Well, if you are using Apple products?
Not good enough quality as they are made in China? 😏
Or, maybe Volvos too?
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.
Apple seems to have dodged a bullet in going with their own Silicon
- Thread starter pshufd
- Start date
- Sort by reaction score
Not good enough quality as they are made in China? 😏
Or, maybe Volvos too?
I have an i7-10700 build and it's pretty good. Power usage is low and CPU temps are typically in the 30-40 degree range. It often runs similarly to my Mac Studio in thermals. But that's because I designed it to run cool and quiet. What's been annoying is Intel requiring motherboard changes frequently and their apparent approach of increasing performance through higher power requirements.
If I needed a new Windows build right now, I'd look at AMD. I'm not a gamer. I only have one program that runs a lot better on Windows than it does on Apple Silicon and, at the moment, I'm back to running it on Apple Silicon so that I don't have to have a Windows system on my desk. I may go back or may not.
My default has always been to go with Intel except back around 2004 when we bought two AMD systems and they were great.
I think we are starting to see the end of x86, performance gains from each new generation are tiny and they seem to be getting harder to manufacture. I have just sold my 7950x3d 4090 gaming PC and gone back to an M3 Pro chip. Single core performance is noticeably faster on the M3 Pro than the 7950x3d for a fraction of the heat and power. I just use GeForce Now for gaming on the Mac.
This is a minor blip and a not unusual engineering problem which has been somewhat sensationalised. Sometimes there are manufacturing flaws that show up in production that you don't find before hand. And it's really quite difficult getting to the bottom of the problem.
I would suggest that if you think this is bad you want to see what happens during post market surveillance in the pharmaceuticals industry.
What has been wonderful was the share price tank which I bought a crap load of at the bottom of the tank.
Edit: going into slightly more detail, this is a mortal risk for Apple too. They are an early adopter of TSMC nodes. If TSMC make the same mistake, which is certainly not impossible, then things will be similarly problematic. When I worked in the industry a long time ago, we used very old processes (350nm CMOS) because we needed stuff to work reliably.
I would suggest that if you think this is bad you want to see what happens during post market surveillance in the pharmaceuticals industry.
What has been wonderful was the share price tank which I bought a crap load of at the bottom of the tank.
Edit: going into slightly more detail, this is a mortal risk for Apple too. They are an early adopter of TSMC nodes. If TSMC make the same mistake, which is certainly not impossible, then things will be similarly problematic. When I worked in the industry a long time ago, we used very old processes (350nm CMOS) because we needed stuff to work reliably.
Apple is an early adopter of TSMC? I thought Apple is TSMC client since 2013-2014? maybe im mistaken and more important it worked by now probably in billions of devices iphones ipads and now macs. Of course ipads and iphones are not macs or servers to work 24h/7 to test its reliability
Maybe im getting this wrong
Maybe im getting this wrong
Early adopter of TSMC nodes (processes). They go for the most advanced, smallest and highest risk process every time.
The range for Intel has been about $12 to $76 this century.
It may be a minor blip but what's more annoying for customers is the lack of candor in talking about the problem, identifying the chips with the problems and then proposing a way to make those customers whole. It appears that the initial problem was in 2022 and this is the first that I'm hearing about it.
I've traded many companies that have crashed due to self-inflicted errors in strategies or product issues. Come up with a plan, communicate it to shareholders and take care of your customers. I've not seen any of those three things from Intel. Sometimes a stock gets hit because of market issues, overvaluation or recession. That is not the case with Intel here.
If things were actually as great as you say, why are they laying off 15k employees this year?
“Simply put, we must align our cost structure with our new operating model and fundamentally change the way we operate,” Intel CEO Pat Gelsinger wrote in a new note to employees. “Our revenues have not grown as expected—and we’ve yet to fully benefit from powerful trends, like AI. Our costs are too high, our margins are too low. We need bolder actions to address both—particularly given our financial results and outlook for the second half of 2024, which is tougher than previously expected.”
Intel will also aim to reduce its operational costs, shave down its teams’ portfolios, and do its best to reduce costs. It’ll also suspend its stock dividend. “I have no illusions that the path in front of us will be easy. You shouldn’t either,” Gelsinger added. “This is a tough day for all of us and there will be more tough days ahead. But as difficult as all of this is, we are making the changes necessary to build on our progress and usher in a new era of growth.”
Intel Plans to Lay Off 15,000 Employees
Intel will lay off 15,000 employees, or roughly 15 percent of its staff. Intel will also aim to reduce its operational costs, shave down its teams’ portfolios, and do its best to reduce costs. It’ll also suspend its stock dividend.
Indeed. They are bread and butter for OEMs so you probably can't lose over time.
I think that's a lot of sensationalism talking from social media and the YouTube channels rather than an honest engineering discussion and I would suggest that the latter is what we should focus on. Here are some problems which tend to occur here which can make investigation difficult:
- The SKUs which were reported as affected were the -K and -KS ones initially which are firstly unlocked and secondly the lowest number of shipped units from the series.
- Most of those SKUs are in end user hands so there are at least two layers (store / distributor) in front of Intel when it comes to getting failed samples back in their hands. Having worked with buying large quantities of Intel stuff from distributors before, it's very difficult to get them to report anything back to Intel to the point that Intel tended to reach out to even low volume resellers to get info out of them.
- It's really difficult to get the full picture of the failure conditions because we have a combination of a third party OEM board manufacturer, another third party BIOS/UEFI vendor and the final system integrator who all have to get together and produce something that works.
- It's even more difficult picking out statistical failures (yes some failures are expected!) from unexpected failures. It takes quite a large sample to even work that out.
- On top of that they have to filter out all the people who have blown up their CPUs through overclocking them, poor cooling or incompetent engineering.
Intel are still shipping millions of boring CPUs absolutely fine. This is a fairly low amount of noise.
This is the start of the current tech collapse which is a combination of the AI bubble bursting, our current geopolitical situation and market conditions. Also 100% not specific to Intel - they are just the first company to (wisely) react to the market collapse. They went all in on AI because Microsoft demanded that they start shipping AI PCs on the market and Intel had to meet that market demand or lose business to an ARM OEM. Well it turns out that was and still is a completely non viable market so they are literally shaving off the mistake quickly before the real down turn kicks in. By mid-2025 nearly all the companies out there that invested in this will have little if nothing to show for the investment and that's when it hurts. The big investment companies already pulled out and are letting the individual investors pay for their gains.
Anyway, look forward a year if they don't screw it up. They have new nodes online, geopolitical security, a viable low power mobile strategy and a better architecture to deliver what people really want: millions of corporate laptops exactly but less hot and slightly faster than the last generation. That's a much better situation than the competition. Where do you think the attention is going?
The biggest risk on the market is nothing to do with any fab or hardware provision: it's the software fads.
Companies would lay off people all the time, as living beings are hampering the profit income. Machines, computers, AI and whatnot would be used in increasing profit as they don't get sick, don't have families to look after, won't go on vacations etc, etc. If a company says that they care for their customers, or even their employees, they are lying. It is always profit, nothing else. In board meetings they don't discuss how to lose money.
Hahah, you got me 😔... Generally speaking low standards but not when it is Designed in California 😏
(All countries have good products, I didn't mean to be a hater)
Clearly there have been a lot of INTC losers over time.
So you're saying that there's a problem but it only shows up in higher SKUs. How do we know that this won't show up in lower CPUs?
Server farms are reporting the problem and those aren't exactly the type of customers that overclock.
Yes, it is really difficult to get the full picture. Why is that? They've had the problem in their hands for five months. Or back to 2022 from the manufacturing side.
I worked in corporate engineering support for several years so I know the headaches involved getting problems from local support through regional support up to my level, reproducing it, diagnosing it or working with an engineer to diagnose it and then come up with a workaround for the customer, and then a permanent fix. And also doing the serviceablity engineering work so that local support people could better diagnose problems. And yes, we dealt with a ton of intermittent failures. Sometimes you're in a room with VPs in a meeting discussing the problem and the impact that it's having on their business. An operation to provide this kind of service is expensive but Intel has a lot of money. Where are the results? Have they put a hold on the product? I had to put an engineering hold on a product once. The product manager was going nuts because of the amount of revenue that would be lost but it was the right thing to do. And then we came up with a plan to diagnose, resolve, fix and send the fixes out to customers. Why isn't Intel doing at least some of that?
Google started the layoffs in January 2023 and repeated them in 2024. That was the real start. I've been in the tech world since the 1970s so I've seen many economic cycles and new tech wiping out old tech. Intel is getting $8.5 billion from the CHIPs act. You'd think that they could at least reassign their workers instead of doing mass layoffs. I read some of the stuff that the CEO wrote and this looks more like cost controls more than right-sizing. Suspending the dividend isn't exactly a vote of confidence for shareholders.
The software fads are great for trading. I'd rather buy CRWD than INTC right now.
They are laying off 15K after getting $8.5 billion in Chips act money, ostensibly to add a lot of employees in the US.
Increasing profit? If that's the case, why are they also suspending the dividend?
"Designed in California" doesn't say much, does it? By whom would be more appropriate. I have a feeling that, after all that "Designed in California"talk, most of it is designed and made in China.
When did a company told you the truth?
If they did, there won't be much profit, would there?
Jim Cramer has an acronym: UPOD, underpromise, over-deliver.
I know that's challenging for executives that depend on personal stock options wealth.
At a board meeting, they never discuss value to the customer, but value to themselves and their companies. When you hear value added, it means profits added, and as the customer, you are at the losing end, always.
AMD CPU and Nvidia GPU is more common for sure. Nvidia has c.80% market share of dedicated GPUs. There is very little ‘special sauce’ associated with having both components from AMD.
I was thinking specifically of Apple in this case, though they probably have a global design system, most is probably made at Apple Park. I agree with "By whom would be more appropriate." that's what really matters in the end.
I listen to conference calls as I don't have access to board meetings. In conference calls, they sometimes talk about value added to the customer and sometimes talk about future profits and sometimes they talk about both. Ideally, you are making products that do more for your customers and add to your customers value equation. I feel that Apple Silicon was in that category. Technology should be able to do that. Give some of the benefits to the customer and some to the shareholders.
What do I expect from Intel as a customer? What the problem is and how they are going to fix it. What do I expect if I were a shareholder? I've said that above.
We don't. Yet. It's impossible to tell.
Well the server farms in question are not what you'd call "enterprise" customers. They are running game servers with CPUs that probably shouldn't be crammed in a rack anyway. And they didn't exactly run the hardware in on a test bed or anything for a couple of months. Straight into production. Then they talked among themselves and to YouTubers.
Yep. I've seen defects in software take years to even getting around to being reproduced. Having worked on the hardware side, it's even harder there. Just getting time on the SEM at the company is hard enough!
They are definitely doing all that. The communication is the issue here. I suspect it's probably corporate politics more than anything.
Large corporations tend to have poor control of recruitment and direction due to the layers. That is fairly standard. Also why I refuse to work for them.
Yes it's cost controls. They overspent on fab and senior shareholders are pissed that they don't get their dividend. They want money not stable technology progression.
I wouldn't buy CRWD until they have worked out if they can sue them or not.
I've already traded it once off the bounce. I'm trying to determine if it's a worthwhile swing. They basically have a monopoly and will hopefully radically improve their QC processes.
I worked in a large cloud company before and the stuff I've seen with regards to bugs ...
Fortunately nobody actually ran into them.
Cloud bugs can compromise all of your customers at the same time. And all of their customers.
The customers might get some value, if that value brings in more profit to the company, otherwise, that value is forgotten and deliberately hidden from the customers.
You should discuss that with the Intel (or Intel forums, if they have something like that), but whether you'd get a coherent answer is doubtful. Apple had stopped being customer of Intel, and had become more like a competitor. But, Apple might go back to Intel, after the ARM experiment, who knows.
The M1 was really a monumental change for Apple. It's a competitive chip even heading towards the end of 2024 and the onslaught of lots of competition. Apple Silicon quite literally saved the Mac in my opinion. Post M1 we got the incredibly fast, efficient, and quiet MacBook Air, the MacBook Pro with all the ports back, amazing displays, incredible speakers, and insanely powerful chips. We got the Mac Studio which is arguably a less expensive much more compact Mac Pro. I think Macs are the most exciting products in Apple's lineup now, which is opposite of the way it was pre-M1.
The end of X86 would be decades away. The slowdown and stop of x86 advancements might come sooner though. For most folks, as long as ARM or RISC-V can emulate any and all exiting x86 programs (including those dated back in 1980s) with zero compromises, including performance, then that’s the point when x86 hardware becomes truly obsolete for the majority. I dunno if we will get there eventually however.
HP stopped selling DEC Alpha systems in 2007 according to Wikipedia. I recall selling some of our software assets running on Alpha a long time ago and talked to the people that did the support and they said that they ran on x86 hardware running emulating or translating Alpha code. So it's something that could eventually happen. I personally think that computer users should migrate when they can and put some effort into it instead of being reactive. Of course next summer, there will be a ton of calls for IT folks to upgrade offices full of PCs from Windows 10 to Windows 11 and lots of other organizations trying to figure out how they can still run Windows 10 with support.