Agreed! And, after Nov 4, it still won’t beat the Max in processing 8k ProRes video.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Intel CEO Hopes to Win Back Apple by Making Better Chips Than Apple
- Thread starter MacRumors
- Start date
- Sort by reaction score
Agreed! And, after Nov 4, it still won’t beat the Max in processing 8k ProRes video.
How do you know that? It scores 18000 muti in geekbench 5
Does the m1 support pcie gen 5 and ddr 5 memory?
Because the M1 has hardware acceleration on the chip for the ProRes codec. As ProRes is not important to the vast majority of Intel’s customers, you can bet that they haven’t spent the money to license the codec from Apple to include in their current chips.
Oh, and the M1’s it will continue to be the fastest chips that ship in systems that natively support macOS (since Apple won’t be using any of Intel’s).
Intel made their initial bad decision LONG ago when they declined to provide Apple a processor for the iPod. I feel that they COULD have, but they had been teaching the PC industry for years that “small doesn’t = powerful” by gimping their lower wattage chips (removing structures that COULD have been there just to provide a tiered structure).
If they had made a small, high performance x86 compatible processor for Apple, folks would have been clamoring for that type of innovation from the i3 line of processors, for example. And if they had made something that was good but not x86, then doubts would have been cast on x86. So, in SOME ways, I can see why they turned it down.
However, if they knew then what we know today… that an Apple making their OWN processors is eventually going to cause those headaches for them anyway, they may have been more ready to bite the bullet and deal with the fallout then. Who knows, that very early focus on performance per watt might have driven them to be able to avoid the whole Skylake situation
If they had made a small, high performance x86 compatible processor for Apple, folks would have been clamoring for that type of innovation from the i3 line of processors, for example. And if they had made something that was good but not x86, then doubts would have been cast on x86. So, in SOME ways, I can see why they turned it down.
However, if they knew then what we know today… that an Apple making their OWN processors is eventually going to cause those headaches for them anyway, they may have been more ready to bite the bullet and deal with the fallout then. Who knows, that very early focus on performance per watt might have driven them to be able to avoid the whole Skylake situation
Intel is delusional...they have failed to keep up with chip designs that work for their (previously) large customer. Now their all gung-ho? 3 years and a trillion dollars short...
Intel has become like IBM (though IBM is far worse). More focus on shareholder return; quarterly numbers, appeasing Wall St. analysts; not caring about the top talent, being happy with the market share; a top-heavy organization. IBM is dead, hopefully, Intel gets better.
Most of corporate America is focused on performing to shareholder and investors than to anyone else.
Look at Boeing! They are NOT, apparently an engineering company. Likely news to all of their engineers and all those that are retired. They are apparently a company that pulls all the strings to generate as much profit as they can, using as much unskilled, non-union, and foreign labor as possible, and trying to get government to go along for the ride. (They 'captured' the FAA and were in charge, horrifically, for inspecting their own planes, while the amount of 'FOD" that is found in their finished planes has skyrocketed. Tools, rags, food, fasteners, helmets, goggles, boots, notes from employees, etc)
I feel that the current devotion to the investor class can't last, and this also transcends politics and industries. Wall Street is enraptured by the investor class. It's infected the entire country.
But Intel thinking after flipping off Apple, and other manufacturers, that they can 'win them back'? With what? The only way Intel would likely be able to woo Apple back would be to bid on a contract to manufacture their newer chips, and likely take a massive loss to do it. Intel's modem chips are still a loser, even after Apple spending money to fix them. It's sad. But Intel makes most (all) of their chips in China, Costa Rico, Malaysia, etc, so what do 'they' actually make? Not much... Which makes it more sad that Intel is a shell of a company that help invent the microprocessor, and they are falling al over themselves to save themselves, and also dance to investors. It can't last...
Would Apple be confident that access to the extreme details of the designs and manufacture of Apple chips won't boost Intel's ability to deliver competing devices?
TSMC appears to have a more appropriate level of non-competition with customers than would Intel acting as a contract manufacturer.
Without a doubt, hell no... But look at China. I'm sure that everything they build for US Investors is helping them build their own tech and make it incredibly impossible to fight.
Who thought that giving our manufacturing ability to a communist country was a great idea? It was InSaNe!!!
Absolutely not confident at all. I mean think about it this way.
Intel had to open the x86 architecture spec that could not require a license, which is how Cyrix, AMD, and one other manufacturer was able to jump into the competition. Intel lost their monopoly on the PC market because of that, as their CPUs were too pricey for everyone, and the others got into the market by producing the same performance and selling at a lower price point.
Conversely, AMD did similarly the same with the x86_64 spec, and Intel jumped in on that. They created faster CPUs but kept the higher selling point. AMD lost more on the speed and the cores during the Core Duo/Core-xx days, but because of Intel's mindset of keeping shareholders happy over making better products, they lost out on that, and are scrounging for what they can get to bring them back into the game. Apple would be fools to open up Silicon to them. If anything, Apple would have learned from Intel and AMD from their tit-for-tat over the last 25 years.
BL.
For the same, or lower, power input?
Aside issues like battery life, heat production/cooling, the simple fact is that energy costs have absolutely rocketed.
I have no problem with the idea of an x86 chip matching an AS chip in flat out performance. They might be able to adopt features of AS (such as the interconnects, unified memory, etc.) to level up in some areas.
Simplistically, you can switch x86 features between being implemented in pure hardware and other techniques. But it seems to me that the x86 architecture cannot ever be made as power efficient as AS (or other RISC and RISC-like approaches). It will have to do more work. Even if the hardware were pure RISC, it would be having to do more work.
(I'm no chip designer or expert. This is just how I see it in my little corner of ignorance.)
You’re correct. At the core of it, supporting the x86 instruction set, including all the cruft that’s still in there from years ago, requires a good deal of work during the decoding step. For EVERY instruction. Intel would have to drop compatibility with a lot of legacy code in order to take a shot at simplifying decoding (they’ve tried it before), but the market Intel’s in rewards backwards compatibility over raw performance, so that’s not an option.
Additionally, Rosetta appears to run Apple Mac Intel code faster than Intel not because ARM can run any random set of Intel code faster, it’s because Apple, for awhile, has continually dropped older methods from being produced by their compiler. At this point, every Apple Silicon chip knows exactly what subset of Intel methods they’re ever likely to have to execute and all those methods have been tuned for performance. Any non-macOS ARM implementation would have considerably more work to do trying to handle the vast x86 history (including 32-bit instructions) that Apple’s cut out, meaning that’s not a shortcut that’s available for Intel either.
Remember the fuss when Windows dropped 16-bit support?
(I know there have been various virtual ways of getting some 16-bit code to run.)
Could it? Can "the end" propagate as fast as the speed of light?
With a propagation time for light of about 69 milliseconds from one side of the earth to the other, and a blink being around 100 milliseconds, I suspect it could lead to a whole series of arguments and counter-arguments.
That was from a physics class where the prof said that *everything* could cease to exist at the blink of an eye because there is no finite reason for why everything continues to exist, moment to moment. Given the idea that quantum mechanics can apparently prove that we live in such a crazy world, it then *is* apparently possible that everything really could just not exist, in an instant. That's the same branch of physics that says it is *possible* to actually walk through walls You just have to make all the atoms in your body line up with the empty spaces in the corresponding wall's atoms. Simple I know, right?. That's the problem as it has never actually happened (so far as we know. (Not many people are found stuck halfway through walls)). So is it *possible* that the entirety of life, the universe, and everything could *POOF* disappear? Apparently yes. Is it likely, sadly perhaps, no. But in my earlier post, that didn't stop someone from having 'a moment', a massive panic attack needing therapy and definitely a change in educational direction.
But I also love thinking of an 'endless' universe but it freaks the hell out of many people. *shrug*
Last edited:
British politics has already been scrambling my brain for at least days - without adding that!
Definitely unstable.
But apple have only just started …
The m1 knocked it out of the park , the future looks very bright 💡
No wonder intel are worried 😂
The m1 knocked it out of the park , the future looks very bright 💡
No wonder intel are worried 😂
Hopefully. But Intel’s CEO went on a whole spiel about how he’s going to invest in new fabs because it’s the right thing for Intel, and he doesn’t need government handouts, but now he’s refusing to build them unless Congress gives him $50 billion (after Intel spent the last decade doing billions in buybacks instead of ASML machines). So the plan here seems to be “too big to fail”.
They're not handouts, they're "investment in the future" and "creating jobs."
Seriously, corporate welfare is even more sacrosanct than welfare; politicians love to say "I created those jobs here" even when the reality is, if it's profitable, that company would build teh plant without the handout. Company's like to get locals to bid against each other to get the best deal.
I did some work with an economic development agency, and one of the most important things was to work with the politicos to be sure to schedule the announcement of the money so they could be there with a huge check and photo op.
Of course. Intel's problem is it's become too big to innovate. I suspect they've been the top dog so long that people have become afraid to try something with high potential but may fail before it finally works; and failure may be a career ender so they simply no longer stick their neck out.
It's hard to change a culture that has become risk adverse because they fear the impact on their career of being seen as having failed. You also get a lot of "we tried that and it didn't work" types that have a million reasons why you should not do something new or different.
Another problem is the engineers themselves - they often disdain management and pointy haired MNAs, and do not aspire to high management positions or deign to learn about marketting, finance, sales, etc. and so the ultimate decision makers care about the bottom line and have no real passion for innovation, all the slogans aside.
Ah! Just like IBM. (IBM is the worst) IBM borrowed money to pay dividends and do buybacks. Let us not talk to dozens of dubious acquisitions and spin-offs
The stock buyback swindle works on two levels, one to give more money to the management teams and investors, but two it is used to show that everything's fine because stock prices go up when companies do that. They can 'prove' that the economy is doing gang busters becasue stock prices go up, which is a lie. They steal money from the company to do a reach around and put money into well fed management who are largely immune from the effects of reality, and float away on golden parachutes when everything craters.
Which is exactly why up until Reagan stock buybacks were illegal and considered price manipulation.