Intel CEO Expands on Desire to Make Apple Silicon Chips, Touts 'Competitive Fun' With Apple

[Andropov]

Is it normal for sinking ships to increase YoY revenue and generate billions in profits and free cash flow?

Intel Reports Fourth-Quarter and Full-Year 2020 Financial Results

www.intc.com

• Full-year revenue set an all-time Intel record of $77.9 billion, up 8 percent YoY.
• In 2020, Intel generated a record $35.4 billion cash from operations and $21.1 billion of free cash flow (FCF) and returned $19.8 billion to shareholders.

Nokia's all-time earnings record year was actually one year after the original iPhone was released.

 

[pshufd]

Nokia's all-time earnings record year was actually one year after the original iPhone was released.

Pandemic is a special case. More servers needed and shortages of AMD parts which are more desirable.

 

[Rigby]

It makes no sense to say a process node has “limited frequency scaling.” That’s not a thing.

Gate size, leakage and quantum effects at today's transistor sizes absolutely limit frequency scaling. Some of these issues can be mitigated by clever design.

 

[Andropov]

Not only does APPLE miss out on killer Intel CPU's coming down the Pipeline but also Killer New Memory DDR5!
Twice as fast.

Samsung Unveils Next-Gen Memory for Data-Hungry AI and Computers

(Bloomberg) -- Samsung Electronics Co. will release a new generation of memory chips in late 2021, its first in seven years, that promises to double speeds and offer the biggest capacity yet to keep pace with the growth of data centers and artificial intelligence demands.The world’s largest...

www.yahoo.com

Intel has been notably slow adopting some memory technologies for its CPUs. So much that Apple had to switch back from LPDDR4 RAM to DDR4 RAM to support 32GB on MacBook Pros, since Intel's chips were always one-year-away from supporting more than 16GB of LPDDR4 RAM (which is still true to this day).

 

[sdwaltz]

I'm not saying I believe lizard people are real, but hypothetically, if I thought they were real, this guy definitely looks like he'd be the king of the lizard people.

 

[Arsenikdote]

If you’re American and you’re cheering against Intel I think you need to think more about this situation, and how eventually, that will impact you.
You want your country to be competitive in the global landscape. It is sad that Apple (an American company) has to look to a foreign country for its chip making needs, when there’s Intel still around just down the road.

it’s bad for the American industry ultimately if Intel were to die.

Huh. Well, I am not a fan of how Intel has gone from a company who other chip companies try and model after to a corporation who can't seem to find their way. It has created an unnecessary stagnation. I would LOVE for a disruption to focus the industry and spur on a whole new generation of chip designs. Is that feasible? Is there room left with our current understanding of materials and physics? I don't know enough to say what head room we have left, but I WISH for a future in which we see these advancements. And what Intel has done with the company in the past decade has not shown them to be the industrious leader they once were. This is their issue to own and has created the negative sentiment, not the other way around.

Your point about considering what that means for the US is a thought experiment since again public sentiment does not dictate the technical and business ability of Intel. Intel may or may not fail. A new company may emerge as the new leader and that company may or may not be based in the US. It's not like global tech advancement is sitting around waiting to bend to the whims of some nation, even the US. Technology will continue to advance and public sentiment about Intel doesn't make those companies magically exist or not exist in the US. Your point as a thought experiment might be interesting, but you saying it like a warning, as if people's actions have consqeuences seems like a massive overreach of cause and effect!

Is it normal for sinking ships to increase YoY revenue and generate billions in profits and free cash flow?

Intel Reports Fourth-Quarter and Full-Year 2020 Financial Results

www.intc.com

• Full-year revenue set an all-time Intel record of $77.9 billion, up 8 percent YoY.
• In 2020, Intel generated a record $35.4 billion cash from operations and $21.1 billion of free cash flow (FCF) and returned $19.8 billion to shareholders.

I guess defining "sinking ship" is important here. How fast is it sinking? Is it salvageable? Etc...

For someone like Intel YoY revenue growth is almost a necessity. I mean just remaining nearly stagnate would produce a YoY revenue increase just with inflation. So that metric is honestly a poor metric if you want a single metric to judge a company on. The other one you mentioned is also not a great one because with the size of Intel, Billions in profits and free cash flow also are not necessarily good things. Billions in profits are a bare minimum for their size. Also, free cash flow might show a company who doesn't have good leadership or strong direction and so can't invest their money. Not saying I think Intel is a sinking ship, just saying that your case for why they are not is not very credible.

Intel is failling behind and others are catching up. This is known and this is what is causing the reaction. Are they doomed? If they continue to operate for the next 10 years like they have for the previous 10 years, I believe they will become irrelevant. I don't think they can withstand another decade of operations in this similar path. I still don't think they are going to "sink", they will likely just get acquired by someone else, or pivot and get niche. But they won't be "Intel" for all purposes unless the drastically alter their trajectory. And the threat is not just from Apple...

 

[pshufd]

Huh. Well, I am not a fan of how Intel has gone from a company who other chip companies try and model after to a corporation who can't seem to find their way. It has created an unnecessary stagnation. I would LOVE for a disruption to focus the industry and spur on a whole new generation of chip designs. Is that feasible? Is there room left with our current understanding of materials and physics? I don't know enough to say what head room we have left, but I WISH for a future in which we see these advancements. And what Intel has done with the company in the past decade has not shown them to be the industrious leader they once were. This is their issue to own and has created the negative sentiment, not the other way around.

Your point about considering what that means for the US is a thought experiment since again public sentiment does not dictate the technical and business ability of Intel. Intel may or may not fail. A new company may emerge as the new leader and that company may or may not be based in the US. It's not like global tech advancement is sitting around waiting to bend to the whims of some nation, even the US. Technology will continue to advance and public sentiment about Intel doesn't make those companies magically exist or not exist in the US. Your point as a thought experiment might be interesting, but you saying it like a warning, as if people's actions have consqeuences seems like a massive overreach of cause and effect!



I guess defining "sinking ship" is important here. How fast is it sinking? Is it salvageable? Etc...

For someone like Intel YoY revenue growth is almost a necessity. I mean just remaining nearly stagnate would produce a YoY revenue increase just with inflation. So that metric is honestly a poor metric if you want a single metric to judge a company on. The other one you mentioned is also not a great one because with the size of Intel, Billions in profits and free cash flow also are not necessarily good things. Billions in profits are a bare minimum for their size. Also, free cash flow might show a company who doesn't have good leadership or strong direction and so can't invest their money. Not saying I think Intel is a sinking ship, just saying that your case for why they are not is not very credible.

Intel is failling behind and others are catching up. This is known and this is what is causing the reaction. Are they doomed? If they continue to operate for the next 10 years like they have for the previous 10 years, I believe they will become irrelevant. I don't think they can withstand another decade of operations in this similar path. I still don't think they are going to "sink", they will likely just get acquired by someone else, or pivot and get niche. But they won't be "Intel" for all purposes unless the drastically alter their trajectory. And the threat is not just from Apple...

Long-term stock charts are a great quick and dirty of what the market thinks of companies.

Lots of big cap tech companies are above their 2000 highs. Not Intel. They may get there but that's a long time to wait for all-time highs.

 

[Arsenikdote]

Long-term stock charts are a great quick and dirty of what the market thinks of companies.

View attachment 1749342



View attachment 1749343

Lots of big cap tech companies are above their 2000 highs. Not Intel. They may get there but that's a long time to wait for all-time highs.

Heh, I would be interested to see Nokia....as others have pointed out they were doing quite well the year before the iPhone was released

 

[pshufd]

Heh, I would be interested to see Nokia....as others have pointed out they were doing quite well the year before the iPhone was released

 

[cmaier]

Gate size, leakage and quantum effects at today's transistor sizes absolutely limit frequency scaling. Some of these issues can be mitigated by clever design.

What? No. These are orthogonal issues.

A smaller gate size has less capacitance and therefore charges and discharges (e.g. switches) *faster*.
Quantum effects prevent the ability to fully turn off the transistor by applying a gate voltage, but that doesn’t affect switching speed - it just means you have static current leakage.
And leakage doesn’t affect F-sub-t either. It just means you have leakage. So the speed doesn’t change, but your total power dissipation does.

And since you are talking about scaling the frequency of the chip, and not a single transistor of arbitrary size, none of this matters anyway. If I have a transistor with a gate size of 15nm on a 5nm process or on a 14nm process, it’s the same transistor. It will switch at the same speed (assuming doping, etc. is equal).

And of course “frequency” of a chip is a function of the number of transistors and amount of wire in the most timing critical path, and is not really affected by the f-sub-t of the transistor, because the f-sub-t of the transistor is much much much much higher than the frequency of the chip, anyway.

I have no idea what you are trying to get at, but as a cpu designer who worked on many of the fastest chips out there at the time, i can tell you that’s not how it works.

 

[MBAir2010]

i would type something here about intel being good, great track record and owning the market,
but feel the need to somehows educates meselfs!
so i wonts!

 

[quarkysg]

So much that Apple had to switch back from LPDDR4 RAM to DDR4 RAM to support 32GB on MacBook Pros, since Intel's chips were always one-year-away from supporting more than 16GB of LPDDR4 RAM (which is still true to this day).

I wouldn’t be surprised if Apple already have DDR5 memory controller baked into their SoCs now. That’s what I think will go into the Mac Pros, and probably higher end iMacs when the Apple Silicon versions are released, and likely with a high number of memory channels. Let’s hope the supply is enough for Apple to proceed with DDR5.

 

[pshufd]

I wouldn’t be surprised if Apple already have DDR5 memory controller baked into their SoCs now. That’s what I think will go into the Mac Pros, and probably higher end iMacs when the Apple Silicon versions are released, and likely with a high number of memory channels. Let’s hope the supply is enough for Apple to proceed with DDR5.

I am expecting the high-end iMacs and Mac Pros to offer DIMM slots for user-supplied RAM as second tier RAM. I will be somewhat disappointed if they don't.

 

[quarkysg]

I am expecting the high-end iMacs and Mac Pros to offer DIMM slots for user-supplied RAM as second tier RAM. I will be somewhat disappointed if they don't.

With the the increased bandwidth provided by DDR5, I think it'll just be DIMM slots for system RAM. I don't think Apple will go with soldered RAMs for the Mac Pros.

As for the iMacs, it uses SO-DIMM modules since it's inception (IIRC). So the iMacs may go all soldered RAMs. Having said that, with the removal of dGPUs from iMacs, it may make room for DIMM slots to be include in the main board. We can always hope.

 

[pshufd]

With the the increased bandwidth provided by DDR5, I think it'll just be DIMM slots for system RAM. I don't think Apple will go with soldered RAMs for the Mac Pros.

As for the iMacs, it uses SO-DIMM modules since it's inception (IIRC). So the iMacs may go all soldered RAMs. Having said that, with the removal of dGPUs from iMacs, it may make room for DIMM slots to be include in the main board. We can always hope.

The 27 inch models have always had user-serviceable RAM but that might not work with a much thinner iMac. They had user-serviceable RAM in the 2008 MacBook Pro and they could do something like that. DDR4 or DDR5 wouldn't matter to me. I just want to be able to add RAM down the road if needed.

 

[quarkysg]

DDR4 or DDR5 wouldn't matter to me. I just want to be able to add RAM down the road if needed.

Yeah, I very much prefer user replaceable RAM too. I added more to my 2010 27" iMac and it's still running perfectly fine hosting Linux VMs to allow me to hack around with my router's firmware.

The problem with DDR4 is the low bandwidth. It'll starve the M1 of needed bandwidth which will result in bad performance. AFAIK LPDDR4X RAM (currently used for Ax and M1s SoCs) does not come with any DIMM/SO-DIMM modules, so we won't find any LPDDR4X RAM modules for sale.

Not sure if DDR5 will have SO-DIMMs intended for notebook use.

 

[pshufd]

Yeah, I very much prefer user replaceable RAM too. I added more to my 2010 27" iMac and it's still running perfectly fine hosting Linux VMs to allow me to hack around with my router's firmware.

The problem with DDR4 is the low bandwidth. It'll starve the M1 of needed bandwidth which will result in bad performance. AFAIK LPDDR4X RAM (currently used for Ax and M1s SoCs) does not come with any DIMM/SO-DIMM modules, so we won't find any LPDDR4X RAM modules for sale.

Not sure if DDR5 will have SO-DIMMs intended for notebook use.

If they allow user-serviceable RAM, I would expect it to be two-tier with latency costing built into the operating system for making decisions on where to place things.

 

[e.clinton]

If you’re American and you’re cheering against Intel I think you need to think more about this situation, and how eventually, that will impact you.
You want your country to be competitive in the global landscape. It is sad that Apple (an American company) has to look to a foreign country for its chip making needs, when there’s Intel still around just down the road.

it’s bad for the American industry ultimately if Intel were to die.

People cheer against Intel because Intel has been performing terribly. If Intel were to disappear today, no one would miss them... crazy thing to say when 10-15 years ago they were god of tech.

 

[Michael Scrip]

Exactly.

Nokia was still riding high in 2007-2008 when the iPhone was available in what, only 4 countries?

But it didn't take long for Nokia to disappear from the smartphone conversation.

And it wasn't just the iPhone that signed Nokia's death warrant... it was also Android.

Nokia got hit by two avalanches.

See also: Blackberry

 

[wanha]

No.
It’s about a company falling behind plain and simple.
It’s about complacency, We see it happen to major iconic companies all the time.
It happens to complete industries. American car companies in the 80’s. Japanese cars filled a void where all three American car companies has become lazy and compacent and ate our lunch.

What you say is all true.

What is also true is that the weird mind games Intel is playing is very reminiscent of what's going on generally in America at the moment.

There's a lack of clarity all around.

 

[Falhófnir]

If you’re American and you’re cheering against Intel I think you need to think more about this situation, and how eventually, that will impact you.
You want your country to be competitive in the global landscape. It is sad that Apple (an American company) has to look to a foreign country for its chip making needs, when there’s Intel still around just down the road.

it’s bad for the American industry ultimately if Intel were to die.

So what, the US taxpayer needs to write a blank cheque to keep Intel going no matter how uncompetitive it becomes or no matter how much of a loss it might crash to? That’s been tried many times and it’s never sustainable and never has the desired effect. If Intel collapses there’s the opportunity for fragments of the company to spring back as innovative or competitive entities. Propping it up keeps those suppressed by the dead weight. If Intel has a future it’s one it needs to make itself, standing on its own two feet. They’re down but not out at the moment, if they act boldly (for starters embracing Arm rather than trying to fight against it, this is perhaps a first tentative step in that direction?) they might yet turn things around anyway.

 

[cmaier]

So what, the US taxpayer needs to write a blank cheque to keep Intel going no matter how uncompetitive it becomes or no matter how much of a loss it might crash to? That’s been tried many times and it’s never sustainable and never has the desired effect. If Intel collapses there’s the opportunity for fragments of the company to spring back as innovative or competitive entities. Propping it up keeps those suppressed by the dead weight. If Intel has a future it’s one it needs to make itself, standing on its own two feet. They’re down but not out at the moment, if they act boldly (for starters embracing Arm rather than trying to fight against it, this is perhaps a first tentative step in that direction?) they might yet turn things around anyway.

Given how you spelled check, why are you worried about the us taxpayer?

And given how Intel’s foreign competitors are often subsidized by foreign governments, why should us taxpayers do the same to avoid the loss of jobs, technological prowess, and supply chain disruptions?

 

[ArPe]

Our problem is not Intel anymore it’s the chronic chip shortage caused by inefficiencies.

Crypto miners especially those stupid single purpose ASIC chips have gobbled up the silicon supply and put strain on the fabs. As usual, crypto always causes socialized losses to consumers and the climate so that a small minority of crooks can take all the money with their manipulated market.

Their whole inefficient system of millions of chips doing the exact same work is based on greed and no utility except crime and we all have to pay the price in more expensive processors or delayed processors.

Then comes along that 💩 Elon and he promotes it because the chip supply strains hurts other auto car makers. He knows crypto is stupid but he is using it as a proxy to damage the auto competition.

 

[Falhófnir]

Given how you spelled check, why are you worried about the us taxpayer?

And given how Intel’s foreign competitors are often subsidized by foreign governments, why should us taxpayers do the same to avoid the loss of jobs, technological prowess, and supply chain disruptions?

You seem to have missed my point, it's not about the taxpayer (though I don't like to see ordinary people being taken advantage of to line corporate pockets) it's that if you want to be commercially relevant, the last thing you want to be doing is using public money to subsidise failure. Intel has no lack of funds currently, privately, if it's able to get out of this mess it can do it alone, if it can't then declaring Intel is the US’ national chip champion and that it can’t be allowed to fail no matter what ain't gonna help.

 

[cmaier]

You seem to have missed my point, it's not about the taxpayer (though I don't like to see ordinary people being taken advantage of to line corporate pockets) it's that if you want to be commercially relevant, the last thing you want to be doing is using public money to subsidise failure. Intel has no lack of funds currently, privately, if it's able to get out of this mess it can do it alone, if it can't then declaring Intel is the US’ national chip champion and that it can’t be allowed to fail no matter what ain't gonna help.

The u.s. Subsidized the auto industry here a few years ago, and saved several companies from going under. That seems to have worked out great.