More leaks regarding Intel’s upcoming Arrow Lake have emerged, and this time it comes from @OneRaichu (via Videocardz) detailing three of the K/KF-series models and their clock speeds. It appears that its P-Cores will be clocked slower than the preceding 14th Gen Core lineup, while the E-cores are getting a healthy boost in clock speeds.
Arrow Lake: Slower P-Core Clocks, Faster E-Core Clocks
To simplify, here’s all the numbers courtesy of the leaker:
SKU | P-Core Boost (vs. 14th Gen) | All-core Boost | E-Core Boost (vs. 14th Gen) | Core Count |
---|---|---|---|---|
Core Ultra 9 285K / 285KF | 5.7 GHz (-300 MHz) | 5.4 GHz | 4.6 GHz (+200 MHz) | 8P+16E |
Core Ultra 7 265K / 265KF | 5.5 GHz (-100 MHz) | 5.2 GHz | 4.6 GHz (+300 MHz) | 8P+12E |
Core Ultra 5 245K / 245KF | 5.2 GHz (-100 MHz) | 5.0 GHz | 4.6 GHz (+600 MHz) | 6P+8E |
While the P-Cores across all three (six) models feature reduced clock speeds, it’s very unlikely to see performance regression if the IPC (instructions per clock) improvements is big enough the mask the clock speed deficit. Notably, all models will feature E-cores clocked at 4.6 GHz, and assuming these are Skymont cores – which Intel has previewed in Computex – we could be looking at a fairly powerful cluster of efficiency-focused cores to bump up the processors’ overall multi-core performance.
We also know from this leak that Intel will continue to split its lineup between standard and graphics-less parts (F/KF-series), and all of them will lack hyper-threading. However, it is believed that the multi-core performance remains competitive, with notable uplifts seen in the recent qualification sample benchmark data.
Intel is expected to formally announce Arrow Lake under the Core Ultra 200 series in September this year, while general availability is expected to be in the following month.
Pokdepinion: I think the new E-cores will significantly close to performance gap to P-cores – does that mean Ryzen 9000 will be under threat? We’ll have to see.