Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/intel-earnings-announcement-gloomy-2023.17375/page-2
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Intel earnings announcement - gloomy 2023

The problem is that the process lead alone is not enough and sometimes it can be misleading or leading to attention be misplaced. There are many factors and capabilities need to be in place such as yield, IP, ecosystem, supply chain, timing, implementations, execution, cost, price, and capacity, etc.
I agree, and we chemical engineers live for much of that stuff. But the way I see it there are two parts to this.

People who question Pat and intel’s credibility, thinking that all nodes for the rest of time will be 10nm electric boogaloo, any claims that things are going well are lies, i4 won’t begin HVM until late 2024, intel engineers are too incompetent to ever be innovators again, blah blah blah. These people also often ignore TSMCs current roadmap and expect far too much from what is very much TSMC diping their toes into the HNS waters. In my opinion these people are doing TSMC a disservice by overhyping N2. There’s nothing wrong with N2, it is exactly what TSMC needs it to be (a strong low risk foundation to build off of). I don’t want people to start thinking less of those brilliant engineers and scientists at TSMC when N2 turns out exactly like TSMC said it would (a slightly better performing N3 at lower power).

Then there is the second group with the more realistic concerns of are intel’s node on time (excellent) or off by a couple of quarters (not great but still a MASSIVE improvement)? While they will probably have enough capacity for internal products how much will be available for i3/18A external customers? How does intel navigate the financial balancing act of raising the necessary capacity for becoming the number two foundry? How does intel’s wafer cost (with and without IFS’s margin) compare to what intel/other external customers would be paying TSMC and Samsung to make the same chip? What is the value add from intel’s presumed PPW leadership, and how much extra would customers pay for it over say Samsung? I think it is a safe assumption that the IP/design ecosystem for new intel nodes is much better than what it was during the custom foundry days, but how does it compare to Samsung and TSMC? I would assume that somewhere in the middle is a reasonable expectation but who knows this really is a whole new world for intel.
 
Last edited:
I agree, and we chemical engineers live for much of that stuff. But the way I see it there are two parts to this.

People who question Pat and intel’s credibility, thinking that all nodes for the rest of time will be 10nm electric boogaloo, any claims that things are going well are lies, i4 won’t begin HVM until late 2024, intel engineers are too incompetent to ever be innovators again, blah blah blah. These people also often ignore TSMCs current roadmap and expect far too much from what is very much TSMC diping their toes into the HNS waters. In my opinion these people are doing TSMC a disservice by overhyping N2. There’s nothing wrong with N2, it is exactly what TSMC needs it to be (a strong low risk foundation to build off of). I don’t want people to start thinking less of those brilliant engineers and scientists at TSMC when N2 turns out exactly like TSMC said it would (a slightly better performing N3 at lower power).

Then there is the second group with the more realistic concerns of are intel’s node on time (excellent) or off by a couple of quarters (not great but still a MASSIVE improvement)? While they will probably have enough capacity for internal products how much will be available for i3/18A external customers? How does intel navigate the financial balancing act of raising the necessary capacity for becoming the number two foundry? How does intel’s wafer cost (with and without IFS’s margin) compare to what intel/other external customers would be paying TSMC and Samsung to make the same chip? What is the value add from intel’s presumed PPW leadership, and how much extra would customers pay for it over say Samsung? I think it is a safe assumption that the IP/design ecosystem for new intel nodes is much better than what it was during the custom foundry days, but how does it compare to Samsung and TSMC? I would assume that somewhere in the middle is a reasonable expectation but who knows this really is a whole new world for intel.
I agree that people blame Pat for far too much. The sins of previous management set the bed that Intel must now lay in. Pat is shooting for the fences which is what he should do. I am just pessimistic about Intel financial capacity to execute at this point. Technically they seem just fine. Truth me told my curiosity really lies with what TSMC has for N2+ or whatever they call it. We know that’s when BPD is coming and I’m extremely curious as to what is cooking behind the scenes. Speaking of which does anyone know when TSMC’s NA tech symposium is this year? I haven’t heard anything yet.
 
I find myself thinking INTC is a buy at the current levels of pessimism and piling on Pat.

The bull case: Intel can do 18A in Hillsboro. They have the fab sites and some shells to build HVM for this node. 5 nodes in 4 years means in 2027 Intel will be in a process leadership position (4 years from 2023).

Bear case: Too little too late, 4 years is an eternity, and in the meantime cash flow will erode and competitors will respond. There is demand for 18A today, but firm demand? Will demand still be here in 2027? There is a monumental fab re-equipping task, and a monumental yield learning task, that needs to be complete in 2-3 years, to be in production in 4. Competitors are not standing still and don't need to do monumental, only average, to beat Intel to 18A.

To buy Intel's stock at this moment? May the Force be with you!
 
If most people are thinking that, it's often the right time to buy a stock indeed...
Not always. Personally I don’t see why anyone would touch Intel until the turnaround is truly well under way. Unless you are 100% confident it would be better to see things start to pan out first imo. Also I don’t see how the dividend doesn’t get cut at some point. I know management is currently dead set on maintaining it but boy is it a 100 pound weight around the proverbial swimmer treading water. If the dividend gets cut there will be a final flush out.
 
Look at how Intel now performs at the consumer market, and how it performs on the server market.

One thing for sure, Intel has went on incredibly huge, unprecedented price cuts on the notebook CPU market to claw market share from AMD at any cost.

I see less, and less AMD powered laptops in supermarkets with each month.
 
Not always. Personally I don’t see why anyone would touch Intel until the turnaround is truly well under way. Unless you are 100% confident it would be better to see things start to pan out first imo. Also I don’t see how the dividend doesn’t get cut at some point. I know management is currently dead set on maintaining it but boy is it a 100 pound weight around the proverbial swimmer treading water. If the dividend gets cut there will be a final flush out.


I'm not convinced that Intel can afford not to change its dividend practice without damaging its own financial health.


Intel CEO: 'We are committed to the dividend'

 
Because first is literally impossible and Intel surely knows it. TSMC is a pureplay foundry and the best one at it (service and process wise). Part of TSMC's recently growing margin has been their relatively recent process leadership. Just like how TSMC can't magically have the capacity if intel wanted to go fabless, intel can't magically have the capacity to supply 100% of the big 10 fabless firms in addition to their own demands. Also TSMC won't thrash it's margins for years to kill IFS. This isn't total war, it's business. TSMC's margin makes shareholders fat and happy and enables their colossal re-investment. To ruin it would piss off shareholders and potentially sabotage TSMC's future as an innovation powerhouse.

Margin wise let's say TSMC lowers their margin to a "measly" 45 or 50%. This still leaves plenty of room to undercut. If intel has PPW leadership, and your customer service is better than Samsung, then I see no reason why intel can't catch many customers with a 35% margin (some firms might even pay the same price for a small foundry service hit if it lets them have better products than their competitors). The made in the west, and best not TSMC alternative might even allow for higher than 35% margins. Obviously this is nowhere near intel's old 60+% margin but it is probably enough to subsidize their node development like the memory business does for Samsung foundry. IFS also serves as a nice way to monetize the business that intel and AMD's CPU bushiness will eventually lose to systems/software companies. This alternate revenue stream can also prop up the company if the design side goes through a rough patch, and further monetize tools/fabs that are too old to continue using.

At the end of the day design will still be the profit leader but there is plenty of room for foundry to be a big side business that can monetize intel's investments and add a fair amount of extra revenue that can make the company more sustainable.
Good arguments (on both sides actually)

Let me add one point, Qualcomm and Nvidia seem to hold to their belief that they need 2 foundry's, and so are other fabless companies thinking, those folks are afraid of an event where something happens with TSMC and they will be left with no other foundry to their rescue, and they are the target for Intel, Intel must not lower prices that much against TSMC in order to win those alternative seekers, they need PPW to be equal or close to TSMC and provide great costumer service, reliability, etc.

And lets not forgot, if intel is winning on PPW leadership and doing so by low operating cost, it will still be a huge win for the intel X84 CPU business, they just need to go back to the good old days where Intel was making huge profit without any external ISF, after all for intel making just 30% margin on foundry is not a serious problem if they make another 50% margin on the design side of the CPU and it sets Intel in a much better position over AMD which will pay more for TSMC wafers.

And to round up the chances for Intel foundry to compete, they already promise to disclose some wins later this year, it seems they are actually wining atleast some clients.

But yes, those doubts if we can trust intel's claims, and doubting the magnitude or meaningfulness of those promising deals are based on the harsh reality intel is facing, regardless of the Current CEO competency they are currently behind and they deal with TSMC as a competitor while the entire semi industry is now in a downcycle, not easy.
 
Last edited by a moderator:
Good arguments (on both sides actually)

Let me add one point, Qualcomm and Nvidia seem to hold to their belief that they need 2 foundry's, and so are other fabless companies thinking, those folks are afraid of an event where TSMC is going south and they will be left with not other foundry to their rescue, and they are the target for Intel, Intel must not lower prices that much against TSMC in order to win those alternative seekers, they need PPW to be equal or close to TSMC and provide great costumer service, reliability, etc.

And lets not forgot, if intel is winning on PPW leadership and doing so by low operating cost, it will still be a huge win for the intel X84 CPU business, they just need to go back to the good old days where Intel was making huge profit without any external ISF, after all for intel making just 30% margin on foundry is not a serious problem if they make another 50% margin on the design side of the CPU and it sets Intel in a much better position over AMD which will pay more for TSMC wafers.

And to round up the chances for Intel foundry to compete, they already promise to disclose some wins later this year, it seems they are actually wining atleast some clients.

But yes, those doubts if we can trust intel's claims, and doubting the magnitude or meaningfulness of those promising deals are based on the harsh reality intel is facing, regardless of the Current CEO competency they are currently behind and they deal with TSMC as a competitor while the entire semi industry is now in a downcycle, not easy.

1. When Google, Microsoft, Nvidia, Qualcomm, Mediatek, Amazon, and Broadcom talk to Intel for potential manufacturing contacts, they will ask for at least 10% - 30% cheaper price and/or other concessions compare to what they can get from TSMC. One of the reasons is that they are taking huge risks to spend huge amount of money on an unproven upcoming player, Intel Fab Service (IFS). Otherwise they have no immediate and measurable benefits to use IFS as an additional foundry while they already have TSMC (Taiwan, Japan, US, possible Germany) and Samsung (Korea, US) or in some cases, Globalfoundries (US, Singapore, Germany) on hand.

2. Intel's cost is much higher than TSMC and it's very difficult for them to change it in two or three years. If IFS wants to compete, they not only need to be cheaper than TSMC but also need to cut their already higher cost.

If we measure the employee productivity by net profit, TSMC is significantly stronger than Intel.

2021 TSMC average net profit per employee, based on 65,000 employees and US$21.354 billion net profit for 2021:
US$ 328,523

2021 Intel average net profit per employee, based on total 121,000 employees and US$19.87 billion net profit for 2021:
US$ 164,214

Source: https://semiwiki.com/forum/index.ph...’t-fix-the-talent-bottleneck.15998/post-52515
 
Last edited:
Intel CEO: 'We are committed to the dividend'
I saw some people saying that Pat's 'we are committed to the dividend' is nothing more than talk because he didn't say 'we will keep the dividend' and they believe that he will either end the dividend or freeze it at its current level soon. Lots of times, CEOs come out and say this and then have to cut the dividend months later - at least this means that its on the chopping block and is being/has been considered.
 
1. When Google, Microsoft, Nvidia, Qualcomm, Mediatek, Amazon, and Broadcom talk to Intel for potential manufacturing contacts, they will ask for at least 10% - 30% cheaper price or other concessions compare to what they can get from TSMC. One of the reasons is that they are taking huge risks to spend huge amount of money on an unproven upcoming player, Intel Fab Service (IFS). Otherwise they have no immediate and measurable benefits to use Intel as an additional foundry while they have TSMC and Samsung or in some cases, Globalfoundries on hand already.

2. Intel's cost is much higher than TSMC and it's very difficult for them to change it in two or three years. If IFS wants to compete, they not only need to be cheaper than TSMC but also need to cut their already higher cost.

If we measure the employee productivity by net profit, TSMC is significantly stronger than Intel.

2021 TSMC average net profit per employee, based on 65,000 employees and US$21.354 billion net profit for 2021:
US$ 328,523

2021 Intel average net profit per employee, based on total 121,000 employees and US$19.87 billion net profit for 2021:
US$ 164,214

Source: https://semiwiki.com/forum/index.php?threads/semiconductor-investments-won’t-pay-off-if-congress-doesn’t-fix-the-talent-bottleneck.15998/post-52515
Net profit per employee is hardly an apples to apples comparison. I don’t see TSMC helping oems design laptops, creating usb standards, being the largest contributor to the linux kernel, ect. By the same metric intel has a much smaller fab network than TSMC, and probably has a smaller design enablement group. TSMC has their large technician pool to pull up their average (whereas intel’s software people pull down their avg).

There is also capability to consider cloud providers using i3/18a should have better ppw than n3e/n2. This is the single largest thing for cloud providers, with cost being a tiretary concern (if it matered a ton they wouldn’t be spending tens of millions on tapeouts designs and the team to do it). At least for the time being they don’t need heaps of capacity like say Apple Mediatek or AMD do either. They also don’t compete with intel (intel competes with them but that is of no concern to cloud providers who just want the best chip).

As I’ve stated before cost difference while undoubtedly present is overblown. At 14nm(lots of SADP) intel could churn out larger die’d products at higher margins than AMD with better architected chips at 14/12nm(only a bit of SADP) (GF was also probably selling those wafers to AMD at a very low margin). If intel was inherently WAY more expensive, then you would think that AMD’s magin should’ve been much closer to intel than it was regardless of the density difference. At 10nm intel’s architectures have gotten much better, die sizes have gone down, they adopted heavy use of SAQP, and their margin has only gone down. Obviously this is probably mostly due to the loss of business to cloud, AMD, and Apple. However to think that 10nm’s high structural cost does not have a palpable contribution to intel’s margin would be foolish. Ben S at the i4 presentation literally even said that i4 was a big cost per transistor reduction.
 
Last edited:
Net profit per employee is hardly an apples to apples comparison. I don’t see TSMC helping oems design laptops, creating usb standards, being the largest contributor to the linux kernel, ect. By the same metric intel has a much smaller fab network than TSMC, and probably has a smaller design enablement group. TSMC has their large technician pool to pull up their average (whereas intel’s software people pull down their avg).

There is also capability to consider cloud providers using i3/18a should have better ppw than n3e/n2. This is the single largest thing for cloud providers, with cost being a tiretary concern (if it matered a ton they wouldn’t be spending tens of millions on tapeouts designs and the team to do it). At least for the time being they don’t need heaps of capacity like say Apple Mediatek or AMD do either. They also don’t compete with intel (intel competes with them but that is of no concern to cloud providers who just want the best chip).

As I’ve stated before cost difference while undoubtedly present is overblown. At 14nm(lots of SADP) intel could churn out larger die’d products at higher margins than AMD with better architected chips at 14/12nm(only a bit of SADP) (GF was also probably selling those wafers to AMD at a very low margin). If intel was inherently WAY more expensive, then you would think that AMD’s magin should’ve been much closer to intel than it was regardless of the density difference. At 10nm intel’s architectures have gotten much better, die sizes have gone down, they adopted heavy use of SAQP, and their margin has only gone down. Obviously this is probably mostly due to the loss of business to cloud, AMD, and Apple. However to think that 10nm’s high structural cost does not have a palpable contribution to intel’s margin would be foolish. Ben S at the i4 presentation literally even said that i4 was a big cost per transistor reduction.
So much is contingent on 18A in fact being better than whatever TSMC has at that moment. Do we know this to truly be the case for sure? And even if it is. What will the yields be?
 
Last edited:
So much is contingent on 18A in fact being better than whatever TSMC has at the moment. Do we know this to truly be the case for sure? And even if it is. What will the yields be?
Customers and CCG will tell you all you need to know. Samsung has no large volume customers (not even themselves) on 4LPE or 3LPE and barely anyone on "5nm". TSMC did not have any high volume N3 products launch in 2022. Do ARL and whatever the lead product for 18A end up in many laptops across the price range by 2025? How long does it take for there to be 20/18A server chips? Do a large number of customers fab something at 18A? Failing that do a few customers fab a sizeable amount of product on 18A? The answer to these questions will tell you everything you need to know on yield and PPAC.

So far it must look at least okay; otherwise why even consider it? If 18A was looking like 10nm nobody would be taping in, if 18A was too expensive nobody would buy it, and if intel wasn't offering wafer agreements with substantial capacity who would waste their time designing for it no matter how good it was? If nothing else, IFS must be looking better than Samsung on the most important metrics, because if not; who would bother?
 
Net profit per employee is hardly an apples to apples comparison. I don’t see TSMC helping oems design laptops, creating usb standards, being the largest contributor to the linux kernel, ect. By the same metric intel has a much smaller fab network than TSMC, and probably has a smaller design enablement group. TSMC has their large technician pool to pull up their average (whereas intel’s software people pull down their avg).

There is also capability to consider cloud providers using i3/18a should have better ppw than n3e/n2. This is the single largest thing for cloud providers, with cost being a tiretary concern (if it matered a ton they wouldn’t be spending tens of millions on tapeouts designs and the team to do it). At least for the time being they don’t need heaps of capacity like say Apple Mediatek or AMD do either. They also don’t compete with intel (intel competes with them but that is of no concern to cloud providers who just want the best chip).

As I’ve stated before cost difference while undoubtedly present is overblown. At 14nm(lots of SADP) intel could churn out larger die’d products at higher margins than AMD with better architected chips at 14/12nm(only a bit of SADP) (GF was also probably selling those wafers to AMD at a very low margin). If intel was inherently WAY more expensive, then you would think that AMD’s magin should’ve been much closer to intel than it was regardless of the density difference. At 10nm intel’s architectures have gotten much better, die sizes have gone down, they adopted heavy use of SAQP, and their margin has only gone down. Obviously this is probably mostly due to the loss of business to cloud, AMD, and Apple. However to think that 10nm’s high structural cost does not have a palpable contribution to intel’s margin would be foolish. Ben S at the i4 presentation literally even said that i4 was a big cost per transistor reduction.

"Net profit per employee is hardly an apples to apples comparison. I don’t see TSMC helping oems design laptops, creating usb standards, being the largest contributor to the linux kernel, ect. By the same metric intel has a much smaller fab network than TSMC, and probably has a smaller design enablement group. TSMC has their large technician pool to pull up their average (whereas intel’s software people pull down their avg)."

I was talking about Intel's overall cost is too high. I didn't try to explain what caused it but it doesn't change the fact: Intel's cost is too high.

For 2021, TSMC achieved larger "net profit" ($21.354 billion) than Intel ($19.87 billion) while TSMC had less revenue (TSMC $56.84 billion vs Intel $79 billion) and much smaller employee headcount (TSMC 65,000 vs Intel 121,000). I haven't calculated the number for 2022 but Intel's number for net profit per employee should go down further.

This kind of number might not be important in the old days. But now Intel Foundry Service wants to compete directly against TSMC then Intel's cost must be more competitive, for both IFS and Intel product division.
 
This kind of number might not be important in the old days. But now Intel Foundry Service wants to compete directly against TSMC then Intel's cost must be more competitive, for both IFS and Intel product division.
Yes and no. Once IFS starts charging design for wafers then that comparison begins to make sense if you compare IFS's P&L and headcount (not intel as a whole) to TSMC. Even then this isn't totally "fair" since presumably intel design will get sweetheart wafer agreements. This would of course reduce IFS's revenue and margin even if wafer costs were brought close to TSMC.

Regardless of how IFS compares to TSMC, comparing intel in aggregate will never seem logical to me. If anything comparing intel's headcount vs TSMC+AMD+Redhat+IBM makes more sense, because in aggregate intel does a similar amount of work to all of these firms combined (while this is obviously not one for one, the important thing is that manufacturing is only one part of intel). Where it gets interesting is how poorly monetized some of these parts are. I could be wrong but I think it is safe to assume that being a leading member in the bodies that create the various computer standards and the biggest Linux contributor are very low margin activities (and they might even be money sinks). Additionally creating new laptop form factors and helping OEMs design their laptops doesn't directly result in any extra revenue (of course the goal is to promote intel powered designs and branding for these new form factors but you understand my point). These sort of activities also often benefit intel's competitors as much as they benefit intel. Obviously this results in intel having far more employees per dollar earned. But I'm not going to complain since this sort of work benefits almost everybody on earth, and if intel didn't do it there is no guarantee that others would put the same amount of effort in.
 
Regardless of how IFS compares to TSMC, comparing intel in aggregate will never seem logical to me. If anything comparing intel's headcount vs TSMC+AMD+Redhat+IBM makes more sense, because in aggregate intel does a similar amount of work to all of these firms combined (while this is obviously not one for one, the important thing is that manufacturing is only one part of intel).
It's too much of a stretch to include IBM in your list. IBM is mostly a systems company and a cloud computing and services provider. While both Redhat and Intel are big contributors to open source, Redhat gets billions of dollars in software sales and licensing and support revenue. Intel doesn't, not at anywhere near that scale. The TSMC + AMD analogy sort of works though, especially now that AMD acquired Xilinx.
Where it gets interesting is how poorly monetized some of these parts are. I could be wrong but I think it is safe to assume that being a leading member in the bodies that create the various computer standards and the biggest Linux contributor are very low margin activities (and they might even be money sinks). Additionally creating new laptop form factors and helping OEMs design their laptops doesn't directly result in any extra revenue (of course the goal is to promote intel powered designs and branding for these new form factors but you understand my point). These sort of activities also often benefit intel's competitors as much as they benefit intel. Obviously this results in intel having far more employees per dollar earned. But I'm not going to complain since this sort of work benefits almost everybody on earth, and if intel didn't do it there is no guarantee that others would put the same amount of effort in.
Agreed. There's a reason why the Linux Foundation (and Torvalds) are based in The Portland metro area, and that reason is Intel. Also, no other company has contributed more to successful industry hardware and software standards than Intel. It is nice to see the big cloud companies like Google and Microsoft putting more of their technology in the open source or industry specifications domain too.

I also agree with you about Intel's enabling efforts in pushing forward client computing, and even multi-socket servers.
 
Yes and no. Once IFS starts charging design for wafers then that comparison begins to make sense if you compare IFS's P&L and headcount (not intel as a whole) to TSMC. Even then this isn't totally "fair" since presumably intel design will get sweetheart wafer agreements. This would of course reduce IFS's revenue and margin even if wafer costs were brought close to TSMC.

Regardless of how IFS compares to TSMC, comparing intel in aggregate will never seem logical to me. If anything comparing intel's headcount vs TSMC+AMD+Redhat+IBM makes more sense, because in aggregate intel does a similar amount of work to all of these firms combined (while this is obviously not one for one, the important thing is that manufacturing is only one part of intel). Where it gets interesting is how poorly monetized some of these parts are. I could be wrong but I think it is safe to assume that being a leading member in the bodies that create the various computer standards and the biggest Linux contributor are very low margin activities (and they might even be money sinks). Additionally creating new laptop form factors and helping OEMs design their laptops doesn't directly result in any extra revenue (of course the goal is to promote intel powered designs and branding for these new form factors but you understand my point). These sort of activities also often benefit intel's competitors as much as they benefit intel. Obviously this results in intel having far more employees per dollar earned. But I'm not going to complain since this sort of work benefits almost everybody on earth, and if intel didn't do it there is no guarantee that others would put the same amount of effort in.

"Regardless of how IFS compares to TSMC, comparing intel in aggregate will never seem logical to me."

There is only one stock symbol INTC for Intel and only one stock symbol TSM for TSMC. Investors, financial analysts, and shareholders have to make decisions based on Intel's or TSMC's overall cost and performance.

Until one day Intel Foundry Service and Intel Product/Design Division are spilt into two independent companies, Intel's earnings and stock performance will be tightly related to overall what Intel is making and spending. A dedicated PnL number for Intel Foundry Service won't change it.
 
"Regardless of how IFS compares to TSMC, comparing intel in aggregate will never seem logical to me."

There is only one stock symbol INTC for Intel and only one stock symbol TSM for TSMC. Investors, financial analysts, and shareholders have to make decisions based on Intel's or TSMC's overall cost and performance.
Okay do you subscribe to comparing Amazon to AMD then? By your metrics AMD is a far better company; higher revenue per employee and MUCH better margin. Should we close that failure of a semiconductor division that literally generates $0 in revenue every quarter? What about that loser Amazon.com website? Look at how many employees they have for how little profit they generate!

Obviously these comparisons are unfair and unreasonable, because AMD and Amazon have different core businesses. But if you think these are unreasonable statements then how is it reasonable to make these comparisons for Intel and TSMC given how different their core businesses are?

Until one day Intel Foundry Service and Intel Product/Design Division are spilt into two independent companies, Intel's earnings and stock performance will be tightly related to overall what Intel is making and spending. A dedicated PnL number for Intel Foundry Service won't change it.
How would a detailed breakout of IFS's books not tell you most of what you need to know about wafer costs? Let's say in a given Q IFS revenue is 12B (2B being external) and their margin is 18%. Let's also say that in this quarter intel launched a new laptop chip with intel 3 as the SOC die rather than N3E. This information literally tells you almost everything you could possibly want to know about intel's wafer cost. if they are using i3 over N3E (roughly comparable node) then it can be said without a doubt that IFS+internal margin is cheaper then TSMC+margin. If intel does fix many of their cost structure issues you would also see it reflected in CCG/DCAI's margins improving (because I would assume that these savings would be passed on to design rather than pocketed).

I think it is fair to assume that intel design would get charged a lower margin than external customers; for this example let's say that IFS charges intel design 15%. Given that internal is 10/12 of IFS's revenue in this example, that means that external customers would be paying 35% (a margin that would seem healthy since intel is mostly concentrated on the lower margin leading edge). These numbers would then lead to profits from internal customers totaling 1.5B and external customers at 0.7B. The exact numbers for this example aren't really important, but my point with this thought exercise is to show how IFS's customer mix can greatly impact that BU's revenue and margins.
 
It's too much of a stretch to include IBM in your list. IBM is mostly a systems company and a cloud computing and services provider. While both Redhat and Intel are big contributors to open source, Redhat gets billions of dollars in software sales and licensing and support revenue. Intel doesn't, not at anywhere near that scale. The TSMC + AMD analogy sort of works though, especially now that AMD acquired Xilinx.
The redhat bit was more so from an amount of work they perform and headcount. The bit about redhat actually making money from this software was kind of what I was talking about when I mentioned that intel's software efforts are poorly monetized (this of course hurts the $/employee metric). IBM was included to make up for the fact that intel does more collaboration with OEMs than AMD (IBM's mainframe business maybe can act as a stand-in for this headcount wise), and that intel design plays in many markets that AMD simply doesn't/can't (5G towers, quantum, network switches, ASICs, embedded, ect).
 
If Intel successfully launch Intel 3, 20A, and 18A on schedule but with big loss, what will stock market react?
 
Back
Top