting its server orientation.

Memory support differs: the Intel part accepts both DDR4 and DDR5, while the EPYC supports only DDR5. The EPYC's twelve-channel bus versus Intel's dual-channel bus creates the massive bandwidth disparity noted earlier. Both support ECC memory.

The Intel part has a launch MSRP of $589, while the EPYC 9115 has a launch MSRP of $726. The EPYC was released in October 2024, while the Intel part came later in March 2026.

FAQ

Q: Which processor wins more benchmarks overall?

A: The AMD EPYC 9115 wins 14 out of 17 head-to-head benchmarks, while the Intel Core 9 273PQE wins 3.

Q: How large is the Intel part's single-thread advantage?

A: In Passmark single-thread, the Intel Core 9 273PQE scores 4573 versus the EPYC 9115's 3360, a 36.1% lead. This is the largest margin in either direction across all recorded tests.

Q: Where does the EPYC 9115 show its biggest wins?

A: The EPYC's largest victories are in Passmark physics (34.2% ahead), prime number finding (31.5% ahead), and random string sorting (24.2% ahead).

Q: How do the memory bandwidth figures compare?

A: The EPYC 9115 provides 576.0 GB/s via twelve-channel DDR5, while the Intel Core 9 273PQE provides 89.6 GB/s via dual-channel DDR4 or DDR5.

Q: What are the core and thread counts?

A: The Intel Core 9 273PQE has 12 cores and 24 threads. The AMD EPYC 9115 has 16 cores and 32 threads.

Q: Which processor has a higher average benchmark score?

A: The EPYC 9115 averages 69288 across all tests, placing it in the 94th percentile. The Core 9 273PQE averages 66099, placing it in the 93rd percentile.

Where Each One Wins

The AMD EPYC 9115 is the clear winner for multi-threaded and server-oriented workloads. Its 16 cores and 32 threads, combined with 64 MB of L3 cache and 576.0 GB/s of memory bandwidth, deliver consistent advantages across Cinebench multicore tests (6.7% ahead), Passmark multithread (5.8% ahead), integer math (9.4% ahead), and data encryption (11.5% ahead). The physics benchmark result, where the EPYC leads by 34.2%, suggests it excels in simulation and computational physics tasks. Its 128 PCIe Gen 5 lanes make it suitable for systems requiring extensive I/O expansion.

The Intel Core 9 273PQE wins in single-thread performance by a wide margin. Its Passmark single-thread score of 4573 versus 3360 indicates superiority in applications that cannot leverage multiple cores effectively. The floating-point math win (125546 versus 113853, a 10.3% edge) makes it attractive for specific scientific computing tasks that rely heavily on FPU throughput. Its 5.90 GHz boost clock, the highest recorded between the two, likely drives these wins. The inclusion of integrated graphics and support for both DDR4 and DDR5 memory makes it a more flexible desktop option.

For users deciding between these two, the choice hinges on workload type. Server deployments, virtualization, data processing, and heavily threaded applications favor the EPYC 9115. Desktop use cases with single-threaded legacy software, or applications where floating-point math is the bottleneck, favor the Core 9 273PQE. The EPYC's higher average benchmark score and 94th percentile ranking suggest it is the more capable processor overall, but the Intel part's single-thread dominance cannot be ignored for certain use cases.