Skip to main content
  1. Home
  2. Computing
  3. Opinion
  4. Features

Why we can’t get too excited about Intel Arc Alchemist discrete GPUs just yet

Add as a preferred source on Google

Going into CES 2022, I was focused on Intel Arc Alchemist. Since the announcement of Arc and the generations that will make it up, Intel has committed to a release date in the first few months of 2022. But at the start of the show this year, we knew almost nothing about Arc Alchemist outside of its name.

Unfortunately, Intel’s CES 2022 keynote didn’t shed much light. The company confirmed that Arc discrete GPUs will show up in 50 laptops and desktops throughout 2022, but still wouldn’t provide hard performance numbers or even so much as the number of cards in the range. The clock is ticking for Intel Arc Alchemist, and as murmurs of internal struggles mount, that’s a bad sign for the future of the company’s discrete GPUs.

As great as Intel Arc Alchemist may end up being — and I’m hoping it is — some optimistic skepticism is important.

Struggling to get Arc on track

Intel announces Intel Arc dGPU for Alienware x17 laptop.
Image used with permission by copyright holder

Intel hosted its CES keynote a day before the show floor opened, as it normally does. The company spent six minutes talking about Arc Alchemist, half of which was taken up by an Intel Deep Link demo — a technology Intel launched in 2020.

In that time, we learned that Intel is shipping cards to its manufacturing partners, that over 50 mobile and desktop designs are “coming soon,” and that XeSS, and upscaling tech similar to Nvidia’s Deep Learning Super Sampling (DLSS), is coming to Death Stranding Director’s Cut on PC. Intel also reaffirmed, once again, Arc Alchemist’s launch date of sometime during the first quarter of 2022.

Three days after the keynote, Intel scrubbed this release date from Arc Alchemist’s product page. It wasn’t just in a single spot, either. Anywhere the company mentioned “Q1 2022,” you’ll now see just “2022.” Normally, this wouldn’t be a big deal, but it follows weeks of rumors of internal delays and release date struggles at Intel.

Intel ARC DG2 FULL Lineup Leak: Pricing, Segmentation, Release dates, and Battlemage whispers!

Well-known leaker Moore’s Law is Dead detailed release date issues in a video released in December, with one of the sources saying the release date had changed about a dozen times. Another said something “seems off” about the way Intel graphics is managing and executing the Arc Alchemist launch.

It’s never a good idea to take rumors as law. But given how tight-lipped Intel has been about a product line that should be arriving in a few short weeks, I’m skeptical. Even if the cards launch in the first few months of the year, it’s looking the like the majority of the range will arrive later in the year.

That’s not inherently a bad thing. Delays, in most cases, are a good thing, allowing designers to spend more time to get a product right instead of just releasing it into the wild with flaws. That doesn’t seem like Intel’s approach here, though. Given the reinforcement of the release date, it seems like Intel wants to have its cake and eat it, too — meet the promised release date while delaying most of the range in the background. I’m speculating, but that’s the impression I was left with after Intel’s CES keynote.

If you’ve got it, show it

A render of an Intel Arc Alchemist graphics card.
Image credit: Wccftech Image used with permission by copyright holder

You can try to glean information from cryptic marketing and the telephone game of release date speculation, but you don’t need to go that far. If Intel could show off Arc Alchemist at CES, it would have. Or the very least, it should have.

Intel dumped a load of Arc Alchemist information at Architecture Day in August 2021. We learned about XeSS, the N6 manufacturing process, and the Xe-HPG architecture powering the range. And since then, everything we’ve heard has been piecemeal and drawn out to the point that even minor announcements, like a single game supporting XeSS, seem important.

Even without specs, Intel hasn’t even so much has hinted at what Arc Alchemist is trying to accomplish. Where do the cards fit in the current GPU market? Can buyers expect flagship, midrange, or low-end performance? How many cards are even in the range? I could spend a couple of hundred words just on rhetorical questions, and that’s not good for a product that’s due out at any moment.

Intel isn’t typically shy about sharing performance data. Even if Intel wasn’t ready to share concrete numbers, I still expected info on relative performance —  a vague graph showing the competitors to the flagship card, or a line graph showing how the performance scales across the range.

These numbers generally don’t mean anything, and they’re always skewed to favor the brand presenting them. But they set expectations, and they show that a product is making progress toward a goal promised by the company.

Hold your breath

LEDs form the Intel ARC logo.
Image used with permission by copyright holder

Intel Arc Alchemist aims to shake up the AMD/Nvidia duality that has dominated PC graphics for more than two decades, but CES was a worrying sign for what’s supposed to be a disruptive product. At this point, Intel isn’t disrupting much. The company has thrown its hat in the ring, but that’s it.

I’m not discounting Arc Alchemist, and truth be told, I’m rooting for Intel to bring some much-needed competition to graphics. After CES, though, my enthusiasm has waned. I’m looking forward to Arc Alchemist, but I’m remaining skeptical until Intel will share some more concrete information about the range.

That doesn’t mean I’ve lost all of my enthusiasm about Arc Alchemist. XeSS and hardware-accelerated ray tracing are a winning duo, and Intel has vastly improved its graphics products over the last few years. Arc Alchemist looks great from what we know — but that’s the problem right now; we don’t know much.

Jacob Roach
Former Lead Reporter, PC Hardware
Jacob Roach is the lead reporter for PC hardware at Digital Trends. In addition to covering the latest PC components, from…
The best password managers for 2026
have i been pwned owner uncovers 13 million plaintext passwords leaked from free webhost is a safe password even possible we

Passwords are still a fact of digital life, even as passkeys slowly start to change how we sign in to our accounts. Apple, Google, and other platforms have also made their built-in password managers much more capable, giving people more options than ever for keeping track of their credentials. A dedicated password manager still has an important advantage, though: it can bring passwords, passkeys, two-factor authentication, secure notes, and shared credentials together across the different devices and platforms you use.

The best password managers also make good security habits easier to maintain. They can generate unique passwords instead of leaving you to come up with another variation of the same one, flag compromised credentials, and make it easier to share access without passing passwords around in messages. For families and people who regularly move between operating systems, they can be particularly useful, while privacy-focused services offer another reason to look beyond the tools built into your phone or browser.

Read more
Anthropic wants everyone to take a chill pill at cooking worryingly powerful AI models
Anthropic’s CEO thinks frontier AI is advancing faster than its safeguards, and his proposed fix amounts to giving the industry a speed limit
Claude Anthropic Featured

Anthropic CEO Dario Amodei thinks the AI industry is moving too fast for its own safety work to keep up. He still wants more powerful AI. He just wants companies to take longer getting there.

In a new essay, Amodei is calling for frontier AI companies to deliberately slow how quickly their models improve. The extra time would go toward understanding stronger systems and making sure safeguards work before another leap arrives. He argues AI progress would still remain fast.

Read more
OpenAI AI agents were linked to a cyberattack on RubyGems before the Hugging Face incident
Rogue AI concerns grow after OpenAI agents disrupt RubyGems in May attack
OpenAI logo on Microsoft surface

Artificial intelligence agents being tested by OpenAI were involved in a previously undisclosed cyberattack against RubyGems in May, an incident that is now raising uncomfortable questions about how much control humans really have over increasingly autonomous AI systems.

The attack overwhelmed RubyGems, a popular service used by software developers to publish and access Ruby packages, forcing operators to suspend new account registrations for four days. According to a report by The Wall Street Journal, OpenAI confirmed that its agents had been involved, but said they were using the platform to perform benign tasks and retrieve publicly available information during a training run.

Read more