Intel's Bartlett Lake lineup is one of the more interesting CPU families to come out of Intel in recent years. The chips are not available on consumer platforms and are instead targeted at embedded and edge computing applications. What makes them stand out is their pure P-core design, with no E-cores in sight. The flagship of the lineup is the Core 9 273PQE, which packs 12 P-cores running at up to 5.9 GHz, 36MB of cache, and a 125W TDP, all based on the same Raptor Cove architecture found in Raptor Lake.

Since Bartlett Lake chips are not supported on consumer motherboards, enthusiasts have had to get creative. We previously covered how a modder successfully booted the Core 9 273PQE into Windows on a standard Z790 motherboard, an impressive feat given the platform's limitations. Shortly after, we also covered benchmarks from German reviewer Zed Up Gaming, which showed the 273PQE outpacing the Core i9-14900K by up to 10% in gaming. That was a genuinely exciting result for a chip that Intel never intended for consumer desktops.
Now, PC Games Hardware has published a much more thorough set of gaming benchmarks, and the results tell a very different story. Using a proper Bartlett Lake-compatible workstation board, the ASRock IMB-X1714 with a W680 chipset, and DDR5-5600 C46 memory, the 273PQE was tested across roughly 15 games. Across those tests, the chip failed to pull ahead of the Core i9-13900K, a processor that is now over four years old.

Looking at the actual numbers from PCGH, the 273PQE averaged around 64.8 FPS across the tested games, while the Core i9-13900K averaged approximately 69.9 FPS on the same platform, putting the older chip around 8% ahead. In the broader ranking, the chip lands somewhere between the Core i5-14600K and the Ryzen 7 7800X3D in gaming, which is not where you would expect a 12 P-core flagship to end up.
The gap between these results and the earlier Zed Up numbers likely comes down to the platform and memory situation. Running the chip on a consumer Z790 board with faster, better-tuned memory gives it a meaningful advantage that the native embedded platform cannot replicate, given the very restricted DDR5-5600 C46-qualified vendor list.
This also goes some way toward explaining why Intel chose not to bring Bartlett Lake to consumers in the first place. If 12 P-cores cannot consistently outperform an older 8 P-core consumer chip in gaming, the case for a mainstream launch becomes very hard to make.




