A pair of prototype Intel Xeon Platinum 8468 Sapphire Rapids 48 core server processors have been not too long ago examined in Geekbench 5, as shared by @BenchLeaks on Twitter. They featured some spectacular efficiency figures, matching AMD’s better of the perfect twin EPYC 64-core configuration in the identical benchmark – and these chips aren’t even the flagship 56-core variants.
The Sapphire Rapids system shared was geared up with twin Xeon Platinum 8468 CPUs that includes a complete of 96 cores and 192 threads. The 2 chips mixed outputted a single-threaded rating of 1,257 factors, and a mlti-threaded rating of 74,586 factors.
One of many highest performing twin AMD EPYC processor configurations we may discover on Geekbench 5 for comparability options twin AMD EPYC 7763 processors with a mixed core depend of 128 cores and 256 threads. The 2 chips hit 1249 factors within the single-threaded take a look at and 75,539 within the multi-threaded take a look at.
This leads to a distinction of 1% in each the only and multi-core for each platforms, basically tieing them with equal efficiency. However, if we take a look at the person efficiency of every core in the course of the multi-threaded workload, Intel is 23% sooner per core in comparison with AMD.
That is necessary to notice as a result of single-threaded benchmarks solely inform you the utmost efficiency of a person core when all the opposite cores are idle, permitting that single core to make use of extra energy and thermal reserves. Because of this, particular person core efficiency underneath load exhibits us what per core efficiency appears like underneath most load.
But it surely’s price noting that Geekbench 5 isn’t a very helpful benchmark for real-world efficiency outcomes. So take this knowledge with a grain of salt. On the flip facet, Geekbench can at the very least give us a common thought of Sapphire Rapids’ efficiency.
A Fast Recap on Sapphire Rapids and Its Delays
Sapphire Rapids is Intel’s upcoming server structure, designed to exchange Intel’s present 14nm Cascade Lake Xeon lineup. Because of this, Sapphire Rapids will characteristic some critical upgrades, together with Intel’s cutting-edge Golden Cove cores working on the Intel 7 (10nm Enhanced SuperFin) node – similar to Intel’s present Alder Lake elements.
Like Alder Lake, PCIe 5.0 and DDR5 will likely be supported (with some elements sporting HBM compatibility). Nonetheless, not like Alder Lake, it won’t assist effectivity cores. From what we all know, Sapphire Rapids will sport a most of 56 cores. However because of multi-socketed motherboards, core counts can go as much as 112 or extra. Sapphire Rapids may even sport further options like AMX, and Intel DSA, which will likely be unique to its server structure.
Sadly, Sapphire Rapids has allegedly was a bug-infested nightmare for Intel, with a number of delays setting again the platform’s launch window. Initially, Sapphire Rapids was meant to face off in opposition to AMD’s Zen 3 EPYC Milan processors (like those we confirmed off on this article), however because of a whopping 500 bugs discovered within the microarchitecture, the launch window has moved to early 2023.
This implies Sapphire Rapids might want to cope with present Zen 3 elements and AMD’s future Zen 4 EPYC processors codenamed Genoa in 2023. Fortunately, it seems Intel’s future server structure can at the very least grasp with AMD’s Zen 3 rivals and outperform it by a modest quantity. However there is not any telling the way it will carry out in opposition to AMD’s future Zen 4 structure, which is already releasing within the type of Ryzen 7000.