Intel has once again pushed back the full-scale launch of its latest generation of Xeon server CPUs, codenamed Sapphire Rapids.
At a conference earlier this week, the company’s head of data center silicon, Sandra Rivera, announced that Intel will need additional time to validate the new chips with different system configurations.
Rivera reassured customers that Sapphire Rapids chips will still be widely available later this year, but reported that unexpected bottlenecks have driven the company to further delay high-volume manufacturing.
Intel Sapphire Rapids delays
The rollout of Sapphire Rapids has been punctuated by repeated delays. Originally slated to launch in 2021, the new Xeon chips were first pushed back to early 2022. And later, Intel was forced to delay high-volume availability yet again, this time to the middle of the year.
The company delivered on its promise to put the new chips in the hands of select customers in Q1, but it now appears that most will have to wait until the end of Q3 to gain access to the silicon.
“With innovation comes a level of complexity, and we see all customers – OEMs, ODMs, the hyperscalers - have strong demand for platform validation with us,” explained Rivera. “At this point, we are building in more platform and product validation time, so we see the Sapphire Rapids ramp being later in the year than we had originally forecast.”
“Intel’s [third-gen Xeon Scalable] Ice Lake will be the highest-volume product for 2022 as we ramp Sapphire later this year and then of course throughout 2023,” she added.
However, Rivera was quick to dismiss concerns that repeated Sapphire Rapid hold-ups are indicative of a delay in the launch of the following generation of Xeon chips, codenamed Emerald Rapids.
She told the audience that Emerald Rapids chips are still scheduled for release in late 2023, bringing with them a “nice performance bump” across memory, networking and overall performance.
Via Tom’s Hardware
