Intel 48 core new Xeon running score exposure: against 3D cache version OF AMD Zen3 results unexpected
According to Mercury Research, AMD’s x86 processor share reached a new high of 25.6% in the fourth quarter of 2021, with server products contributing the most with 10.7%.Obviously, this means that next, Intel and AMD in the server market “fight” will be more intense.AIDA64 is based on Intel 7 (10nm) process, Socket E (LGA-4677) interface, and level 3 cache of 90MB. AIDA64 is based on Intel 7 (10nm) process and Socket E (LGA-4677) interface.270W power consumption, support AVX-512 instruction set, etc.The base band is 2.3GHz, and the acceleration is 3.3GHz. However, the 48-core doesn’t seem to be the flagship in the Sapphire Rapids family, and there is also said to be a 56-core 112-thread monster.Also announced is the running score, which uses a dual-channel platform, compared to the Zen3 3D cache Milan-X (EPYC 7773X), with 64 cores and 128 threads, and the current Ice Lake-SP (Xtron Platinum 8380).Running points no longer list specific test screenshots, directly on the comparison chart.In Cinebench R15 runs, the Sapphire Rapids maintained a single-core advantage, outperforming the EPYC 7773X by about 26 percent.However, because of the thread size gap, the latter’s multi-core performance exceeded, the gap of 12%, it depends on the follow-up flagship of Intel 56 cores can reverse attack.Then came memory and processor cache latency tests. Intel’s new and old platforms both performed well, but in processor cache, MILan-X, supported by 3D cache, is a ubiquitous presence, showing its own unique advantages.Finally, two versions of CPU-Z were tested, v17.01.64 and V19.01.64, and surprisingly, EPYC 7773X beat Intel by a wide margin.