Sunday, May 22, 2022

Intel seems on track to solve its core problems

Must read

Shreya Christina
Shreya has been with for 3 years, writing copy for client websites, blog posts, EDMs and other mediums to engage readers and encourage action. By collaborating with clients, our SEO manager and the wider team, Shreya seeks to understand an audience before creating memorable, persuasive copy.

Intel is finally back on track.

That’s the biggest benefit of the first year of the company’s reboot, brought by new CEO Pat Gelsinger when he took over the reins in early 2021. Intel has a new vision for the future: the “IDM 2.0” approach. The introduction was exactly one year ago today. That makes today a great time to check in and see how the new Intel — which looks a lot like the old Intel — is taking shape.

The past decade or so, leading up to Intel’s recent reinvention, has been a series of misplaced bets and product misses, culminating in a disastrous summer in 2020 when almost everything went wrong for the company: another delay for its next major manufacturing hub, departure. from top executives, and the loss of one of its most premium partners all struck within weeks of each other.

Even Intel seemed to realize that something had to be done: former CEO Bob Swan was replaced by Gelsinger — an accomplished Intel chip designer who had helped develop some of the company’s previous critical chip innovations.

Intel spent years and resources on expensive and flashy new market bets, such as wearable devices in 2014, drones and robots in 2015, VR in 2016, self-driving cars in 2017, smart glasses in 2018 and 5G in 2019. But those days are over. The way forward would be chips and in particular IDM 2.0. Intel would rethink how and for whom it would make processors. And it would emphasize its strength as one of the few remaining integrated device manufacturers in the world that designed, manufactured and sold semiconductor products.

IDM 2.0 would consist of three aspects: strengthening Intel’s in-house manufacturing, expanding the use of third-party foundries from 2023, and Intel Foundry Service – a new standalone business unit that would let Intel build chips for other fab chipmakers such as Qualcomm, something it had never done before in its history.

Intel’s 2021 finally feels like a company on time and on track. The company was able to release its Alder Lake chips (for both laptops and desktops) after years of delay, with its new hybrid architecture and some of its most powerful processors in years. It set out an extremely ambitious roadmap to try to recapture silicon supremacy by 2025 with a return to the rapid cadence of technological advancement the company used to be known for, with its Alder Lake chips already kicking off that process. And it announced plans for billions of dollars in investments to expand its semiconductor production: $20 billion in its existing Arizona facilities, another $20 billion to build a brand new site in Ohio, and $17 billion to build more. a new factory in Germany.

After spending 2020 on the brink of disaster, Intel seemed to be spending all of 2021 doing and saying the right things. And 2022 — and Gelsinger’s second year as CEO — also seems full of potential: Intel’s Arc GPUs are poised to launch for laptops and desktops in the coming months, launching the company into a new product category where it (ultimately) ) on long-term gaming heavyweights like AMD or Nvidia. And the company plans to finally start production on its long-delayed 7nm node that has been the source of so much drama over the years (although it will be curiously branded “Intel 4” to better position its products against from competitors).

Image: Intel

Other areas, such as Intel’s use of third-party foundries from TSMC and Samsung, have yet to really come to fruition. Intel’s upcoming Arc GPUs will be built by TSMC on its more advanced 6nm node, although this is one of the most notable Intel products built by a third-party partner to date. And presumably we’ll be hearing more about that front in the coming months as we get closer to that first 2023 date.

Similarly, Intel Foundry Service is still building business, and Intel has also announced some big deals up front: It will make Qualcomm chips in the future on its upcoming next-generation 20A process (which will launch sometime in 2024). debut at the earliest), in addition to offering packaging solutions for Amazon’s AWS.

All in all, it really looks like Intel has managed to stop its downward spiral and could rise again. Products are starting to come out on time and the roadmap looks encouraging, with a rapid cadence of product releases the likes of which Intel has failed in over a decade.

But there is still much work to be done. Intel’s competition from AMD and Apple has never been stronger. AMD announced new laptop chips at CES 2022 and is gearing up for the next generation of Ryzen 7000 desktop chips later this year, both of which aim to surpass Intel’s best in performance and power efficiency. And Apple’s excellent M1 chips have made the leap to even more powerful desktop models with the M1 Ultra. The Cupertino company is now one computer away from completing the transition of Intel’s processors.

There is also the technological downside of Intel. As of today, Intel is not making the most advanced semiconductors in the world, having long relinquished that title to TSMC and Samsung. Its competitors use more advanced manufacturing techniques (such as extreme ultraviolet lithography) that Intel is just beginning to adopt and build chips with more transistors than Intel’s most advanced products. Intel’s renewed ambition to catch up is a good sign, and the impressive roadmap goalposts reflect that. But it is still catching up with competitors like TSMC, whose plans for a $100 billion investment expanding production over the next three years far exceeds even Intel’s own ambitions.

In addition, there is the ongoing shortage of semiconductors to contend with. Demand for chips still far outstrips supply, and Intel’s big plans to ramp up production and build all of these new plants will take time—the Arizona expansions won’t be operational until 2024, the Ohio plant by the end of 2025 and the German fab in 2027. That means it will be years before we can see whether those billion-dollar bets really pay off or not.

Looking to the next year of Intel’s last livelihood is a much more encouraging prospect than it was a year ago. And while there are still plenty of challenges for the company – especially when it comes to actually meeting the roadmap goals – it finally feels like things are moving in the right direction for Intel.

More articles


Please enter your comment!
Please enter your name here

Latest article