Apple’s M1 chip is Apple’s fastest single-core CPU benchmark on a Mac, and it beats many high-end Intel competitors in multi-core performance. Developer Erik Engheim recently shared an in-depth look at the M1 chip, exploring why Apple’s new processor is so much faster than the Intel chip it replaces. M1 is not a CPU! First, the M1 is not a simple CPU. As Apple explains, itR...
A few days ago, we have seen a 6-core Model of a suspected Intel 10nm Ice Lake-SP server platform from the GeekBench test database. Foreign media said that compared with the existing 14nm products, the multi-threading performance has increased by as much as 118%, which is quite incredible, but the truth is Sex also remains to be tested.
Now, a new Intel 10nm Ice Lake-SP Xeon processor has appeared in the SiSoftware database. The model is unknown, only the word Intel 0000 represents the engineering sample. In terms of specifications, there are 14 cores and 28 threads, and the core base frequency is 2.0GHz (acceleration frequency). Unknown), L2 cache 17.5MB, L3 cache 21MB.
The current second-generation Xeon Scalable family code-named Cascade Lake-SP does not have 14 cores, but the first generation earlier, but the L3 cache is only 19.25MB, so this new U can only be the next-generation 10nm ‘s new product.
So why not the next-gen 14nm Cooper Lake? because of performance.
The test results show that the new 14-core 28-thread arithmetic performance is 360.36GOPS, the multimedia performance is 1471.17Mp/s, and the high-precision encryption performance is 23.04GB/s.
what does this mean? Compared with a second-generation scalable Xeon Gold Medal 6132, with two channels consisting of 28 cores and 56 threads, with a main frequency of 3.2GHz, the arithmetic performance that can basically represent the architectural capability is 749.80GOPS. Up to 54% ahead!
If you look at multimedia performance and encryption performance, the same-frequency improvement is actually larger, reaching about 70% and 100% respectively!
Although Intel’s 10nm process performs generally on the ultra-thin notebook platform code-named Ice Lake, the frequency does not go up, resulting in average overall performance. It has to continue to rely on 14nm to support the scene, but it seems that after continuous optimization, it can finally be used in more The core server platform has come into its own.
For a 14-core processor, 2.0GHz is still qualified as a reference frequency, it depends on how much acceleration can be achieved, and whether it can break through the current 28 cores…