Intel Briefly Shows Lunar Lake Chip; Next-Gen Mobile CPU Uses On-Package Memory


As part of the final segment of Intel’s “Intel Client Open House Keynote” at CES this afternoon, Intel EVP and GM of the Client Computing Group, Michelle Johnston Holthaus, also offered a brief update on Intel’s client chips in the work for the second half of the year. While no demos were run during the relatively short 45 minute keynote, Holthaus did reiterate that both Arrow Lake for desktops and Lunar Lake for mobile were making good progress and were expected to launch later this year.

But in lieu of black box demos we got something more surprising instead: our first look at a finished Lunar Lake chip.

Briefly holding the chip out for viewers to see – and holding the press away least they get too close – Holthaus pulled out a finished Lunar Lake chip.

While details on Lunar Lake still remain very slim – Intel still hasn’t even confirmed what process nodes it’s using – the company has continually been reiterating that they intend to get it out the door in 2024. And having silicon to show off (and shipping to partners, we’re told) is a very effective way to demonstrate Intel’s ongoing progress.

Of note, seeing the chip in person confirms something we’ve all but been expecting from Intel for a few years now: CPUs with on-package memory. The demo chip has two DRAM packages on one of the edges of the chip (presumably LPDDR5X), making this the first time that Intel has placed regular DRAM on a Core chip package. On-package memory is of particular interest to thin & light laptop vendors, as it allows for further space savings and cuts down on the number of critical traces that need to be routed from the CPU and along the motherboard. The technique is most notably (though far from exclusively) used with Apple’s M series of SoCs.

Beyond showing off the physical chip, Holthaus also briefly talked about expected performance and architecture. Lunar Lake is slated to offer “significant” IPC improvements for the CPU core. Meanwhile the GPU and NPU will each offer three-times the AI performance. How Intel will be achieving this remains unclear, but at least on the GPU side, we know that they’ve yet to offer XMX matrix cores within an integrated GPU.

No doubt this is far from the last time we’ll hear about Lunar Lake ahead of its launch. But for now, it’s a bit of a look into the future while Intel continues to ramp production on Meteor Lake for what is now the latest generation of laptops and other mobile devices.



Source link