Intel Expects to Lose More Market Share, to Reconsider Exiting Other Businesses

[TheoneandonlyMrK]

big.LITTLE is basically good for two generations of 'stretching Core further to keep up'.

Dramatic TDP bumps still required, core count only increases on E cores, they're already pushed way beyond efficient curve... All I see here is a continuation of what Intel was doing since Sandy Bridge. It won't last and Pat knows it.

Exactly this.

But also, I think the chip shortage has exacerbated everyone's issues, a little known thing that's happening is chip brokerage firms, they have exploded in size and scope and are now fully exploitative.

In one case I Know of a roll of chip's that used to cost £300 literally costs a million now.

These aren't modern chip's so have a scarcity tbf but still.

I think the fact stuff is costing what it is has directly driven many to Make do, and affected the bottom line of Intel and others.

Hopefully they can hold out, and pull some stuff on they're fluff because they're PR is starting to wear thin on some line's (Arc)

 

[Wirko]

This is the best post so far. Semiconductors pushing electrons through copper wires and transistor gaps is at the end of its technological life.

Studies for my job in the laser industry has shown a huge increase in nanophotonics research. Its time for the age of the photon!

What's the current state of development of optical transistors (or switches or processing elements or however we call them)? Also, memory cells? I get the most information on new technologies from SemiEngineering, and while they do cover photonics a lot, it's about data transmission mostly, not processing.
There's an inconvenient fact about photon(ic)s mentioned here, I never thought of that before but now it seems obvious:

CogniFiber: Photonic Computing

Israeli startup uses light and glass, but no silicon.

semiengineering.com

On top of that, these devices are not small. “The photon is larger than an electron by several orders of magnitude,” Cohen said. “A transistor is much smaller for electrons than a compartment for a photon. By comparison, the minimal size of a photonic element on a current photonic chip (about 100μm²) is about 30,000 times larger than that of a current electron-bearing transistor made in 5nm technology.”

That means transistors are unable to handle the large sizes of photons, so more transistors are required. And when there are too many transistors, reliability becomes an issue due to interference. These challenges have led companies in the past to abandon photonic computing.

Then there's manufacturing. No matter what great new materials and structures scientists develop, be it in electronics or photonics or quantastics, someone must be able to pack billions of components on a chip, then mass produce that, achieve good yields and uniform characteristics, achieve an acceptable price/performance, guarantee long term stability of the product (so no OLED-based photonics, I guess), be able to simulate and optimise a chip design, and so on.
Fabs as we know them can only do so much. And a bit more in the foreseeable future with high-NA EUV and higher power light sources. But what manufacturing breakthrough is expected after that?

If AMD decides to go ecore as well, Intel won't stand a chance.

If AMD decides to go E-core as well, they'll do it in their own way, radically different from Intel's way. One leak(?) stated that cores may come in 1P+1E pairs - maybe Bulldozer cores done right? If that materialises then it surely will behave much differently than the Lakes, with serious advantages as well as disadvantages.

 

[R0H1T]

The only realistic alternatives are graphene based & they're still decades away!

 

[sepheronx]

Yea. I agree with this 100%. Intel's Decline started when the former CEO Brian Krzanich had to resign because He could not keep his pants.

well, how else are you supposed to find a relationship these days? I initially tried at a grocery store at the check out line but the chicks either roll their eyes at me or think I am some predator.

 

[MarsM4N]

well, how else are you supposed to find a relationship these days? I initially tried at a grocery store at the check out line but the chicks either roll their eyes at me or think I am some predator.

Maybe she just doesn't like pineapple? Try out Tinder, heared it's great for making friends.

 

[sepheronx]

Maybe she just doesn't like pineapple? Try out Tinder, heared it's great for making friends.

im a married pineapple now. And Tinder is where you want to end up with an itchy surprise so no thanks. I met my wife at work and didnt end up like Brian Krzanich.

 

[Lycanwolfen]

The Problem with Intel is they are back to the days when the first Athlon 64's was out and they were stuck with Pentium 4's and crappy CPU's that just cannot compete with AMD again. They need to take a step back and just re design a new CPU from the ground up. None of these Core i3 i5 i7 i9 Bull crap anymore. They need a new name and a whole new look at designing CPU again. Remember the Athlon X2 64 bits at the time was spanking intel left right and center. Took Intel 2 years then the first core 2 duo came out. The E6300 was only running 1.83 ghz but the cpu output was like day a night vs the AMD Athlon's at the time. Intel finally redo everything. They need to do that again.

 

[R0H1T]

Can't do that today, modern CPU's are way too complex for a Conroe type of comeback.

 

[pavle]

The Problem with Intel is they are back to the days when the first Athlon 64's was out and they were stuck with Pentium 4's and crappy CPU's that just cannot compete with AMD again. They need to take a step back and just re design a new CPU from the ground up. None of these Core i3 i5 i7 i9 Bull crap anymore. They need a new name and a whole new look at designing CPU again. Remember the Athlon X2 64 bits at the time was spanking intel left right and center. Took Intel 2 years then the first core 2 duo came out. The E6300 was only running 1.83 ghz but the cpu output was like day a night vs the AMD Athlon's at the time. Intel finally redo everything. They need to do that again.

While I like your idea of a ground up design, the thing is that Core (banal name) was not a new design.
A quick history overview: Pentium 3 (3 pipelines, normal clocks) vs. Pentium 4 (2 pipelines, insane clocks).
Since intel could not wind up the Pentium 4 clocks further (material restrictions), the project failed, so they just took Pentium 3 design added a very good branch prediction unit, developed at their intel Izrael branch and there it was the "Core" line (plus some improvements here and there).
They can't just fall back on any good designs these days because they do not have any, all of the good from Pentium 3 and 4 are mixed in these newer CPUs with improvements and added security holes (you know, for security).
The only thing holding them back is their production process lack of progress, first being stuck for ages at 14nm, now they have somehow patched together that 10nm process that they lie is "7" which it is not and are kind of still stuck at it as with 14nm, so nothing strange they crank out 200W+ wonders and lie and call them 65W And 125W all the while AMD uses TSMC's true 7 and lower nm processes (automatically cooler chips because of smaller transistor size).
Intel needs smaller production process, oddly enough they seem incapable of deploy it in so many facilities that they have. They even produce some of their chips at TSMC, quite a shameful practice for a company that used to lead in process advances. Quite an ironic punishment for their past practices...

 

[Wirko]

Can't do that today, modern CPU's are way too complex for a Conroe type of comeback.

Yeah, all they can do now is send the apes to pick a couple of the remaining coconuts near the top of the palm tree every so often.

 

[Patriot]

I hope optane is opened like IBM power.

Even at the prices it sold at, it lost money. They could just not get its fabrication to the scale and yield rate to be profitable.

 

[Lycanwolfen]

While I like your idea of a ground up design, the thing is that Core (banal name) was not a new design.
A quick history overview: Pentium 3 (3 pipelines, normal clocks) vs. Pentium 4 (2 pipelines, insane clocks).
Since intel could not wind up the Pentium 4 clocks further (material restrictions), the project failed, so they just took Pentium 3 design added a very good branch prediction unit, developed at their intel Izrael branch and there it was the "Core" line (plus some improvements here and there).
They can't just fall back on any good designs these days because they do not have any, all of the good from Pentium 3 and 4 are mixed in these newer CPUs with improvements and added security holes (you know, for security).
The only thing holding them back is their production process lack of progress, first being stuck for ages at 14nm, now they have somehow patched together that 10nm process that they lie is "7" which it is not and are kind of still stuck at it as with 14nm, so nothing strange they crank out 200W+ wonders and lie and call them 65W And 125W all the while AMD uses TSMC's true 7 and lower nm processes (automatically cooler chips because of smaller transistor size).
Intel needs smaller production process, oddly enough they seem incapable of deploy it in so many facilities that they have. They even produce some of their chips at TSMC, quite a shameful practice for a company that used to lead in process advances. Quite an ironic punishment for their past practices...

Ya well like a said before they need to redesign everything from the ground up. Stop using old tech and mixing it here and there. Like If I was head of Intel I would look right away at increasing the IPC instructions per cycle, not little steps I would look at huge steps like maybe 4 to 8 to 16 IPC. vs the min 2 and 2.5 we have today. Increasing the IPC like 10 or 20 fold would allow the chips to perform faster and lower speeds. Next cores and P cores and E cores. Now its cool and all but most performance of chips is cores in general. If I worked at Intel I would come out with a 16 or 32 core desktop model for gamers or high end computer users. Not limit it to Xeons only. I mean AMD has threadreaper which yes its expensive as hell but still it can be purchased for a home user. As for process power Its too bad the cores or the multicores could not act as one. Say you have a demanding app or game and the CPU had a feature where all the cores could act as one or some of the cores could be assigned as one. Imagine that tech would be cool. Say you have an 8 core CPU and you wanted to run a game you need at least 1 core to run the operations for the background stuff and the other 7 cores to act as one CPU. Could you imagine the performance, Say 7 cpus running at 4 ghz giving you a 28 Ghz processor all running together. Intel needs a big leap again.

 

[ratirt]

If AMD decides to go E-core as well, they'll do it in their own way, radically different from Intel's way. One leak(?) stated that cores may come in 1P+1E pairs - maybe Bulldozer cores done right? If that materialises then it surely will behave much differently than the Lakes, with serious advantages as well as disadvantages.

I get what you are saying but my point is, Intel has already hit a wall with the pcores and it is rather hard for them to move forward without ecores. AMD has the Zen4 and this one has a significant improvement over previous gen. If AMD went ecores (their own idea of ecores) Intel would not stand a chance in performance nor power. The chiplet design is the best move AMD made so far. Where it will lead we will see. I guess Intel will have to make a chiplet design as well.

 

[Wirko]

I get what you are saying but my point is, Intel has already hit a wall with the pcores and it is rather hard for them to move forward without ecores.

So they will move forward with E-cores. 8+24, why not?
The advantage of P-cores is much better performance when running one thread per core. That's great but why would you need more than so many of them?
The advantage of E-cores is better performance per mm2 of silicon, which is reflected in better performance of another Intel product, known as INTC.
I assume, based on available data for desktop ADL (which is scarce), that two threads on an E-core two E-cores execute as fast or a little faster (on average) than two threads on one P-core.
Issues with P+E scheduling remain but it's a matter of software, which can be improved and tuned over time. Hopefully there will be a couple tuning knobs left to the end user to turn, too.
Heat density be damned, we'll sell you a larger cooler.
Edit: because it was unclear before

 

[TheoneandonlyMrK]

So they will move forward with E-cores. 8+24, why not?
The advantage of P-cores is much better performance when running one thread per core. That's great but why would you need more than so many of them?
The advantage of E-cores is better performance per mm2 of silicon, which is reflected in better performance of another Intel product, known as INTC.
I assume, based on available data for desktop ADL (which is scarce), that two threads on an E-core execute as fast or a little faster (on average) than two threads on one P-core.
Issues with P+E scheduling remain but it's a matter of software, which can be improved and tuned over time. Hopefully there will be a couple tuning knobs left to the end user to turn, too.
Heat density be damned, we'll sell you a larger cooler.

You mean two threads on two E core's beat two threads on one p core.
SMT does only usually amount to +30% the performance of one core.

 

[ratirt]

So they will move forward with E-cores. 8+24, why not?

I think they do not have any other choice.

I assume, based on available data for desktop ADL (which is scarce), that two threads on an E-core execute as fast or a little faster (on average) than two threads on one P-core.

You got it wrong buddy. Ecores don't have SMT. I disagree with the rest to be fair. 2ecores are not equal to 1pcore nor faster. If that would have been the case we would not have Pcores whatsoever and yet here they are. Ecores are growing because it is cheaper for Intel and they take less space but saying these are better performers than pcores is clearly wrong.

Issues with P+E scheduling remain but it's a matter of software, which can be improved and tuned over time. Hopefully there will be a couple tuning knobs left to the end user to turn, too.

Intel has a problem with forcing things their way thinking developers will make it happen for them. I'm skeptical in that department especially because I don't think ecores are to advance but instead make good marketing and cut costs.

 

[stimpy88]

After the 6+GHz announcement, I wonder if Intel has just gaslighted AMD...

 

[oxrufiioxo]

After the 6+GHz announcement, I wonder if Intel has just gaslighted AMD...

Amd isn't going to lose sleep over a 3-400W processor on mainstream desktop..... Don't get me wrong I want to see it benchmarked but it's going to be a silly product if it exist.

 

[Wirko]

You got it wrong buddy. Ecores don't have SMT. I disagree with the rest to be fair. 2ecores are not equal to 1pcore nor faster. If that would have been the case we would not have Pcores whatsoever and yet here they are. Ecores are growing because it is cheaper for Intel and they take less space but saying these are better performers than pcores is clearly wrong.

Of course, no MT on E. I corrected my post above. But I didn't say or imply that E-cores, any number of them, could make P-cores redundant. They couldn't.

Here's what I mean, in terms of relative performance.
1 thread on 1 P-core = 1 (low-threaded CPU load, so SMT is not necessary)
1 thread on 1 E-core = 0.65
2 threads on 1 P-core = 1.3 (SMT used because high-threaded CPU load) (probably a bit less on average)
2 threads on 2 E-cores = 2 x 0.65 = 1.3 (probably a bit less because of shared L2 cache)

On a CPU with 8P+8E cores, the scheduler should put 8 most demanding threads on the P-cores, one on each; next 8 threads on E-cores; and next 8 on P-cores using SMT. (I do realise this is an oversimplification, threads are dependent on other threads, power consumption matters too, etc.)

I don't think ecores are to advance but instead make good marketing and cut costs.

These things are not mutually exclusive. Is it good Intel giving CPU buyers more performance per mm2 of silicon, or bad intel giving INTC buyers more performance because of less silicon used? Or both?