Intel to boost gross margins – new products must deliver 50% gross profit | Hacker News Reader

Intel to boost gross margins – new products must deliver 50% gross profit

tomshardware.com

63 pointswalterbell60 comments

I read the title and all I could think of was Michael Scott, “I declare bankruptcy!” Hoping to will reality to your need.

More seriously, isn’t this signaling that all experimental R&D is out? Nothing but incremental progress can deliver anything but an approximation of profit margin.

I think Robert California telling Andy Bernard to double profits by the following quarter would also fit here.

I think this is a better fit. Just need an incredible string of good luck in a trivia night to make it a reality!

Double, or nothing!

Nothing it is, then.

Such is Intel’s fate off late. They have promising consumer grade GPUs and excellent low power N* CPUs but they face an existential crisis now.

The falling margin is a consequence of falling competitiveness. I don’t think you’ll arrest that by decreeing high margin only going forward. But sure give it a go and see I guess.

This likely means they’ll kill their GPUs? Don’t think those are particularly profitable

Sounds about right, it looked like they were a loss leader trying to get a foot into the space. They have been vaguely successful in getting a small toe in the door of GPU's, but not enough to justify the costs.

Profitable != margins. GPU folks can at least make a case that GPUs can be high margin (points at nvidia stock and shrugs). Some businesses where all competitors are selling at low margins that isn’t possible.

GPUs seem like a space ripe for investment, but this kind of management decree might make follow-up impossible. Do what neither AMD or NVIDIA will: sell me a card with oodles of RAM. Could operate at speeds two generations removed from current top tier and it will still sell.

You want this, but the margins aren't there, which is why NVIDIA doesn't care so much about consumer and workstation cards.

How do we know the margins aren't there? Was there a high-RAM low-oomph card that flopped which didn't make the news?

If customer research are doing their job properly, you won’t hear about the flop because they already investigated and determined it wasn’t worth the investment. Before designing the chip.

Like when AMD decided that Nvidia was barking up a crazy tree when they started targeting general purpose compute with their GPUs instead of gamers? Companies miss opportunities for high-margin products all the time. There is every chance there is a huge market for high-RAM GPUs, the demand has only started appearing in the last few years. Margins could be anywhere in the space right now. I don't think its been tested, if there is a high-ram but otherwise mid-range card I'd like to know about it.

nvidia barely cares about gamers, which is a huge market segment compared to programmers, and even less that want video cards with lots of ram.

Not enough profit when they can chase other, more profitable avenues.

You're correct but it doesn't quite refute my statement.

My statement is that good customer research kills bad ideas before they hit the market (so you don't see the failure). However, it can also be the case that it won't spot a good opportunity, therefore there is no product in the market to succeed or fail.

Basically, your statement of there not being a case of a product failing does not prove that market exists. It also doesn't prove that there's no evidence that the market doesn't exist. It just means that there's no publicly available evidence that the market does or does not exist.

I've got an AMD card I bought last year under $1k with 24GB. Meanwhile Nvidia is putting out these weird 8GB and 16GB cards in that price range.

The professional versions should be very profitable. That said, I am hesitant to think that they can sell very many of the professional ones if they don’t sell consumer ones so that software support is there. Maybe they will get the software support in part from iGPUs, but it is possible to design software that only works on iGPUs if discrete versions are not readily available.

100 comments, https://news.ycombinator.com/item?id=44611098

> After years of innovation and community collaboration, we’re ending support for Clear Linux OS. Effective immediately, Intel will no longer provide security patches, updates, or maintenance for Clear Linux OS, and the Clear Linux OS GitHub repository will be archived in read-only mode.. Intel remains deeply invested in the Linux ecosystem, actively supporting and contributing to various open-source projects and Linux distributions to enable and optimize for Intel hardware.

It's a shame but I've never heard of anybody using it to be honest.

> never heard of anybody using it

5 million downloads of the Docker image, https://hub.docker.com/_/clearlinux/

It was a reference distro for performance optimizations, which will hopefully find their way into upstream projects and mainstream distros, https://www.phoronix.com/news/Intel-Ends-Clear-Linux

  Clear Linux has delivered many great out-of-the-box performance optimizations for Linux systems and shown what can be done for packaging with profile guided optimizations / link-time optimizations, various kernel tweaks, and other innovations over the years. At least the likes of CachyOS have picked up some of these optimizations. Intel engineers are also working with other major Linux distributions on enhancing their Linux distribution performance too.

Over the lifetime of CL that’s not very much. Docker pulls isn’t a great metric either because images tend to get pulled repeatedly in CI pipelines.

When considerable portions of the optimizations are things like trading away journaling (in the kernel), I find it's neither effective or recommended.

If those tweaks did anything under a container, they might risk your data for performance. Clear as a container is a niche of a niche.

This article doesn't seem to actually report anything. It is obvious that no company can arbitrarily choose its own profit margins or they wouldn't stop at 50%. So basically all we have here is the Intel CEO says he wants Intel to be profitable and they're going to prioritise products they think are more profitable than others. I hope this isn't news.

The only thing I can spot in the article that wouldn't come up in a psychic's cold read is speculation that there might be a 20% layoff coming hidden towards the end. I'd suggest reading this sort of article is a waste of attention better spent elsewhere.

Companies can choose their gross margins! Just take the cost of manufacturing, double it, and call that your price. You now have 50% gross margins. But you can't choose your total profit. Maybe almost nobody buys at this crazy price you've set. Now your margins don't seem so clever...

They surely can, this is how we usually end up with layoffs affecting people that were doing their jobs as best as they could.

MBA influenced shareholders set random profit margins, when the fiscal year comes to an end and the targets weren't met, a selected random set of employees gets punished due to missing the desired margins.

Yep, I already went through this %@#££! a couple of times, and seen also happen to others.

>Lip-Bu Tan is "laser-focused" on getting Intel back to maximizing shareholder value

I am guessing this is the board's decision again.

Zen 6c is about to come out for server. That is 256 Core per socket. I haven't followed Intel's roadmap for a long time, mostly due to it be very confusing and not exciting. But I am not even aware of any grand plans with TSMC. Which means they are betting their foundry to deliver, both on cost, capacity, and performance.

But AMD needs to work harder. I have been saying this for so many years it is tiring. They need to further increase market shares in every single market, Servers especially hyperscalers, PC, Laptop, OEMS players. I have given them benefits of doubt for so many years Zen 6 generation will be the last of it. I am not even mentioning GPU.

Literally anything a public company does is a board decision. That's how boards work. They say what to do, and if the CEO disagrees, the board votes to dismiss them and put someone in the position who can follow instructions.

One huge factor in Apple’s historic turnaround with SJ’s return to the company: the bailout funders fired the board and brought in SJ’s hand picked board to rubber stamp his decisions.

The Tesla board is a similarly mindless rubber stamp operation, for better or worse.

That's not true, have you never heard of CEOs packing the boards? Also, boards are in charge of a tiny % of all decisions, so it's the opposite of your "literally"

Reminds me of the CEO who said that "from now on we will focus on low-risk projects with huge upside".

Me before a basketball game. My team to score more points than the other team, all new shots must go in.

My google fu is failing me but this sounds a bit like Marissa Mayer’s no more small ideas at Yahoo memo. Also if I was a competitor this would let me know just how hard I would have to compete for Intel to give up.

> Maximizing shareholder value

Depressing. Intel is going to die under this guy's "leadership". I wonder what company will swoop in and buy them? I wonder how much his golden parachute will be.

I felt like Intel was slowly making progress, but the board was too impatient and let Pat Gelsinger go too soon. I don't understand why Lip-Bu Tan would be a better choice than Gelsinger.

You just need to look at Intel’s stock performance under Gelsinger to see why Gelsinger was let go. His loss of a 40% discount on 3nm from TSMC for upsetting them also did major damage to profits.

Stock price alone doesn't tell the full story. Pat Gelsinger was brought in to fix long-term structural issues at Intel - things like rebuilding its foundry competitiveness and securing leading-edge node capabilities. These aren't quick fixes and takes years to pay off. Yes, the TSMC discount issue hurt, but that also speaks to broader industry tensions and strategic missteps that predates Gelsinger. Letting him go before the turnaround had a chance to work seems shortsighted.

The stock lost 2/3 of its value under his tenure. After nearly 4 years and a minor recovery in the stock’s value that was reversed into a new 10+ year low for reasons that were almost entirely the CEO’s fault, I expect most boards of directors would lose confidence in the CEO and fire him.

Pat Gelsinger killed the royal cores project which handed CPU dominance to AMD. Without leadership in CPUs to keep the company going, there would be no Intel. They reportedly lost their entire Oregon design team to a startup developing RISC-V processors as a result of the cancellation, which will have repercussions for many years. Intel’s stock had been on a small recovery until Zen 5 made it clear that Intel could no longer compete with AMD CPUs on performance, which caused the stock to crater. The power management bugs that were killing nearly all of the high end rocket lake CPUs just prior to this also did not help, as they erased some of the gains during the recovery, before it was entirely wiped out by Intel’s lack of competitiveness against Zen 5. There was nothing in Intel’s pipeline to compete with Zen 5, because Gelsinger had cancelled the project that could compete with Zen 5.

The loss of the TSMC discount did not predate Gelsinger. TSMC pulled the discount directly in response to comments made by Gelsinger that they considered disrespectful. This became public knowledge. As I recall, there had been a rumor/leak around summer of 2024 that the board had been very unhappy about that.

While there were issues that predated Gelsinger, he did nothing to fix them. There is a great writeup on that here:

https://bcantrill.dtrace.org/2024/12/08/why-gelsinger-was-wr...

The only thing that predates Gelsinger that Gelsinger cannot be considered responsible for failing to fix is Intel’s reputation as an IP thief from the 80s and 90s. It is likely a major reason nobody wants to use Intel’s foundries.

Apparently, he also volunteered Qualcomm as a customer before they agreed to it and blew a bunch of money buying the high NA EUV machines when TSMC didn’t even have them in the roadmap

> To reach this, Tan has committed to investing in recruiting and retaining top talent; "I believe Intel has lost some of this talent over the years; I want to create a culture of innovation empowerment."

So that means they're going to start paying well and treating people well? Right?

Because compared to FAANG Intel is a big paycut. Particularly in ML the gap is quite large.

But what does that actually mean in terms of strategy? Server and enterprise products have the highest margin. Normal consumers, aside from the small gaming segment, are price sensitive. Are they shifting away from the consumer PC market?

Presumably, they’ll need to abandon cloud and hyperscalers too. That leaves legacy enterprise. (ARM or AMD are going to have better $/perf, and be closer to the dev workstations/laptops, so the only thing Intel would have left is inertia.)

If they actually execute this strategy, sales implode and the CEO is gone.

If they don’t, he demonstrated he can’t set policy, so he’s gone.

Perhaps they're looking at specific empty slots in the competitive product lineup and doubling-down there.

I'm sort of surprised, for example, that LTS embedded is not more of a thing. Sell something like those little N50/100/150 mini-PCs, (or more likely an ITX mainboard with the same specs) but for an astronomical price and guarantees "we'll stock and support it for 10 years" for firms who need certifiable fixed platforms. Small volume but probably easy to get name-your-price margins.

Anecdata, but looking at the higher end of the "normal people" market Intel has already lost. It's macbooks everywhere. The only non-apple notebooks I see are either for cost reasons or tech experts that want more contol (a small group)

And Apple is silently reducing prices by making the cheap Macbook Air close in specs to the Pro. So they're likely to start taking more and more of the midrange market as well.

> The only non-apple notebooks I see are either for cost reasons or tech experts that want more contol (a small group)

You forgot gamers. I know plenty of people with beefy and expensive notebooks that are not Apple.

How big is the "gamer" market really? How much is it set to grow? Are "gamers" really going to be buying the high margin kit from Intel or are they going to instead spend their money with Nvidia on the beefiest GPUs?

Even today a mid-tier CPU is way more than enough to handle even the most demanding games so long as you've got a beefy CPU. Over the past twenty years games have demanded far more in the way of GPUs than they have in CPUs.

That market doesn't make Intel zero dollars or anything but I don't really see it a market that's going to drive a bunch of growth in Intel's revenue or specifically their margins. Long gone are the days where buying the Extreme Edition CPU got some appreciable extra performance in games.

> How big is the "gamer" market really? How much is it set to grow? Are "gamers" really going to be buying the high margin kit from Intel

I was saying that more about your argument about Apple products.

I don't think Intel benefits from the gamer segment, my point is only that there is a significant segment that donown computers and that segment has no interest in Apple notebooks.

In fact, basically everyone I know that owns a laptop has a laptop primarily to play games, not macbooks.

Looks like Intel is going the same way as Boeing - focus on only shareholder value.

You mean mcdonnell douglass. Who went "we both know who's on top here" when they merged, immediately put all of its own people in positions of power, and then effectively destroyed boeing from the inside out in order to maximize shareholder value, until only their name was left and some 350 people were in literal body bags.

I agree that the bean counters from McDonnell Douglas played a big role in Boeing's decline. but it's important to remember that the merger was a negotiated one, not a hostile takeover. Still, it's sad to see how many once-great engineering companies - like DEC, HP, SUN - have been brought down by poor leadership.

Reminds me of that talk on steel mills: https://bhc3.com/2009/05/11/when-being-rational-kills-your-b... You give up the low-end of your business over and over until one day you wake up trying to compete on quality for a tiny share of the market with no revenue left to fund it.

Intel CEO: "Twenty, 30 years ago, we were really the leader. Now I think the world has changed. We are not in the top 10 semiconductor companies."[1]

Ouch. Although Investopedia rates them as #5.

[1] https://www.tomshardware.com/tech-industry/intel-ceo-says-it...

So i guess overnight we have a new floor for unempirical projections of future project profit :D

A lot of made up numbers now have a min(50.0) equivalent somewhere :P

Killing R&D is usually the direct result of such "focus" which leads to dead end anyway.

What a silly idea. Nobody can predict the future (as Intel has demonstrated many times saying no to iPhone for example).

So you need to experiment and try different things to not become a Dinosaur. Not just the obvious things.

They’re just going to offshore a bunch of jobs for a little short term stock sugar rush before the whole thing falls apart

That's gonna be tough when you're paying your competitor to make your products for you.

This is the type of thinking that got Intel to sell their arm CPU product line.

Or else what?

Yes the company that has the least competitive, least desired chips, must either raise its prices or cut costs by reducing support to get higher margins! That’s the real issue!! /S

Throw the financial people out. Get the builders and chip designers back in control. Andy Grove is rolling in his grave. FFS.

Reply on news.ycombinator.com