Processor ERRORS and marketing missteps take place on both sides.

Intel has problems with their latest AI PC processors having a performance killer bug in the microcode.  So far, this bug is seen an un-fixable bug that will require a factory microcode change at the minimum to fix it (i.e. a freshly built processor made with the fix integral to it).  

Think of something on the order of Meltdown or Specter for severity and size.

No processors with this error have left Intel's control according to Intel, so this mess is controllable but will entail yet another delay for Intel's first market entry.   This turns out to be completely false as all production run processors of the last two generations are busy bricking themselves as we speak.   Qualcomm and Samsung are already out there using ARM based processors, so this screw up further re-enforces Intel's non-leader position in AI PC space.


===================================================


AMD has been caught by review houses overstating some of their metrics and basic performance results on several new processors.  This was due to the test drivers giving inflated answers.  Updated test drivers are promised from AMD to fix these overstatement issues as they only actually affected a few of the most modern AMD processors.  

Intel does shite like this routinely on purpose, but AMD historically never does stuff like this at all --  this was likely a test metric error from the testing vendor or a plain old goof on AMD's part.  

Everybody is using AI PC functions at some level with their newest processors, but expecting equal results from test suites using different set ups and different types of processors is simply asking a bit too much so early in the changeover.

Still, as all the players hustle to get over their starting line AI PC bobbles that are certainly being made on both sides --- then you have to throw on top of the steaming heap the simply ill defined operating standards for AI PC in general and the risk levels for buying right now rise above acceptable levels every which way you go.

Still, the early results say AI PC works faster on some items by a whole bunch of a lot ......

https://www.computerworld.com/article/2514506/copilot-ai-pcs-are-finally-here...


The promise of these new devices is that they’d herald the arrival of AI tools to the Windows masses. That’s not really the case right now.

The AI hype just keeps on coming.

The latest news is the arrival of an entirely new line of Windows computers, Copilot+ PCs, which are specifically designed with artificial intelligence (AI) in mind. Microsoft claims they’ll dramatically speed up AI, offer new features unavailable to other PCs, and deliver improved battery life. The new machines point the way to the future of Windows and of AI, if the company is to be believed.

Laptops from Acer, ASUS, Dell, HP, Lenovo, Samsung, and Microsoft were released several weeks ago, long enough to find out how they perform in real life. So how do they stack up? Are they everything Microsoft claimed they would be, or just one more overhyped new technology?

To find out, let’s start by looking at Microsoft’s promises about what the Copilot+ PCs will do. In a blog post announcing them, the company crows:

“Copilot+ PCs are the fastest, most intelligent Windows PCs ever built. With powerful new silicon capable of an incredible 40+ TOPS (trillion operations per second), all–day battery life and access to the most advanced AI models, Copilot+ PCs will enable you to do things you can’t on any other PC. Easily find and remember what you have seen in your PC with Recall, generate and refine AI images in near real-time directly on the device using Cocreator, and bridge language barriers with Live Captions, translating audio from 40+ languages into English. “

The laptops are based on Qualcomm Arm-based processors, which include a neural processing unit (NPU) to handle AI-related tasks. Normally, AI processing occurs in the cloud rather than on a local PC, potentially slowing things down AI. On Copilot+ PCs, Microsoft claims, much of that processing will stay local on the machine.

Recalling Recall
Microsoft went into hype overdrive when touting the new machines’ Recall feature. There’s good reason for that. Anyone who has spent too much time trying to remember and open a specific email, website or file they worked on months ago would want it — and that pretty much means all of us. It’s clearly the killer app that could sell countless Copilot+ PCs.

But Recall has an Achilles heel. As I wrote earlier, it could be the ultimate security and privacy nightmare. It works by constantly taking screenshots of everything you do, storing them on your PC, creating a searchable database of them, and then using AI tools on them so you can find what you want quickly.

Initially, Microsoft claimed that because all that work is done locally rather than in the cloud, it wouldn’t lead to privacy or security issues. But many security researchers and analysts disagree.


Jeff Pollard, vice president and principal analyst at Forrester, told Computerworld “I think a built-in keylogger and screen-shotter that perfectly captures everything you do on the machine within a certain time frame is a tremendous privacy nightmare for users.”


If a hacker gains access to your PC, researchers found, he or she can read the database, which isn’t even encrypted. At first, Microsoft tried to convince everyone that the privacy issues were much ado about nothing. But then it backed off. The company announced in a blog post that the feature won’t be available on Copilot+ PCs when they launch. Microsoft says it will make Recall available some day — though it won’t say when.

That means the biggest reason for buying a Copilot+ at the moment remains elusive.

Other Copilot+ PC woes
These machines have other issues, too. One of the most head-scratching ones is that the Copilot app on Copilot+ PCs appears to be less powerful than the app on traditional PCs. On Copilot+ PCs, Copilot runs as a traditional Windows app rather than as a sidebar pane, as it now normally does on traditional PCs. So, you can resize it, move it around the screen, and do anything with it that you can do with any window.

That’s not the problem. The problem is that Microsoft also took away some Copilot features. When run as a sidebar pane, Copilot can perform some basic Windows tasks for you, such as turning dark mode on or off. The app on Copilot+ PCs can’t do that. (By the way, Copilot as a Windows app is now also available for non-Copilot+ PCs, and it has the same problem as the Windows app on Copilot+ PCs.)

Another oddity: Although the new Copilot+ PCs have a dedicated Copilot key, the PCs won’t allow you to launch Copilot with the keyboard shortcut Windows key-C as you can on other PCs. Go figure.

And there’s more, according to Computerworld and PC World contributor Chris Hoffman. On the new machines, he says, “Copilot doesn’t run offline or use the new integrated neural processing unit (NPU) hardware to do anything at all.”

Running AI offline was one of the big promises of the new line. Perhaps someday that will happen, but as Hoffman notes, that day isn’t yet upon us.

Emulation: Thumbs up or thumbs down?
Because Copilot+ PCs run Windows on an Arm chip, they have to run Windows apps via emulation. Theoretically, that could be problematic or slow apps down. Microsoft contends that the chips are so fast that the apps run fine.

Not everyone agrees. Many reviewers generally report no serious problems, but Android Authority warns: “emulation is hit-and-miss.”

PC World’s Mark Hachman found that most apps work fine, with one big caveat: “There’s a good chance your favorite games won’t even run” on a Copilot+ PC.

The upshot
So, should you buy one of these machines? I won’t hem and haw. The answer is no. Their two most important AI-related features — Recall and local AI processing — aren’t yet available. And running games on one, if that’s a priority, is iffy at best.

There are plenty of very good thin, powerful Windows laptops out there. If you need a new PC, buy one of those, not a Copilot+ PC. Even if you’re looking for true AI power, you’d do better to wait.

In the past few days Microsoft "system update based errors" have temporarily shut down many international airports and banking centers.   Although relatively quickly fixed by manual programming efforts, this sort of episode will repeat itself multiple times as the "system update errors" spread out through the world of MS based sofwares.

Knowing human nature, these first attempts at using AI PC will likely be seen very very shortly as "dangerous" and likely be temporarily abandoned and/or banned as an official pathway by governments and key businesses.

However, the use of 40-70 TOPS co-processor power will continue to be quietly developed and quietly utilized by select softwares and applications.



Applications that can grab on to a co-processor and do their core work much much faster will be competitive winners in the real world.

Intel's ongoing interrupt flaws have now been analysed by several impartial outside sources at this point in time.

Consensus is that the high voltage boosts Intel was using to claim their current processor speeds are (within 6 months no less) was eroding the processors core compute processes to the point of failure (which result in the fatal interrupts) and these have accelerated to become an early version of complete processor failure.

Give it another six months and the processor WILL brick itself.

Intel had tried sending out updates that cut back on the high processor voltage, but the already damaged processors continued to decline and the loss of advertised processor speed became a real issue all by itself.  

People applied for RMA on the defective processors, but Intel initially denied these claims and pushed it off on the motherboard makers, who then took legal actions against Intel because they had designed and built their motherboards using Intel's very own spec sheets .......    Intel had even certified all the motherboards after an extensive Intel review process, so the thermal degradation error was 100% all Intel's fault.

Intel has now begun accepting the required RMAs and there will be enough of them to put Intel in serious danger of losing all credibility and also losing profitability on the last three years worth (two generations) of Intel processors.    

Intel currently has no good processors to replace the defective processors with, anyway.

Please note,  AMD has not got this issue and will be very careful not to get into any of it.

And again, remember to buy only AMD processors and avoid all the big box vendors who still remain locked into Intel's nasty little over-voltage mess.

https://www.theverge.com/2024/8/1/24210656/intel-is-laying-off-over-10000-emp...

In the last two days, Intel has issued cut back notices that are intended to keep the company from going under as the total cost of their over-voltage driven processor crashes and the costs to repatriate the damages arising from the issue become more clear.  

Here’s the full memo from Gelsinger:

Team,

We have moved our All Company Meeting to today, following our earnings call, as we are announcing significant actions to reduce our costs. We plan to deliver $10 billion in cost savings in 2025, and this includes reducing our head count by roughly 15,000 roles, or 15% of our workforce. The majority of these actions will be completed by the end of this year.

This is painful news for me to share. I know it will be even more difficult for you to read. This is an incredibly hard day for Intel as we are making some of the most consequential changes in our company’s history. When we meet in a few hours, I’ll talk about why we’re doing this and what you can expect in the coming weeks. In advance of that, I wanted to preview some of what’s on my mind.

Simply put, we must align our cost structure with our new operating model and fundamentally change the way we operate. 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.

These decisions have challenged me to my core, and this is the hardest thing I’ve done in my career. My pledge to you is that we will prioritize a culture of honesty, transparency and respect in the weeks and months to come.

Next week, we’ll announce a companywide enhanced retirement offering for eligible employees and broadly offer an application program for voluntary departures. I believe that how we implement these changes is just as important as the changes themselves, and we will adhere to Intel values throughout this process.

Why Now?

Since introducing our new operating model, we have taken a clean-sheet view of the business and assessed ourselves against benchmarks for high-performing foundries, fabless product companies and corporate functions. This work made it clear our cost structure is not competitive.

For example, our annual revenue in 2020 was about $24 billion higher than it was last year, yet our current workforce is actually 10% larger now than it was then. There are a lot of reasons for this, but it’s not a sustainable path forward.

Beyond our costs, we need to change the way we operate – something many of you shared as part of our Employee Experience Survey. There’s too much complexity, so we need to both automate and simplify processes. It takes too long for decisions to be made, so we need to eliminate bureaucracy. And there’s too much inefficiency in the system, so we need to expedite workflows.

Key Priorities

The actions we are taking will make Intel a leaner, simpler and more agile company. Let me highlight our areas of focus:

Reducing Operational Costs: We will drive companywide operational and cost efficiencies, including the cost savings and head count reductions mentioned above.

Simplifying Our Portfolio: We will complete actions this month to simplify our businesses. Each business unit is conducting a portfolio review and identifying underperforming products. We are also integrating key software assets into our business units so we accelerate our shift to systems-based solutions. And we will narrow our incubation focus on fewer, more impactful projects.

Eliminating Complexity: We will reduce layers, eliminate overlapping areas of responsibility, stop non-essential work, and foster a culture of greater ownership and accountability. For example, we will consolidate Customer Success into the Sales, Marketing and Communications Group to streamline our go-to-market motions.

Reducing Capital and Other Costs: With the completion of our historic five-nodes-in-four-years roadmap clearly in sight, we will review all active projects and equipment so we begin to shift our focus toward capital efficiency and more normalized spending levels. This will reduce our 2024 capital expenditures by more than 20%, and we plan to reduce our non-variable cost of goods sold by roughly $1 billion in 2025.

Suspending Our Dividend: We will suspend our stock dividend beginning next quarter to prioritize investments in the business and drive more sustained profitability.

Maintaining Growth Investments: Our IDM2.0 strategy is unchanged. Having fought hard to reestablish our innovation engine, we will maintain the key investments in our process technology and core product leadership.

The Future

I have no illusions that the path in front of us will be easy. You shouldn’t either. 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.

When we began this journey, we set our sights high, knowing that Intel is a place where big ideas are born and the power of what’s possible triumphs over the status quo. After all, our mission is to create world-changing technologies that improve the lives of every person on the planet. And at our best, we have exemplified these ideals more than any company in the world.

To live up to this mission, we must continue to drive our IDM 2.0 strategy, which remains the same: re-establish process technology leadership; invest in at-scale, globally resilient supply chain by expanding manufacturing capacity in the U.S. and EU; become a world-class, leading-edge foundry for internal and external customers; rebuild product portfolio leadership; and deliver AI Everywhere.

Over the past few years, we have rebuilt a sustainable innovation engine that is largely in place and on track. It’s now time to focus on building the sustainable financial engine needed to drive our performance. We must improve our execution, adapt to new market realities and operate as a more agile company. That’s the spirit of the actions we are taking – knowing that the choices we make today, as difficult as they are, will strengthen our ability to serve our customers and grow our business for years to come.

As we take these next steps in our journey, let’s not forget that there has never been a greater need for what we do. The world will increasingly run on silicon – and the world needs a healthy and vibrant Intel. That’s why the work we are doing is so consequential. Not only are we remaking a great company, but we are also creating technology and manufacturing capabilities that will reshape the world for decades to come. And this is something we should never lose sight of as we push forward in pursuit of our goals.

We’ll talk more in a few hours. Please come with your questions so we can have an open and honest discussion about what comes next.

Update, August 1st: Added more details from the earnings call about Intel’s expectations for its upcoming Lunar Lake and Panther Lake chips.

https://www.tomshardware.com/pc-components/cpus/intels-stock-drops-30-overnig...

Intel's stock drops 30% overnight —company sheds $39 billion in market cap


Intel's stock dropped around 30% overnight, shaving some $39 billion from the company's market capitalization since rumors of a pending layoff first emerged. The devastating results come after the chip giant reported a loss for the second quarter, complained about yield issues with the Meteor Lake CPU, provided a modest business outlook for the next few quarters, and announced plans to lay off 15,000 people worldwide.  

When the NYSE closed on July 31, Intel's market capitalization was $130.86 billion. Then, a report about Intel's massive layoffs was published, and the company's market capitalization dropped sharply to $123.96 billion on August 1. Following Intel's financial report yesterday, the company's capitalization dropped to $91.86 billion. Essentially, Intel has lost half of its capitalization since January. As of now, Intel's market value is a fraction of Nvidia's worth and less than half of AMD's.

As Intel's actions look rather desperate, analysts believe that Intel's challenges are existential. "Intel's issues are now approaching the existential," said Stacy Rasgon, an analyst with Bernstein, told Reuters.

Indeed. Intel is fighting numerous rivals. On the one side, it is competing against AMD, Nvidia, and now Qualcomm for revenue share in the consumer PC market. For now, Intel outsells all three companies easily in this market, though Nvidia's gaming business looks to be more profitable.

On the other side, the company competes against AMD, Nvidia, and Arm chips in the data center space. Thanks to Nvidia's highly popular AI GPUs, Nvidia outsells AMD and Intel combined by nearly 3.3 times. But TSMC is arguably the biggest fight for Intel. The Taiwanese company makes chips for all of Intel's rivals and Intel itself.

On the roadmap side, Intel looks quite competitive both in terms of performance and, eventually, in terms of costs. Yet, the company has to prove that it can make money making chips not only for itself, but for others. To do so, it needs to persuade TSMC's customers to use Intel's technologies, which isn't easy given that the Taiwan foundry can efficiently produce chips with great yields.  

Rasgon believes that under different circumstances, there might be discussions about the company's viability. However, Intel could boost its balance sheet by $40 billion by 2025 through its current actions, subsidies, and partner contributions, ensuring its survival in some form.

"Intel will survive (in some form) to continue the fight," Rasgon told Reuters.

Intel reported $12.8 billion in revenue for Q2 2024 and faced a substantial loss of $1.6 billion, a significant drop from a $1.5 billion profit in Q2 2023.

Looking ahead, Intel's projections for the third quarter of 2024 are concerning. The company anticipates revenue between $12.5 billion and $13.5 billion, which would be approximately $1.2 billion less than the revenue reported in Q3 2023 and just a bit higher at the midpoint than in Q2 2024. Although the company's product divisions make a profit, the company's Intel Foundry manufacturing arm lost $2.8 billion in the second quarter alone. As the company does not expect a rapid recovery, it decided to lay off some 15,000 personnel to reduce its costs.

Well, why don't they have AI design the chips?