Intel Has To Be Better Than 'Lifestyle Company' Apple at Making CPUs, Says New CEO (theverge.com) 215
Intel's new CEO, Pat Gelsinger, doesn't start his new role until February, but he's already prepping the company to take on Apple's M1 chips. From a report: The Oregonian, a local newspaper in Oregon where Intel maintains a large presence, reports that the chip maker held an all-hands company meeting yesterday, and Gelsinger attended. "We have to deliver better products to the PC ecosystem than any possible thing that a lifestyle company in Cupertino" makes, Gelsinger reportedly told Intel employees. "We have to be that good, in the future." Intel has been facing increased competition from both Apple and AMD recently. Apple announced its transition to its own silicon back in June, calling it a "historic day for the Mac." The transition has gone well, with M1-based Macs providing impressive performance and battery life compared to existing Intel-based Macs.
Too late to catch up (Score:5, Insightful)
The situation at Intel was so bad that not only is AMD surpassing them, but other brands aren't even bothering to switch to AMD. There's plenty of room in between for Apple to jump in.
Intel made bad choices more than a decade ago by stagnating and raking in profits and now they can't catch up.
Re:Too late to catch up (Score:5, Insightful)
He's focused on the right things, though. Improving quality, making better chips.
If they had hired a CEO that wanted to cut the workforce 10% to improve profitability, you would know Intel is dead.
Re: (Score:3)
Say and Do, completely different they can be!
Re: (Score:3)
I hope they stay relatively competitive, because I don't want a single company making all the CPUs. For the same reason I don't want AMD to go out of business.
Re: (Score:3)
The future belongs to neither Intel nor AMD, but ARM.
Apple's gamble on the M1 is paying off. It is faster and way more power efficient, giving hours more battery life. Rosetta runs legacy apps with decent performance, often better than they run natively on x86.
The same will soon happen in the Windows world. Microsoft is improving Windows-on-ARM, and has a x86-to-ARM translator [windows.com] similar to Rosseta.
x86 is dead.
Re: (Score:3)
Apple's gamble on the M1 is paying off. It is faster and way more power efficient, giving hours more battery life.
That's great for that particular market segment, but Apple doesn't make the fastest CPU around, and you can still buy cheaper systems with more CPU power. Even a $300 laptop these days has 8 hours of battery life if you're word processing or performing other light tasks, which is provably good enough for most purposes. So it remains to be seen whether Apple's gamble on ARM is going to ultimately pay off, or whether they're just going to alienate more power users by dropping x86 emulation in two years or wha
Re: (Score:2)
I agree, much has to do with "what" the company is focused on, is it getting every dime out of aging infrastructure (I once worked at a semi co that was sitting on tons of 4" silicon wafers hoping a market would develop for them) while returning revenues to investors and ignoring the future, OR has it been investing in a long chain of research and development in order to remain competitive...
imo, Intel beat the snot out of AMD by shortcutting security [theregister.com] and took a real performance hit when they had to drop so
Re: Too late to catch up (Score:2)
We know Intel is dead anyway. Their performance advantage was one part process technology and one part security holes (speed holes confirmed!) Without it they have only inertia and that won't save them because 1) the competition is compatible or can emulate and 2) instruction set matters less now anyway. It doesn't really matter what they do, short of lobbying the feds to bomb the opposition.
Re: (Score:2)
I feel like Intel has a lot of smart people, they ought to be able to take advantage of that some way, at least in theory.
Re: (Score:2)
Re: (Score:3)
Re: (Score:2)
Ooh, looks like I got a SIGSEGV in the threaded section. This will be fun
Delete that shit.
I wouldn't go that far (Score:3)
Re: (Score:2)
there just aren't that many folk in the world who can design CPUs.
That's not really true, though.
Re: (Score:2)
Re: (Score:3)
I imagine most of them have designed and built a CPU. Certainly I did it in a 200 level EE class in college.
When you are building a modern super-scalar CPU, of course it is too hard for a single person to do everything. And you will probably hire a person to build your clock circuitry who has done nothing his entire career but design clock circuitry. But if they hire someone new, the biggest problem will be for that person to get used to all of Intel's internal tools, not for them to understand the fundamen
Re: (Score:2)
Certainly I did it in a 200 level EE class in college.
::Snork:: thanks for the laugh. I studied CPU design in college as well. Did you design a new 32 core multi-threaded gigahertz-level CPU with the thousands of bells and whistles of a modern processor, or did you design the equivalent of a rudimentary 4004 like I did? You and I aren't the guys they are talking about, assuming you're not a rock-star in the CPU world either.
probably hire a person to build your clock circuitry who has done nothing his entire career but design clock circuitry.
That's what rsilverGun is talking about. You not only want the person whose worked their entire career designing clock circuitry, but
Re: (Score:2)
But if they hire someone new, the biggest problem will be for that person to get used to all of Intel's internal tools, not for them to understand the fundamentals of CPU design.
That's kind the point we're arguing.
It's really not. Learning Intel's internal tools isn't going to be as hard as learning fundamentals of CPU design.
Re: (Score:2)
Learning Intel's internal tools isn't going to be as hard as learning fundamentals of CPU design.
Ok. Now I don't even know what we're arguing about, lol. We're in agreement, on this part, I think. I'd much rather hire someone that knows the subject matter and teach them my tools and policies.
there just aren't that many folk in the world who can design CPUs.
That's not really true, though.
Maybe that's the disconnect. OP didn't specify the quality of the people that are available. We've established that both you, I, and most other EE grads are likely perfectly capable of doing CPU design work. I know the fundamentals of baseball, even play some beer league softball, but the Twins aren't looking
Re: (Score:3)
CPU designers are a dime a dozen compared to TSMC EUV engineers, which is what they lack.
I suspect TSMC strongarmed ASML to let them put their own virtual pellicle technology into the machines, that's what they need to replicate.
Re: (Score:3)
Although the 486 was a decent processor for its time it's not really the same as modern processors with multiple cores, hyperthreading (fake cores) and a lot of other features intended to speed up things.
But given that an 8 year computer (Core i7-2600 with about 117kMIPS) still works fine today means that the pace of development has slowed down considerably. The speed between an 80286 (mid life) was about 1.28 MIPS while the Pentium some 8 to 10 years later had 188 MIPS. A speed increase of about 140 times.
Re: (Score:2)
It all depends on spending (Score:2)
Too late to catch up
Face it, it is all about spending. While Apple may have more money they also have more areas they need to be spread it over. If Intel can significantly outspend Apple on CPU design they may catch up.
Look at the PowerPC. It too looked better than x86. Intel's ultimate win in that contest was not a failure on the PowerPC side. It was really that Intel pulled off absolute miracles in getting performance out of x86 that no one had imagined possible. Basically Intel outspent Apple/IBM/Motorola and pulled ahea
Re: (Score:3)
Look at the PowerPC. It too looked better than x86. Intel's ultimate win in that contest was not a failure on the PowerPC side. It was really that Intel pulled off absolute miracles in getting performance out of x86 that no one had imagined possible. Basically Intel outspent Apple/IBM/Motorola and pulled ahead.
Well, that and the partners pulled out. Motorola went first, I believe, and IBM liked POWER but wasn't really interested in PowerPC (the "PC" being the operative part). That left Apple as pretty much the only advocate for PowerPC (if you don't count the Mac clone makers), so development of that line stagnated at the same time that Intel was investing.
Re: (Score:3)
Look at the PowerPC. It too looked better than x86. Intel's ultimate win in that contest was not a failure on the PowerPC side. It was really that Intel pulled off absolute miracles in getting performance out of x86 that no one had imagined possible.
AMD also beat PowerPC, for pennies on Intel's dollar.
The "miracles" Intel pulled off were partly due to superior process technology, which is now over. (It's not due to inferior security because POWER processors were some of the few chips which were also fully vulnerable to MELTDOWN like Intel's.)
Re: (Score:3)
If Intel can significantly outspend Apple on CPU design they may catch up.
Intel reported $29.6B on R&D and capital investments [intc.com] in their last 10-K on 1/24/2020. Apple reported $18.8B [apple.com] on R&D. While both companies do not break down exactly what they spent their money, I would guess the Intel spends way more on CPUs R&D than Apple as Apple has other product lines.
The problem for Intel the last 5 years is not in their CPU design. The problem for Intel has been their manufacturing failures. First 10nm and now 7nm have been plagued with low yields. Apple relies on TSMC for m
'Lifestyle Company` (Score:5, Insightful)
Intel thinking of Apple as a `Lifestyle Company` is one of the reasons that they are in this situation.
Re: (Score:2, Funny)
Basically, the only reason anyone ever buys any Apple product is to look cool in front of their friends. All their products are stupid and their customers are stupid because if they were smart they'd buy whatever it is I buy because it is better.
Or some minor variation on that.
Comment removed (Score:4, Interesting)
Re:'Lifestyle Company` (Score:5, Informative)
Then there are entire companies built on using OSX because it's unix and we still have strong enterprise tool to manage them.
I and almost every developer I know uses OSX because it's unix. Good linux laptops really don't exist, I've bought a few and I've always been disappointed. Dell was a back breaker (15inch), cheap netbooks are too slow, other companies charge as much or more than apple for less just to get linux.
In the end OSX gives me a proper terminal, a pretty good UI, tons of commercial applications, and my IT department doesn't tell us 'there are no good MDM tools, no good EDR tools, etc). It ties well into Okta, AD, etc.
Apple is more for developers than designers at this point.
Re: (Score:2)
In the end OSX gives me a proper terminal
Please. OS X terminal sucks. Every time I have to use it, it feels like I'm using a Linux from the 90's. Missing common options to various base utilities, poor completion (did they enable bash completition yet?), no package manager out of the box. At least it provides an ssh client so that you can connect to a real, useful terminal and do what you have to do.
Re: 'Lifestyle Company` (Score:2)
OSX is certified UNIX. That makes it UNIX. Or at least it was, I'm not keeping up.
Re: (Score:2)
More mistakes in one post - or s ingle line - than ever.
NextStep is Unix.
BSD is Unix.
MacOS is Unix.
They use it because Apple forces you to use it when developing iOS.
You can develop on any machine you want, you just can not push it to the App Store, facepalm, idiot.
Re: (Score:2)
NextStep is not Unix. It has its own kernel which is separate from Unix. Its called the mach kernel. Mac OSX also uses it. Just because something is POSIX compliant and has some open source utilities included does not automatically make it Unix. Windows also has some Unix derived utilities and code in it. Is it Unix? Of course not.
Re: (Score:2)
Normally I wouldn't care, but my major issue is that the entire industry - phone and computer - seems to think "Well, if Apple can get away with it, we can too!",
Unfortunately for many of these folks, it's a cargo cult mentality.
Re: (Score:2)
But nobody outside of these groups is buying Apple equipment to get functionality that's significantly better than its rivals. They're buying for the image.
Is that an American disease?
I live half in Europe and half in Asia.
Never met anyone buying/using an Apple Computer for any "image reason" ... people who might buy it for "image reason" are probably to stupid to use it anyway.
So: is that an American myth? Are you really convinced of that bullshit attitude?
Re: (Score:2)
Is that an American disease?
I live half in Europe and half in Asia.
Never met anyone buying/using an Apple Computer for any "image reason" ... people who might buy it for "image reason" are probably to stupid to use it anyway.
You could have said you are not part of that group but to say it doesn't exist (or would be somehow limited by geography to a single continent) likely means you are part of it and lack self-criticism.
I even read one slashdot user who was proud to say he liked the fact that his kids were using iDevices to communicate, even thought it excluded some other kids from their group of friends (cause they were using some Apple-only communication protocol from what I understand), because that would make his kid less
Re: (Score:2)
But nobody outside of these groups is buying Apple equipment to get functionality that's significantly better than its rivals. They're buying for the image.
You are exactly the sort of person I was referring to.
It is an objective fact that Apple products work differently than their rivals. Even in some checklist style comparison (both have lighted keyboards, both have a screen, etc etc) the way Apple implements their functionality is probably different.
What is completely subjective is whether that way is better or worse than another option. You may not think it is better or you may not care but that doesn't mean the only possible answer is people are chasin
Re: (Score:2)
I was immediately and severely butthurt for a brief period of time. ;)
Re:'Lifestyle Company` (Score:4, Insightful)
Intel thinking of Apple as a `Lifestyle Company` is one of the reasons that they are in this situation.
What's wrong with what he said? Isn't Apple a lifestyle company? The iPhone is definitely a fashion accessory. You need a useless hole in the case just so others can see that apple logo.
Re: (Score:2)
You need a useless hole in the case just so others can see that apple logo.
LOL I'm glad someone else recognizes this. I always look for cases without that hole, and sometimes they're hard to find.
Re: (Score:2)
Re: (Score:2)
They don't need it to be a requirement. Most people buying Apple products will avoid cases hiding the Apple logo. Because they care about how their phone looks, and how they will be perceived by others if the logo is proudly visible.
Re: Just ignore the vast hardware work from Apple (Score:2)
The differentiating factor for a lifestyle brand is not how much work they do. It's the reason people give them money. If it is for technical superiority and the buyers are educated well enough to know the difference, it's not a lifestyle brand. Most of Apple's customers are buying them for style reasons, though. They are not well-educated, which is why Apple is a lifestyle brand.
They do occasionally have superior technology, though they have never managed that for very long. Someone has always handed them
Re:'Lifestyle Company` (Score:5, Insightful)
I have seen a lot of good companies die, because they had failed to respect what the "Consumer Grade" competition is.
Big iron mainframe Companies (sometime division within a company) (Prime Computer, VAX, IBM, Wang...) Who were making million dollar mainframes, Failed to see the threat of workstation and mini-computer market like from (SUN Microsystems, SGI, DEC Alpha...) Who failed to see the Threat from standard Desktop PC Market with (Intel, AMD)...
With the older company pressed on the threat of the newer company, their response is something like "We are the big boys in the industry, those guys are fine if you just want to play with little toys, but if you want do do REAL WORK you need to work with us"
Mainframe Guys pointed out how much processing bandwidth they had, allowing a lot of users to run simple apps on their system.
Mini-Computer guys pointed out how well it balances performances for high end calculations to allowed for fast request and response apps.
Where back in the early 2000's Intel had a Fast Cheap Chip, that often was faster than the big boys especially for the work that was requested for it, and often much of the other guys advantages, didn't make up for its price. So you can get 2 PC Based that servers for the price and runs 75% each as fast of one Mini-Computer Server.
From what I have been hearing the new Apple Chip is superior to Intel offerings, which I will take a face value, as I am not going to spend money an Apple Computer just to see how fast the CPU is. But Intel is just discrediting Apple saying they just make toys, while Intel is suppose to do real work.
Having been an Apple Computer owner in the past, (I guess I am still one, however it is just in my closet for the past 4 years) they are perfectly acceptable computers that you can do a lot of work with. I just haven't continued getting Apple Computers because of their limited options. Meaning if I am going to pay a few grand for a computer, I want features that I want to have vs features what Apple wants me to have. For my current computer, I wanted a better GPU, at the expense of battery life and form factor.
Re: 'Lifestyle Company` (Score:2)
âoe Intel thinking of Apple as a `Lifestyle Company` is one of the reasons that they are in this situationâ
But heâ(TM)s right. Apple IS a lifestyle company. One that just happens to make computers on the side.
Apple is an electronic luxury goods maker now, not a computer company. They sell the dream of an affluent life, no different than BMW or similar companies. They just sell luxury goods with silicon chips in them.
Re: (Score:2)
The fact that people are triggered by Apple being described in that way says more about them than anything. Apple has been a "lifestyle company" since it was saved from the brink by multicolored iMacs and 5MB iPods. It has been solely a lifestyle company longer than some posters here have been alive. With Apple it's not what you do, it's how you do it. It's not about getting work done, it's that your work centers around using Apple products. Every ad they produce is 100% pure lifestyle, it's not even c
Re: (Score:2)
Apple is not a tech company, they're a fashion company that uses tech to get its fan base to over pay for pretty stuff that is guaranteed to be obsolete in a short time
Re: (Score:2)
Re:'Lifestyle Company` (Score:4, Insightful)
Hopefully, that's just for spin purposes. If their thought processes are that badly off-reality, they've got no chance at competing. It would be a shame. I'm really rooting for Intel, but they've been king for so long that they may have forgotten how to take a competitor seriously.
Re: (Score:2)
Apple has been a lifestyle company for decades, Intel has never made a calculator. There is no way to interpret your post as have even the slightest insight.
Re: (Score:3)
And CEOs who view technology as a market segment is how they dug in so deep. Intel *should* be the best at making CPUs, but they dropped the ball, repeatedly. They did this because their brains left them and went to AMD, Apple, ARM, etc. and Intel is left with an upper layer of idiot MBAs who can't find their ass with both hands.
Re: (Score:2)
Intel thinking of Apple as a `Lifestyle Company` is one of the reasons that they are in this situation.
It's possible he actually thinks that way (apparently his original Intel recruiter 40 years ago referred to him as "smart, arrogant, aggressive - he'll fit right in).
But it's also possible he knows his audience, and realizes that Intel is unwilling to see itself as lagging behind anyone - so he played to his base, while still correctly identifying what they need to do.
Re: (Score:2)
"lifestyle company" is an outstanding description of Apple.
Re: (Score:2)
Re:'Lifestyle Company` (Score:5, Insightful)
That's not at all what it means. Lifestyle brands focus more on the brand than the product as a way to sell. It doesn't matter as much what they are selling, the customer of lifestyle brands will buy the product because they want to be seen as the kind of person who owns %product%. And that is very true for many Apple customers. Apple long ago succeeded in making their consumer electronics appear as luxury fashion items. In this way they are pretty unique in the electronics world. Most lifestyle brands I can think of started as clothiers. Nike. Supreme. Porsche and Ferrari even got into it.
Intel in contrast is not a lifestyle brand. Nobody wears their branding trying to look cool.
Re: (Score:2)
Re: (Score:2)
Lifestyle brands focus more on the brand than the product as a way to sell.
Indeed. Apple sells the sizzle and not the steak.
Intel's problem is that once the sizzle fades, Apple's steaks are actually very good. Better than Intel's.
Re: (Score:2)
You could compare to the term "consumer electronics," where "consumer" is the operative word. We're not looking for high-end technical excellence, we're looking for something that works well for most people's usage and that looks nice in their living room.
The typical consumer doesn't care if their MP3 player has the best D/A converter in the world if the result is it costs $10,000.
Likewise, the lifestyle consumer doesn't care if the processor in their laptop can beat the competition on benchmarks, if talkin
Re: (Score:3)
It's when a company sells themselves more on their brand name than on competitive product features. For example, Intel 15 years ago.
Re:'Lifestyle Company` (Score:5, Insightful)
The problem though, for this new intel CEO and most anti-Apple zealot, is that they keep trashing Apple for using "lifestyle" for their marketing and thus thinking the hardware is second-rate. Unfortunately for them, Apple has more than competitive features and it's a mistake to only look at Apple's marketing department. Something can be both shiny and powerful.
The fact that their new, first-generation M1 CPU/GPU can beat the crap out of Intel's and AMD's offering at the same power envelopes is proof enough. I can't wait to see the next generations (M1X, M2, etc).
Re: 'Lifestyle Company` (Score:2, Informative)
Sorry but, no. The M1 is a solid mid-range chip. It is nice that Intel and AMD have new competition which is good for everyone else but M1 hasn't crushed anything. Both Intel and AMD produce a fuck ton of chips that vastly out perform the M1. And if I wanted to fill a data center full of CPUs the M1 wouldn't be high on my list.
I own plenty of apple consumer crap at home, (iPhone, iPad, etc) but let's not go all Apple fanboy over the M1. It is undeserving she viewed from a purely technical perspective o
Re: 'Lifestyle Company` (Score:5, Informative)
You missed the "at the same power envelopes" bit. These are not performance chips. They're efficient mid-range chips. But efficiency is important if you want to scale up. Later generations are certainly going to try to fill the gap to run the Mac Pro.
Re: (Score:2)
Efficiency is important if you want to scale up but it's also harder to stay efficient as you scale.
Apple can't cram 128GiB of ram on top of their chip. At some point they're going to have to have external busses. Choices are external RAM which will give a performance hit due to less bandwidth and a, power hit because of the bus, or they have a huge inter chip bus. That's the route Fujitsu has taken, it comes with a power hit of course, but the other thing is it's a bottleneck, and makes your machine NUMA.
Re: (Score:2)
There's no question they're going to have to move beyond SoC and soon. But for right now, it's really not a bad choice for mid-range machines. Their efficiency is not because it's all in one die. It's mostly because the CPU cores are good.
Re: (Score:2)
You missed the "at the same power envelopes" bit. These are not performance chips.
OK, so I think it's clear to all that what we're talking about is that the M1 excels "at the same power envelopes" but only "for a certain range of power envelopes."
If you want the same architecture as the entry-level, low-power chip, only with BFG monster-blasting, mega-performance, where power consumption is only an issue to the extent that it drives up HVAC costs, Apple can't provide that. And there's no clear indication that it plans to.
The point is that talking about Apple as a "lifestyle" company is n
Re: 'Lifestyle Company` (Score:2)
The M1 is crushing everyone at performance per amount of power consumed, which really matters in mobile.
Re: (Score:2)
It really matters in desktop too. Computers are noisy and fans fail early when TDP gets much above 95W. You can't throw cores at something if they're going to just run into the KWh to power.
Re: (Score:2)
Re: (Score:2)
WTF is a "lifestyle company?"
In a generation, the wheel has made one complete revolution. Remember when mainframe makers called all microprocessor-based computers "toys?"
Re: (Score:3)
I think they make condoms.
Re: (Score:2, Informative)
Re: (Score:3)
Go to a physics conference sometime and see all the Macs. So these physicists are into "lifestyle"?
Re:'Lifestyle Company` (Score:4, Informative)
Re: (Score:2)
They most likely run:
Mathlab
R
Jupiter or other python tools.
Biggest mistake for a human: underestimating a physicist.
Re: (Score:3)
A company who spams "Intel Inside" everywhere.
Re: (Score:3)
You mean those uninitiated in homophobic posts like yours.
There is a difference between lifestyles and alternative lifestyles, otherwise the modifier wouldn't be needed.
Also, homosexuality is not a lifestyle.
Pat Gelsinger seems EXCELLENT. (Score:2)
I would like to know more about the negotiations. (Score:2)
I would like to know more about the negotiations. Why $116 million?
VMWare annual revenue for 2020 was $10.811B, a 12.46% increase from 2019. [macrotrends.net] I have no idea why VMWare has so much income.
JOKE: Is it sad that Pat Gelsinger is so poor? He won't have enough money to buy one of the Hawaiian islands! Here's what it's like to stay on the lush Hawaiian island Larry Ellison bought for $300 million [cnbc.com].
Not silicon, though (Score:4, Interesting)
Silicon is reaching the end of its useful lifetime. If they're going to try and step things up, Intel will have to be the first big chip maker to switch to a whole new process. Gallium nitride, possibly, or even a Hail Mary move towards optical processing.
Intel has has a lot of this stuff on the back burner for a while now, it's time for them to find out if they can actually do something useful with it.
Re: (Score:2)
I am expecting them to look to memristor and massive parallelism to keep silicon alive for some time longer
Re:Not silicon, though (Score:5, Informative)
There are plenty of people at Intel who should be able to do this.
I'm a condensed matter physicist and nanotechnologist. When I look at the network of people I've worked with and interacted with over the last 20+ years, Intel is the company that employs the largest chunk of those people. This is not a secret. There's a wide pipeline of scientists from good PhD programs into Intel.
It's time to see if Intel's investment in many of these niche fields will pay off.
It may be too late, or the culture could be too "big company" to allow real innovation to flourish. If they only bring in new managers at the top (C-level) and do nothing about technical leadership (I'm looking at you, Intel Fellows program), then nothing's really going to change. Essentially, their technical leadership program is built to emphasize making incremental improvements in their existing technology, and works too much like a "lifetime achievement award" than something designed to spur growth and creativity.
Re: (Score:2)
Silicon has never been more useful than it is now. It's not reaching the end of anything and won't until there is something better.
Amazon too (Score:2)
Re: (Score:2)
All the cool kids are doing it [theregister.com]
Intel is giving its customers what they want (Score:5, Interesting)
Realistically speaking, Intel's been delivering exactly what its customers want: decent-performing chips at price points that work.
None of Intel's customers want to disrupt their own product pipeline by introducing something that's twice as fast. That would make old inventory harder to sell, and make people realize that "hey, chip performance doesn't actually matter to me."
From a server point of view, who cares what process die you're on? Most data centers don't care, because they don't operate at the scale where the difference is material.
For corporate desktops? Who cares how fast most of them are.
That's the vast majority of Intel's market.
Intel, though wants to see itself as a technology company, not an enterprise vendor. That's really the disconnect.
Re: (Score:2, Insightful)
"Realistically speaking, Intel's been delivering exactly what its customers want: decent-performing chips at price points that work."
Really? I thought AMD did that. Right now, Intel delivers the opposite of what its customers want: CPU full of security holes*, high power requirements and lower processing power per watt.
* AMD too, but not as numerous and/or critical.
Re: (Score:2)
I wouldn't say AMD has fewer (at least with certainty). Most of AMDs are discovered by researchers trying things on Intel chips and then when a flaw is found the technique is also tested on AMD. Intel just has a big marketshare and more rewarding to look for bugs on. Nobody ever thought of security when developing speculative execution - I wouldn't have imagined it as an attack vector either, though I'm not an expert on things like that.
Re: (Score:2)
Re: (Score:2)
A chip is basically secure by default. The OS is the problem.
Re: (Score:2)
whoo eee! are you guys stinking up the place
The point of security holes like Meltdown is that it was Intel specific and related to shortcutting on Intel's part, on the chip allowing one processor thread to read another processor thread's memory.
Meltdown vulnerability impact:
Read arbitrary kernel memory from user space applications
Fully virtualized machines are not affected (guest user space application cannot read host user/kernel space memory)
Hypervisor escape is possible in paravirtualized environment (Xe
Re: (Score:3)
From a server point of view, who cares what process die you're on? Most data centers don't care, because they don't operate at the scale where the difference is material.
Huh? I'm certainly not an expert in these matters, but I'm pretty sure smaller processes yield reduced power consumption, and I'm pretty sure that the electricity used in powering/cooling servers is a significant chunk of a data center's operating costs. I would think data centers would care about that precisely because of the scale at which they operate.
Re: (Score:2)
but I'm pretty sure smaller processes yield reduced power consumption
Only if you make the same processor. Usually you don't. You make a bigger one, and start at square 1.
Re: (Score:2)
How can they be giving their customers what they want, when they are losing market share? I don't say that sarcastically.
AMD has risen a fair bit. I recently built a new PC and I went with an AMD Ryzen 5 3600. Was it better than Intel's comparable offering? I don't know. My last PC was an older core I5. All I can say is that I chose AMD. Maybe it was just marketing? Maybe AMD has a better price/performance point? Sales like mine should have been Intel's.
It's largely the same with mobile chips. Intel has off
Re: (Score:2)
None of Intel's customers want to disrupt their own product pipeline by introducing something that's twice as fast. That would make old inventory harder to sell
That is some incredibly shortsighted thinking that I very much doubt the OEMs are engaging in.
If Intel were putting out significant improvements year-after-year, sure, it renders old inventory obsolete sooner, and that is an issue you want to avoid, but it renders consumer purchases obsolete sooner too, which is something almost every company wants to see. People have a reason to upgrade on a regular cadence if there are massive improvements YoY because if they don’t keep up everything will slow down
Too many business degrees (Score:5, Interesting)
"we're failing as a company ... (Score:5, Funny)
... but at least we're not lame like those other guys."
I don't think anyone is impressed when someone deflects like this. It's painfully transparent.
It's not going to happen (Score:5, Interesting)
Until the path to promotion and recognition at Intel no longer requires being involved in the design and sale of high-margin CPUs and server chips, nothing at Intel is going to change.
Being stuck in a team that designs low-margin products is career suicide at Intel, and yet that has been the foundation of the entire mobile revolution, which the company completely missed out on.
I think the situation at Intel is going to get a lot worse before it gets any better.
This is their own fault. (Score:5, Insightful)
For 20+ years they little to no competition. They were a virtual monopoly. Look at Linus's rants about lack of ECC on consumer devices and the comment about Xeon, 'Twice as many cores for 4x the price'. They've been pushing profits over innovation for at least 15 years. Now they suddenly they have competition ( AMD, Apple M1 and a Host of other ARM CPU's) they are crapping in their pants. I agree with the other posters. I hope they figure it out more competition is better. However, they have no idea how to compete. All they know is being a monopoly!
Better watch out for Nvidia (Score:5, Interesting)
He should be more worried about Nvidia getting into the CPU business with Arm than about what a "lifestyle company" has been doing with Arm CPUs. Apple has demonstrated what one can achieve with Arm CPUs on the desktop, but when one considers Nvidia's graphics, AI and super-computing experience and chip-making ability, and their purchase of Arm, then there is a far more serious competitor they should be worry about. With Intel weakened and AMD making their own GPUs, does Nvidia not have a strong partner in the x86 business as they used to. Nvidia could decide to step up and to make Arm-based designs of their own. What then, Intel?