Get the App
SLTechnology News&Howtos  ›  IT Information  › 

It is reported that Intel Arrow Lake-S processor large core L2 cache will be upgraded to 3MB

Shulou Source: shulou.com Published: 2023-11-24 18:43:54 10月06日 Update

Thanks to CTOnews.com netizen South China Wu Yanzu for the clue delivery! CTOnews.com August 13 news, a new generation of Arrow Lake-S series processors will use the Core Ultra (2) brand, is expected to be unveiled at the end of 2024 as the successor to the Meteor Lake series Core Ultra processor.

The flagship model of this series of processors will be based on the LGA 1851 socket, with an 8-core + 16-core design, the same as the current flagship i9- 13900K, with an expected CPU boost of up to 21%(compared to Raptor Lake), and the core performance of the new architecture will also double.

According to the description of the @ Golden Pig upgrade package, the L2 cache of each large core of the new generation Arrow Lake-S series processor will be increased to 3MB, which is 50% larger than Raptor Lake.

This means that Arrow Lake-S L2 cache is 2.4 times larger than Golden Cove, which is a huge boost for L2 cache. Other than that, the family CPU will retain the same L3 cache as Raptor Lake.

CTOnews.com comes with three generations of processors:

Arrow Lake-S (Lion Cove)-3 MB L2 cache per core

Raptor Lake-S (Raptor Cove)-2 MB L2 cache per core

Alder Lake-S (Golden Cove)-1.25 MB L2 cache per core

According to recent intelligence, the Intel Arrow Lake-S should debut with 800 series motherboards, based on Lion Cove (P core) and Skymont (E core), and Lunar Lake will also use the same architecture (expected to target low power platforms).

Arrow Lake-S is planned to use the "Intel 20A" process, but there have been rumors recently that Intel has decided to fully switch to TSMC 3nm for outsourcing production (the specific N3 or N3E process is unclear).

Tags: Cache processor processing same Intel Intel process next Generation flagship Architecture messaging upgrade huge clear Motherboard rumor Power consumption Daniel Wu Brand Model Apple Docker Huawei Linux macOS MariaDB Microsoft MySQL NVidia OPPO Reno Redmi macOS MariaDB Apple Microsoft