Intel mentioned its next-generation Xeon Scalable CPUs will launch in January subsequent 12 months, shortly after a brand new report from Taiwan mentioned mass manufacturing of the server chips has been delayed till 2023.
The semiconductor big introduced a launch date of January 10 subsequent 12 months for the oft-delayed processors, codenamed Sapphire Rapids, on Tuesday. The corporate mentioned the mass-produced Sapphire Rapids models now meet product launch necessities and it has began ramping up manufacturing. This echoes statements made by Intel CEO Pat Gelsinger through the firm’s earnings name final week.
“We’re ramping up output as we communicate… We’re anticipating this to be our quickest Xeon to 1 million models. And we will push it fairly aggressively, and factories are ramping up as we communicate,” he mentioned.
Whereas Intel has introduced a launch occasion for Sapphire Rapids, it isn’t but identified when the processors shall be commercially out there to the final IT market subsequent 12 months. More and more in recent times, there have been gaps between product launch occasions and common availability.
Intel’s launch announcement comes after Taiwanese analysis agency TrendForce mentioned this morning that mass manufacturing of the Silicon Valley stalwart’s server chips had been moved from the fourth quarter of this 12 months to the primary half of subsequent. as a result of manufacturing issues.
TrendForce mentioned that Intel was experiencing efficiency points with the Intel 7 node, saying the present manufacturing throughput fee for Sapphire Rapids is estimated to be within the 50 to 60 p.c vary. The agency mentioned this was affecting fashions with increased core counts.
What a coincidence!
When Register When requested about these claims, an Intel spokesperson didn’t handle them instantly, as an alternative pointing to Tuesday’s launch announcement, which got here hours after the TrendForce report.
“As we share within the revenue, Sapphire Rapids quantity SKUs are certified for product launch and really sturdy quantity progress is predicted throughout all SKUs. In actual fact, we simply introduced our January tenth launch at this time,” the spokesman mentioned.
At a minimal, Intel’s January 2023 launch occasion for Sapphire Rapids signifies that the server processors will arrive early subsequent 12 months, after experiencing a number of delays over the past 12 months and a half.
AMD confirmed that it’s going to hit the market first with new server chips
Intel’s plan to launch Sapphire Rapids early subsequent 12 months means AMD will beat the corporate to the mainstream market with the upcoming Epyc competitor processors, codenamed Genoa, given what we have simply realized from a significant vendor. of servers.
Hewlett Packard Enterprise executives confirmed to Register that programs outfitted with 4th era Xeon Scalable won’t ship earlier than servers with AMD’s Genoa CPU. HPE mentioned the brand new Epyc-powered servers will ship later this month and would not but have a date for the arrival of the brand new Xeon machines.
TrendForce referred to as the delay of Sapphire Speedy mass manufacturing a “blessing” for AMD, which can now have the primary information center-class processors out there to assist DDR5 reminiscence, PCIe 5.0 connectivity, and Compute Categorical Hyperlink (CXL) 1.1 requirements. .
The analysis agency additionally says it now expects AMD’s x86 server CPU share in opposition to Intel to succeed in 15 p.c in 2022 and develop to 22 p.c in 2023.
A collection of delays, new questions concerning the well being of the Intel 7 node
The most recent delay implies that Sapphire Rapids will arrive a couple of 12 months after the corporate initially mentioned server chips could be out there to the final market.
Up to now, Intel has mentioned it could ship new Xeon CPUs to the cloud and different massive prospects in This fall 2021, then to everybody else within the first half of 2022. These targets had been later moved to Q1 and Q2 2022. . , respectively.
Intel then signaled one other delay when the corporate’s information heart chief, Sandra Rivera, informed traders in June that the corporate had pushed manufacturing of Sapphire Rapids “later within the 12 months than we had initially forecast.” . They nonetheless believed then that it may beat AMD to market with Sapphire Rapids, although Rivera warned that the delay would scale back Intel’s “management window.”
On the time, Rivera mentioned the delay was crucial for added “platform and product validation time,” noting that it was not the results of manufacturing points with the Intel 7 node, which was beforehand referred to as the superior 10 course of. nm of the corporate.
Nevertheless, the brand new TrendForce report raises questions concerning the standing of the Intel 7 node for Sapphire Rapids chips. The semiconductor big began utilizing Intel 7 for mass manufacturing with its Alder Lake chips for PCs and laptops final 12 months, and is now utilizing node for lately launched Raptor Lake processors, so it appears attainable that the issues of manufacture are particular to Sapphire Rapids and never all Intel 7 chips.
What separates Sapphire Rapids from Alder Lake and Raptor Lake is that the server chips would be the firm’s first CPU to undertake a chiplet structure. Alder Lake and Raptor Lake, alternatively, have monolithic dies, and the corporate will not transfer its PC processors to a chiplet-based design till Meteor Lake, which is scheduled for launch subsequent 12 months.
Ronak Singhal, a principal investigator at Intel, mentioned Register final month that the semiconductor big is prioritizing high quality with this era of Xeon processors.
“We won’t compromise on that high quality, even when it means delaying the product,” he mentioned. “We’re going via a rigorous validation cycle, and once we discover points, it generally pushes again the schedule, however we predict that is the fitting tradeoff for our prospects and what they anticipate from us.” ®
– Intel’s Sapphire Rapids Xeons allegedly ready for early 2023 • The Register
