Intel Posts Record Results, But Misses Revenue Targets and Issues Weak Outlook

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.


Invalid's argument is that there is no social information showing that AMD has Ryzen and that's why Intel is over priced. Now my question is this: when was the last time you saw an AMD advertisement? That's not Intel's fault.

Case closed.
 

That's a useful clarification, but it needs to be seen in the context of the business reality that Intel cannot just arbitrarily decide to reduce their margins, which is essentially what that would amount to. As noted in the report, their volumes are down (due to making larger dies on a fixed wafer supply), which is why they're actually even more protective of the margins of what products they can ship.

Since not everyone here is familiar with how business works, let's try a little thought experiment. Let's say that Intel decided it wanted to be nicer to its customers and reduce its margins. The resulting drop in profits would directly hit their share price, since investors primarily look at a company's P/E ratio. When investors lose money due to some reckless move or failure to mitigate a foreseeable issue, they will pressure the board to replace the executive team and/or sue the company (which places even more pressure on the board). As for those executives, they'll have difficulty finding their next job. As for their savings, much of their compensation is in the form of stock options, which are now probably underwater, due to the drop in share price.

So, Intel has to do everything possible to protect its gross margins. That's why they're charging as much as they think the market will bear, and even beyond that, have biased wafer capacity in the direction of higher-margin products (as noted in the article).

This is why I said they can't just give away the extra die space, even if it would still be profitable to do so. If they had surplus 14 nm capacity, then perhaps they'd make up the difference in increased volume. However, their much-publicized delays is ramping up 10 nm + competitive pressure from AMD and others has pushed them to ship larger dies on their 14 nm node than they previously planned, leading to the capacity shortage.

It's worthwhile to compare Intel with its peers, but that doesn't mean Intel can suddenly become more like AMD. In fact, to the extent possible, we should expect to see AMD trying to become more like Intel.
 

Legal definitions of monopolies are usually based on market share. And even with all of Ryzen's success, Intel still massively dominates x86 market share, in most/all segments. Legally, I'm sure they still classify as a monopoly.

Interestingly, this does tie Intel's hands, in certain ways.
 

I'm curious to know on what evidence this assessment is based. Do you mean generally, or specifically for CPUs?

IMO, if Intel maintains adequate supply of a given product line, then prices will settle out with standard markups for retailers. Where you tend to see abuses in the channel is when supply is tight.


ORLY? Intel pays dividends, which is a share of those profits you give them. According to this, they recently increased their dividends, and here's an opinion it's still not enough:

https://www.fool.com/investing/2019/01/25/intels-dividend-boost-isnt-enough.aspx

I'm not saying I'd rather line the pockets of retailers, but it's also a mistake to think Intel reinvests 100% of their profits into the business. Sadly, it's not the case.

I found a handy list of Intel's dividends: https://www.nasdaq.com/symbol/intc/dividend-history

No such history for AMD, suggesting (but I'm not 100% sure) that they've never paid any:

https://www.nasdaq.com/symbol/amd/dividend-history
 


This is specific to computer components. CPU's, video cards, memory, etc.. Your last part is correct. But there is zero indication that Intel is intentionally under supplying the chain. This isn't a situation where Intel can just flip a switch and start producing significantly more CPU's. Doesn't matter what AMD can produce, because right now, the majority of the market doesn't want their CPU's. So we are going to be looking at the same market constraints for the forseeable future.



ORLY? Intel pays dividends, which is a share of those profits you give them. According to this, they recently increased their dividends, and here's an opinion it's still not enough:

https://www.fool.com/investing/2019/01/25/intels-dividend-boost-isnt-enough.aspx

I'm not saying I'd rather line the pockets of retailers, but it's also a mistake to think Intel reinvests 100% of their profits into the business. Sadly, it's not the case.

I found a handy list of Intel's dividends: https://www.nasdaq.com/symbol/intc/dividend-history

No such history for AMD, suggesting (but I'm not 100% sure) that they've never paid any:

https://www.nasdaq.com/symbol/amd/dividend-history

Not sure why you would assume I thought 100% of the money Intel brings in they put into R&D. They spend about 20% of their revenue on R&D. Any additional money they bring in would likely have the same allocation to R&D.
 

Intel has announced investments on the order of a couple $B to increase 14 nm production. It also moved some things, like chipsets, back to 22 nm fabs.

From the article:
Intel says its 10nm processors are on-track for a volume retail launch in 2019, and also that it expects the shortage to last through the end of the second quarter.
That should be calendar Q2.


Well, some are not aware of dividends. I, myself, didn't even know whether Intel paid them until I checked for that reply.

The other thing we learned is that if you want more of your money to go into R&D, probably AMD is a better choice. However, simply spending on tech (particularly from publicly-traded companies) does send a signal to the market and encourage more investment, overall.
 
wow what some people would say to defend their favorite cpu maker
some people tend to forget that the only reason why intel abandoned the traditional 4c/8t mainstream i7 and moved to 6core and more was because AMD forced them when they released the Ryzen cpus which cost much less and have more cores than competing intel cpus.
comparing the new i7 and i9 with Haswell i7 released more than 4 yrs ago is also a weak attempt to justify their price increase. The prices may have gone down lately but for quite a while they were a lot higher than their msrp. Maybe some people tend to want to live under a rock.
https://www.tomshardware.com/news/14nm-processor-intel-shortage-9000-series,37746.html
https://www.techpowerup.com/249559/cpu-shortages-will-continue-into-the-second-quarter-of-2019-according-to-asustek-ceo

Using the same manufacturing process and architecture for such a long time, Skylake was released back in 2015, defending the manufacturers inability to come up with something new in such a long time while keep making money on an old architecture and then even run into production problems which lead to shortages and price increases really doesn't make sense at all...

News flash: Intel has bigger gross margins than AMD! zOMG, where have you been? That is not news.

so true.... which is the reason for my criticism
of course they are making tons of money when they keep using the same old Skylake for years
and I am pretty sure I am one of the more loyal intel customers in here having been using their cpus since when I was just a kid, from 1gen Pentium 120Mhz to Pentium4 to their new multi core cpus
 

First, if you're accusing others of being partisan, please be specific.

I think you're so blinded by your own bias that you can't accept any explanations of Intel's actions as anything but Intel partisanship. I think that's sad. Also, unproductive.


I didn't see anyone disputing that. I don't know of any evidence, but I think most people take it as given. In fact, it's what I was referencing, here:
In fact, due their problems with 10 nm, and competitive pressures forcing them to ship more large-die CPUs (not only in the desktop and laptop markets, but also their HEDT


Since you've clearly been following the story of Ryzen, you'll know that its cores are actually smaller than Skylake's, in terms of both transistor count and die area. Coupled with AMD's multi-die strategy, this was a move that Intel could only counter, in the short term, by replicating their existing cores and shipping bigger monolithic dies.

If you combine bigger dies and a strong economy with unprecedented delays bringing up their 10 nm node, it's a recipe for a supply crunch, which has all the consequences that have been mentioned: scarcity-driven price gouging by retailers and an increased need to counter volume reductions with bigger margins. Sorry if you don't like it, but those outcomes are pretty much self-evident.

If you think about it in strategic terms, it looks like a masterstroke by AMD. I mean, they can't have known Intel would've had so much trouble with 10 nm, but this strategy really couldn't have worked out much better, for AMD.


You're wrong on several points. The first uArch to use their 14 nm node was actually Broadwell. Secondly, they've iterated on their 14 nm manufacturing tech to produce 14+ nm and 14++ nm. I honestly don't know much about the differences or whether they upgraded existing production facilities using the earlier versions to produce the later ones.

However, there's a bigger point you're missing, which is no less significant in the recent story of Intel than Ryzen: 10 nm. It was supposed to launch years ago. When CPU designers create a new design, they do so with a particular manufacturing technology in mind. Intel has had uArch designs ready and waiting for 10 nm, for a while. The production problems meant those designs have largely sat on the shelf, forcing Intel to hastily churn out Coffee Lake and Coffee Lake R, as stop-gap measures. Far from saving them money, having to take such measures probably cost them more.

It's not that they couldn't or didn't design new CPUs - it's that those designs got stalled by their 10 nm manufacturing problems. They highlighted this as one of the areas they plan to improve upon, during their most recent Architecture Day, as covered here:
Intel tells us that the company will now design new microarchitectures to be portable between nodes. That will allow the company to move forward even if it encounters roadblocks on its path to smaller transistors.
Source: https://www.tomshardware.com/reviews/intel-sunny-cove-gen11-xe-gpu-foveros,5932-4.html


Except they haven't actually saved money from this, since they did the work on new designs which they haven't been able to use.

Don't believe me? Here's a detailed review of the Cannon Lake core they shipped on 10 nm, way back in 2017:

https://www.anandtech.com/show/13405/intel-10nm-cannon-lake-and-core-i3-8121u-deep-dive-review

The reality serves only to highlight some of the problems with their 10 nm node, but its existence proves they had these designs ready and waiting to go.


I don't care whether you're a jilted Intel fan or a jealous AMD fan. If you want to learn anything or have a productive discussion, you'll need to set aside the idea that everything is necessarily biased towards one side or the other. It's possible to understand why Intel is acting in the way that you've criticized, without endorsing it.

Of course, if you're just here to lash out and vent your frustration, then I think we have little more to say.
 


Hit me up when AMD runs a Super Bowl commercial. 😀
 
so my original comment has made some of the fanboys really upset
"Overall, the company posted yet another full-year revenue record of $70.8 billion"

wow this means people are still buying their cpus even though they keep refreshing their Skylake architecture and increase prices
and they have to thank the "experts" who keep recommending gamers to buy i7s and i9s to play CS and Battlefield

my argument still stands
you still don't need 450-500 dollar cpus to play triple A titles in 2019
Intel is still using Skylake architecture from 2015

Well played Bit, well played. Logic, chickensquack has none.
the name calling shows how mature the fanboys really are
this is hilarious
congratulations for have found your soulmate
 

We do agree on these points.

I have just one question: how long do you think it takes Intel to design a new architecture (e.g. Sandybridge, Haswell, Skylake, etc.)? I mean from the time they start working on it in earnest, to the time they first start selling the chips to the public.
 
chickenballs voted down for this answer
I take that to mean you haven't a clue.

So, you're criticizing Intel for using the same architecture since 2015. You (now) know their 10 nm node is something like 3 years late, and you don't even know how long it would take them to design another architecture for 14 nm, as a better stop gap than what they achieved with Coffee Lake / Coffee Lake R and the improvements to their 14 nm node.

All the while, down-voting all my posts, because I called you out.

You know, the reason I participate on forums like this is to share knowledge and information. It goes both ways. Sometimes, I try to make an argument, only to have it fall apart, as others point out a hole in my facts. Guess what? I still win, because I learn something, when that happens. If we ignore all information that doesn't align with our existing opinions, we don't learn anything!

So, go ahead and refute and down-vote all you want, but I think you're ultimately wasting your time if you're going to remain so stubbornly wed to your opinions. I'm not even saying you need to agree with me, but there are more mature ways to handle an irreconcilable disagreement.
 
10tacle voted down for this answer 2 February 2019 16:14:05
bit_user voted down for this answer 31 January 2019 05:24:09

oh i am sorry for have stooped to your level by voting down your answers
I didn't realize there are intel evangelists in this forum
but then again I am not surprised considering how often people here suggest that gamers need expensive 6core+ intel cpus to play modern games
and my criticism of their pricing and their inability to release a new architecture/node is partially based on several relatively recent articles
if you and your supporters dont like my criticism then tell sites like Toms, Techpowerup and others to stop releasing negative news about Intel
 

I actually prefer intelligent discourse to downvotes. But after seeing you indiscriminantly downvoting my posts with no apparent reason, I figured I'd play along.


Your refusal to see objective facts and argumentation ultimately hurts only you.


It's one thing to say you're not happy with what they're doing, but it's another to suggest they should be doing something different. The first is voicing a legitimate opinion, while the second is ignorant of business and technical realities. I'm not saying these companies always make the right decisions, but there are usually some strong forces and sound logic behind their actions.

Speaking of articles, did you check my links? Why not address my points on their merits, if you think they're invalid?

I think the biggest problem is that you haven't read enough of the CPU coverage on this site. Most of what I've said is taken straight from Tom's and Anandtech's coverage. You'd do well to read the links I posted, assuming you care about anything more than Red vs. Blue taunts and digs, like children in a playground.