Intel’s new Xeons, i.e. the scalable server processors of the 4th generation, are delayed again. Once again, one might say, and the reasons are very diverse. The problems and errors are collected under NDA in internal documents (“so called NDA sights”) and currently amount to almost 500 (!) entries, with the trend continuing to rise. With a total of 12 (!) steppings, they haven’t exactly covered themselves with glory either, and I can’t remember any project so far that needed so many steppings before you could even use it to some extent. This started with A0 and A1, then proceeded via B0, C0, C1, C2, D0, E0, E2, E3 and E4 until currently stepping E5! The market launch was planned about 1.5 years ago and the plan was recently updated again.
Sapphire Rapids will be based on a chiplet design with four 15-core dies/tiles connected via Intel’s advanced EMIB interconnect. A total core count of up to 60 cores is rumored, which is admittedly a first for Intel’s server offerings and a remarkable step forward when compared to Ice Lake-SP. Sapphire Rapids is thus one of Intel’s most innovative server designs, but the grotty implementation and repetitive delays will probably do the CPU in. Sapphire Rapids was originally supposed to compete against AMD’s Epyc Milan processors, but due to the delay, they now have to compete against AMD’s newer Genoa lineup.
Sapphire Rapids uses the Intel 7 process, while Genoa is based on TSMC’s 5nm N5 node. Furthermore, Sapphire Rapids is limited to 60 cores, while Genoa even offers up to 96 cores. AMD’s server offerings also support faster memory, a larger number of PCIe Gen 5 lanes and much larger cache reserves. That’s going to hurt for sure.
Until now, Intel’s customers and partners assumed that the chips would be available in the second half of 2022, but internal information is now becoming a bit more concrete. Intel has now announced the “launch window” for Sapphire Rapids (SPR) for calendar week 6 to 9 (Feb. 6, 2023 to March 3, 2023), while the first shipment to selected recipients is still scheduled for 2022 in two waves. Calendar week 42 for the smallest models (2S) and calendar week 45 for the larger models (4 and 8S) are being speculated.
Thanks to the MCM design, Sapphire Rapids will have larger margins and higher ASPs, which of course nominally means higher profits for Intel and its supply chain partners. But the delay, which has now been concretized, will definitely have a negative impact on this product line, as both demand and competition will increase drastically. AMD’s Epyc Genoa processors of the 4th generation. The 3rd generation will offer top performance in both single and multi-thread operation, so Intel is actually only left with the more expensive HBM variant, which in turn could also fall behind Genoa-X, the 3D V cache variant of Genoa.
Also, yes, there is ARM, as the sheer mass of good ARM designs in the server market is another cause for concern for Intel, as these chip makers often sell at much lower prices and with much higher compute densities. Therefore, if Sapphire Rapids does indeed still come to market next year, Intel’s potential customers and partners may have already accepted competing offers. Sapphire Rapids is more and more mutating into a penny ditch for Intel, while the still relatively young graphics card division should also be highly loss-making. Intel is certainly not on the brink of financial collapse, but things are getting really tight. Emerald Rapids was actually planned for 2023 and could share a similar fate, because so far nothing is running as perfectly as it should and the NDA-sights don’t make friends feel comfortable.