Interesting. Aren't their CCG and DCAI groups below that margin for now?

For a lot of adjacent or side stuff Intel did in past couple decades this does make sense but in current situation with their main product lines where they are at a disadvantage do they really have capability to aim for 50% and no less? CPUs have been profitable, just not that much, and have subsidised the Foundry that is their loss leader for today.

If Intel has gotten their ducks in line, 18A to work and scale properly then maybe...

This means that upcoming Panther Lake, 18A node HVM, Clearwater Forest Xeons, and Jaguar Shores AI accelerators all carry a gross margin of 50% or more, making them viable for Intel and sustainable in the long run.

Well played, Intel, well played. So what you didn't have for decades, you now suddenly have in every upcoming product line? And the end result is you keep pooping out chips like nobody's business on rather similar game plans?

ROFL. Do they put copium in those water chillers now?

I sincerely hope this is just an empty statement to boost short-term stock prices.

If Intel actually operationally adheres to this, it is the (re-)beginning of the end for them. They fundamentally misunderstand how hardware businesses at that scale work. This will be innovator's dilemma on steroids in the bad way.

Back to the bean counter days except with someone who apparently doesn't know how to count beans.

FiredropsI sincerely hope this is just an empty statement to boost short-term stock prices.

If Intel actually operationally adheres to this, it is the (re-)beginning of the end for them. They fundamentally misunderstand how hardware businesses at that scale work. This will be innovator's dilemma on steroids in the bad way.

Back to the bean counter days except with someone who apparently doesn't know how to count beans.

Remember that NVIDIA has a 70% gross margin. That is with billions of R&D. Companies exist for profit first, pleasure second. This is an entirely healthy way for a company to think in the long term.

FiredropsThey fundamentally misunderstand how hardware businesses at that scale work.

Why don't you explain it to everyone else then?

I don't think it is a bad approach at all. nvidia's GPM for last year was 75%, and AMD was exactly 50% - so I don't see why Intel setting 50% as a baseline is egregious at all. If anything, they need their GPM to be much higher than 50% because they need to constantly spend on foundry as well, until the point (and if) that business becomes self-sustaining.

AleksandarKRemember that NVIDIA has a 70% gross margin. That is with billions of R&D. Companies exist for profit first, pleasure second. This is an entirely healthy way for a company to think in the long term.

R&D expense does not figure into GPM calculations. GPM is only about cost of goods (i.e. variable or per-unit cost, not fixed costs), and sale price.

You guys keep taking this message for what it says to be, but that's not what this message is really saying to us.

The message here is that Intel is excusing itself for a price increase, despite a complete stagnation in product offerings. And they sell that to shareholders/investors by talking about margin estimates that are better than what they had.

In the meantime, they just do what they've always done. There are what, 3-4 product lines upcoming? And they're all going to sell? On what exactly? They haven't delivered jack shit.

This is nothing other than a weak excuse for what we see everywhere; profit go up, with the excuse of vastly more costly nodes and production processes. A function of high demand more than anything else. Intel is just telling us they want a piece of that pie. It has nothing to do with their products at all.

Vayra86You guys keep taking this message for what it says to be, but that's not what this message is really saying to us.

The message here is that Intel is excusing itself for a price increase, despite a complete stagnation in product offerings. And they sell that to shareholders/investors by talking about margin estimates that are better than what they had.

In the meantime, they just do what they've always done. There are what, 3-4 product lines upcoming? And they're all going to sell? On what exactly? They haven't delivered jack shit.

This is nothing other than a weak excuse for what we see everywhere; profit go up, with the excuse of vastly more costly nodes and production processes. A function of high demand more than anything else. Intel is just telling us they want a piece of that pie. It has nothing to do with their products at all.

It's not an external-statement. It's an internal statement saying that if you make a mediocre product and the only way to move it is to sell at 20% GPM (like Arc B580), your product and/or department might be at the end of line, so better start delivering products that can move at a price corresponding 50% GPM.

If Intel still has the talent to deliver such products remains to be seen.

So expect price increase for Celestial - if it even releases in dGPU form factor. No more low prices undercutting Nvidia and AMD.

When Intel is at a disadvantage at every possible angle, this reads as a terrible idea, especially after the painful contractions Intel's had snipping off thousands of employees and booting its old CEO. Intel is not in the position it was in the Skylake days; the only way to dig themselves out of the Alder Lake hole is to either provide an immense value proposition or to make leaps and bounds worth of innovation. Neither offer good margins in any sense of the word.

It totally doesn’t work this way. The market decides what margins you can get. Not the other way around. I mean if a company can just pick its margins why not go for 60%? Oh no wait 70%. Aww hell let’s go 80%.

These business geniuses over at Intel just crack me up.

So they want to enforce Nvidia like margins by simply... Requesting them?

I think this rules out most of their current and future products. Imagine the revenue with no more items that don't bring at least 50% margins! ("But we're not selling anything now?")...

The end of x86 PC might happen like that - with everybody chasing those lucrative server AI deals and abandoning everything else that doesn't bring those margins (neither will AI, once this bubble deflates or bursts).

DavenIt totally doesn’t work this way. The market decides what margins you can get. Not the other way around. I mean if a company can just pick its margins why not go for 60%? Oh no wait 70%. Aww hell let’s go 80%.

These business geniuses over at Intel just crack me up.

The point here it is about not developping products where you think 50% margin is not something achievable in general. So if predictions say you would be able to sell it at 20% margin, it's going to be a no.

Does not mean the end result is 50% margin on everything, after all rosy predictions, lower price later in the lifecycle, it's about setting a goal for new products.

AleksandarKRemember that NVIDIA has a 70% gross margin. That is with billions of R&D. Companies exist for profit first, pleasure second. This is an entirely healthy way for a company to think in the long term.

I genuinely didn't think this had to be stated, but the current Nvidia is an outlier among outliers. Even Apple hasn't been able to demand 50+% gross margins across their hardware portfolio, and consumer behavior is immeasurably different between iphones and CPUs.

uftfaWhy don't you explain it to everyone else then?

How much are you paying me? You will certainly get very far in life by snubbing against knowledge. Maybe you can familiarize yourself with HTC and how they thought they could out-compete iphones simply by charging iphone prices without investing in anything else like Apple has.

uftfanvidia's GPM for last year was 75%, and AMD was exactly 50%

You seem to struggle with basic math. If every single product a company makes has profit margins over 50%, their company-wide GPM will exceed 50%. They are not the "example" that you think they are.

ya, well Nvidia raises intel and goes for 70+ gross margins… you move Intel…lol

qlumThe point here it is about not developping products where you think 50% margin is not something achievable in general. So if predictions say you would be able to sell it at 20% margin, it's going to be a no.

Does not mean the end result is 50% margin on everything, after all rosy predictions, lower price later in the lifecycle, it's about setting a goal for new products.

But that’s the problem, planning products in this way is doomed to fail. Intel should be developing the best products possible to achieve the highest margins possible. This has always been the problem with management at Intel. They run the business according to some formula. This cartoon is still relevant.

You need to be crazy to win and go for broke. Plugging numbers into a margin math formula won’t work.

Come on Intel…GET CRAZY!!!

Or

You

Are

Done.

FiredropsI sincerely hope this is just an empty statement to boost short-term stock prices.

Well, in sum, that's what it is.

Intel can set its margin goals all it wants, but at the end of the day, if you have to undercut your competitor's price for having worse products, that strategy goes out the window.

Vayra86You guys keep taking this message for what it says to be, but that's not what this message is really saying to us.

The message here is that Intel is excusing itself for a price increase, despite a complete stagnation in product offerings. And they sell that to shareholders/investors by talking about margin estimates that are better than what they had.

well they have competition soooooo, I can only see them up prices if they manage to get a product that competes

uftfaI don't think it is a bad approach at all. nvidia's GPM for last year was 75%, and AMD was exactly 50% - so I don't see why Intel setting 50% as a baseline is egregious at all. If anything, they need their GPM to be much higher than 50% because they need to constantly spend on foundry as well, until the point (and if) that business becomes self-sustaining.

By requiring a product have 50% margin, you deny your company opportunities to gain share or explore new markets where gross margins won't be 50% out of the gate. It's the "safe" business choice but staying at the top of the silicon design market requires risky choices to continue pushing improvements. AMD's Zen CPUs for example were an extremely risky proposition for the company, as they poured pretty much everything they had left into a CPU architecture that had a moderate to high chance of not working out. Lisa Sue herself wasn't even entirely sure Zen was good enough when seeing the engineering samples. AMD's gross margins didn't hit 50% until recently so Intel's plan here is pretty much to not take big risks like AMD did on Zen.

It'll also be hard to gain share with that profitability requirement as well. Gaining marketshare frequently requires a company to lower prices.

Nvidia's gross margin is 75% because they have a monopoly over AI market and a near one in the gaming gpu market. It's not a good metric to use to then say Intel's 50% figure is reasonable given their current position in the market. 50% is very good for a company that has innovated and is riding that innovation. That is not Intel right now, Intel is in between S-curves and needs a major innovation to start a new S-Curve. 20 - 30% gross margins should be acceptable to gain share and pretty standard when you see a company wanting to catch up. You cannot demand innovator or market leader margins when you are neither.

I know Intel wasted a bunch of resources pursing adjacent markets like wireless adapters, Nuc, etc but that's due to a lack of leadership. A 50% gross margin requirement is no replacement for a good leader that knows what opportunities to pounce on and what not to. This is just a bean counter approach to what should be a calculated leadership decision. If Intel wants 50% or better margins, it should focus on innovating. Wanting them for doing nothing isn't going to get them anywhere, that's seemingly just a different rationalization for what they've already been doing the last 14 years.

uftfaIt's not an external-statement. It's an internal statement saying that if you make a mediocre product and the only way to move it is to sell at 20% GPM (like Arc B580), your product and/or department might be at the end of line, so better start delivering products that can move at a price corresponding 50% GPM.

If Intel still has the talent to deliver such products remains to be seen.

Intel's GPUs were always going to launch with less than 50% gross margins. It doesn't matter how talented the engineers are, you are talking entering a market where competitors have decades of experience and work baked into their drivers.

Cutting every product that doesn't shortly meet 50 gross margins is inane.

I wouldn't predict 50% margin for any product in Intel's public-facing pipeline right now. Should they all be canceled? Also Intel's GPUs are the key to a bigger and higher-margin market than anything Intel is in today, but they probably have low margins today. Canceling dGPU development today because it's less profitable today only shows a lack of vision.

Kind of like how a few years ago Intel reduced emphasis on new nodes, EUV machines, and shipping products on 10nm because 14nm was so profitable.

uftfaIt's not an external-statement. It's an internal statement saying that if you make a mediocre product and the only way to move it is to sell at 20% GPM (like Arc B580), your product and/or department might be at the end of line, so better start delivering products that can move at a price corresponding 50% GPM.

If Intel still has the talent to deliver such products remains to be seen.

Its a PR piece disguised as an internal statement. Any other company would call this a target and we'd never really hear much of it. Doesn't remotely mean the products adhere to it.

The proof is in the pudding - there has been no real change in Intel's approach whatsoever.

This is the last thing I’d want to hear if I were a long-term investor. Intel needs to hurt for as long as it takes for them to learn a long overdue lesson of what happens when you intentionally stagnate.

I hope they don't kill OneAPI.

The only reason I'm considering a Xeon 6 is that it is supported.

Otherwise a 9950x will do the same job at a quarter of the price.

FiredropsI sincerely hope this is just an empty statement to boost short-term stock prices.

If Intel actually operationally adheres to this, it is the (re-)beginning of the end for them. They fundamentally misunderstand how hardware businesses at that scale work. This will be innovator's dilemma on steroids in the bad way.

Back to the bean counter days except with someone who apparently doesn't know how to count beans.

The same bean counter days of nothing but dual core U CPUs For laptops for what 7 generations??

"intel sets 50% margin"... nGreedia says hold my beer :toast:

DavenThis cartoon is still relevant.

You need to be crazy to win and go for broke. Plugging numbers into a margin math formula won’t work.

Come on Intel…GET CRAZY!!!

Or

You

Are

Done.

That old meme hasn't been relevant for well over 5 years already and it had nothing to do with GPUs.