Intel just took the worst beating in earnings in over a decade

Posted:
in General Discussion edited January 2023
Intel's fourth quarter in 2022 encompassing the holiday season was the worst in the company's history, as it reported a 32% drop in revenue since the holiday quarter of 2021.

Intel posts loss for 2022
Intel posts loss for 2022


The company shared its results for the fourth quarter of 2022 and the entirety of the year. Its fourth-quarter revenue of $14 billion was down 32% year-over-year, and its revenue for the entire year was $63.1 billion, down 20% year-over-year.

Intel further said that its earnings-per-share for the fourth quarter was $0.16 and $1.94 for the entire year. It forecasts revenue of $10.5 billion for the first quarter of 2023.

In the fourth quarter, Intel generated $7.7 billion in cash from operations and paid dividends of $1.5 billion.

"In the fourth quarter, we took steps to right-size the organization and rationalize our investments, prioritizing the areas where we can deliver the highest value for the long term," said David Zinsner, Intel CFO. "These actions underpin our cost-reduction targets of $3 billion in 2023, and set the stage to achieve $8 billion to $10 billion by the end of 2025."

In the fourth quarter, Intel generated $7.7 billion in cash from operations and paid dividends of $1.5 billion
In the fourth quarter, Intel generated $7.7 billion in cash from operations and paid dividends of $1.5 billion


Intel believes it can regain leadership in transistor and power performance by 2025, saying that its making progress with its goal of producing five process nodes in four years. The Intel 7 node is now in high-volume manufacturing for client and server markets, and Intel 4 is now ready for manufacturing.

Apple used to be one of Intel's largest customers, but since it started producing its own chips, Intel has had to rely on other companies. However, the chip manufacturer hasn't lost hope.

Intel still believes it can win over Apple again, saying that to do so it must create better chips than Apple Silicon. As an alternative, Intel could produce Apple's chips for the company, alongside Apple's existing relationship with the Taiwan Semiconductor Manufacturing Company (TSMC).

Read on AppleInsider
«1

Comments

  • Reply 1 of 30
    The article doesn't really explain WHY Intel sales were so bad.  Especially during the holiday season.  Is new gear too expensive for consumers?  Are people keeping their computers longer?  Or are consumers moving towards more mobile devices?

    It doesn't help Intel that major sellers like Dell have raised their prices making one take a second thought about buying or upgrading.  
    FileMakerFellerwatto_cobran2itivguy
  • Reply 2 of 30
    thttht Posts: 5,499member
    ITGUYINSD said:
    The article doesn't really explain WHY Intel sales were so bad.  Especially during the holiday season.  Is new gear too expensive for consumers?  Are people keeping their computers longer?  Or are consumers moving towards more mobile devices?

    It doesn't help Intel that major sellers like Dell have raised their prices making one take a second thought about buying or upgrading.  
    Post-pandemic bust.

    MS reported a 40% decline in Windows revenue. PC sales declined liked 35 to 40%. Crypto boomed during the pandemic. Crypto is now busted, so all those server GPU sales are declining. Services (Google, Amazon, MS, etc) boomed during the pandemic - work from home services etc - and now it has busted too.

    Apple will be lucky to hold the line at 20% declines imo. They said Macs would decline 3 months ago. iPhone sales have slowed due to COVID lockdowns at various plants.
    muthuk_vanalingamFileMakerFellerwatto_cobran2itivguymagman1979
  • Reply 3 of 30
    My lap misses my Intel MacBook Pro, on these cold winter nights. My Apple Silicon MBP is always ice cold!
    mknelsonDAalsethhydrogenOferbadmonkpkisselTRAGbaconstanglkruppwatto_cobra
  • Reply 4 of 30
    omasouomasou Posts: 592member
    Keep milking the same cash cow (x86)...eventually it's gonna give out.
    hydrogenwilliamlondondewmewatto_cobra
  • Reply 5 of 30
    Maybe if they plays those promo videos of how Intel laptops have a touch screen and MacBooks don't... maybe that'll fix everything?
    OferbaconstangAlex_Vlkruppwatto_cobramagman1979
  • Reply 6 of 30
    omasou said:
    Keep milking the same cash cow (x86)...eventually it's gonna give out.
    They did try a new chip design … remember Itanium (aka The Itanic)?
    watto_cobra
  • Reply 7 of 30
    DAalsethDAalseth Posts: 2,806member
    omasou said:
    Keep milking the same cash cow (x86)...eventually it's gonna give out.
    They did try a new chip design … remember Itanium (aka The Itanic)?
    Oh yeah I had forgotten about Itanium. Great idea, went nowhere. 
    watto_cobra
  • Reply 8 of 30
    danoxdanox Posts: 2,976member
    tht said:
    ITGUYINSD said:
    The article doesn't really explain WHY Intel sales were so bad.  Especially during the holiday season.  Is new gear too expensive for consumers?  Are people keeping their computers longer?  Or are consumers moving towards more mobile devices?

    It doesn't help Intel that major sellers like Dell have raised their prices making one take a second thought about buying or upgrading.  
    Post-pandemic bust.

    MS reported a 40% decline in Windows revenue. PC sales declined liked 35 to 40%. Crypto boomed during the pandemic. Crypto is now busted, so all those server GPU sales are declining. Services (Google, Amazon, MS, etc) boomed during the pandemic - work from home services etc - and now it has busted too.

    Apple will be lucky to hold the line at 20% declines imo. They said Macs would decline 3 months ago. iPhone sales have slowed due to COVID lockdowns at various plants.
    Windows, and Intel, fates are related to each other, but AMD and Nvidia are in the mix. However, Apple computer isn’t sitting on any crypto farm on a rack, Intel’s problems goes all the way back selling Strong arm and to their refusal to build iPhone CPU’s, on that day their fate going into the future was sealed.

    Apple may have a decline, but it probably will be more like they couldn’t make enough of their products fast enough to sell to the public, I want a larger screen iMac, a M2 24” iMac or possibly a updated Mac Pro but can’t buy any because Apple hasn’t released those products.

    Intel’s problems right now are long term Apple (TSMC) has disrupted them. Intel long-term decline won’t happen overnight, but the era of barn burning CPUs and graphic engines is coming to an end particularly on non-gaming consumer computers (the largest market) and in parallel Intel has been a complete no-show in the mobile market which has taken over computing despite current the downturn.

    Intel will hang on in the back room for a while but arm servers in time, will take up a bigger and bigger chunk of that market behind the scenes, coincidentally, Intel used to own Strong arm but they sold it.
    edited January 2023 baconstangwatto_cobra
  • Reply 9 of 30
    MarvinMarvin Posts: 15,357moderator
    ITGUYINSD said:
    The article doesn't really explain WHY Intel sales were so bad.  Especially during the holiday season.  Is new gear too expensive for consumers?  Are people keeping their computers longer?  Or are consumers moving towards more mobile devices?

    It doesn't help Intel that major sellers like Dell have raised their prices making one take a second thought about buying or upgrading.  
    Their annual report goes into more detail:

    https://www.intc.com/filings-reports/all-sec-filings/content/0000050863-23-000006/0000050863-23-000006.pdf

    2022 vs. 2021
    ▪ Notebook revenue was $18.8 billion, down $6.7 billion from 2021. Notebook unit sales decreased 36%, driven by lower demand in the consumer and education market segments, and notebook ASPs increased 15% due to an increased mix of commercial and consumer products and a lower mix of education products.
    ▪ Desktop revenue was $10.7 billion, down $1.8 billion from 2021. Desktop unit sales decreased 19%, driven by lower demand in the consumer and education market segments, and desktop ASPs increased 5%, primarily from an increased mix of commercial products.
    ▪ Other revenue was $2.3 billion, down $923 million from 2021, primarily driven by the continued ramp down from the exit of our 5G smartphone modem business and lower demand for our wireless and connectivity products.

    Operating Income 2022
    (3,047) Lower gross margin from notebook revenue
    (2,183) Higher notebook and desktop unit cost primarily from increased mix of Intel 7 products
    (1,306) Lower gross margin from desktop revenue
    (1,284) Higher operating expenses driven by increased investments in leadership products
    (969) Higher period charges primarily driven by inventory reserves taken in 2022
    (320) Lower CCG other product gross margin driven by lower demand for our wireless and connectivity products and the continued ramp down from the exit of our 5G smartphone modem business

    DCAI server volume decreased, led by enterprise customers, and due to customers tempering purchases to reduce existing inventories in a softening data center market. Server ASPs decreased due to customer and product mix. NEX revenue increased primarily due to Ethernet ASPs and increased demand for 5G products, partially offset by lower demand for Network Xeon.

    Their future roadmap doesn't look too good either:

    https://www.intel.com/content/www/us/en/newsroom/news/intel-technology-roadmaps-milestones.html

    They are supposedly going from Intel 7 to Intel 4, Intel 3, Intel 20A, Intel 18A in 2 years. Combined they are aiming for 1.8x improvement in performance-per-watt. That only brings them to where Apple is now if they deliver.

    "Intel 7 is in production and shipping in volume with the launch of 12th Gen Intel® Core™ processors and additional products coming in 2022. Intel 4, our implementation of extreme ultraviolet (EUV) lithography, will be manufacturing-ready in the second half of 2022. It delivers an approximate 20% increase in transistor performance per watt. Intel 3, with additional features, delivers a further 18% performance per watt and will be manufacturing-ready in the second half of 2023. Ushering in the Angstrom Era with RibbonFET and PowerVia, Intel 20A will deliver up to a 15% performance per watt improvement and will be manufacturing-ready in the first half of 2024. Intel 18A delivers an additional 10% improvement and will be manufacturing-ready in the second half of 2024."

    Looks like they are going to burn a lot of capital to try playing catchup and it will be 2 years before they start to be competitive. If they can eventually get to a point where they can manufacture chips for Apple, that would be good but they are so far behind right now.
    Appleish said:
    My lap misses my Intel MacBook Pro, on these cold winter nights. My Apple Silicon MBP is always ice cold!
    USB leg warmer:

    https://www.amazon.com/Comfheat-Portable-Temperature-Settings-Compress/dp/B098QPD82T
    FileMakerFellernubuswatto_cobramagman1979
  • Reply 10 of 30
    MacProMacPro Posts: 19,758member
    Multi-year graphs should show a better picture.  It seems silly to look at drops in sales after an artificial boom, such as that Covid created,  in isolation.
    dewmeDAalsethFileMakerFellerbaconstangwatto_cobra
  • Reply 11 of 30
    tbornottbornot Posts: 116member
    It’s also likely that Apple sensed that Intel was a subsidiary of Microsoft, and their needs came first. “What do we have to lose?”,Apple thought as they took the M1 gamble. 
    Alex_Vwatto_cobra
  • Reply 12 of 30
    blastdoorblastdoor Posts: 3,346member
    Intel's roadmap is impressive, but anybody can draw a map. Fact is, they're still on what they used to call 10nm -- a process that was first kinda-sorta introduced in 2018. They refined it a bit and renamed it "Intel 7", but they haven't made advanced a full node in over 4 years. 

    So.... let's see when we get actual products shipping in volume on "intel 4" (formerly called 7nm, which was supposed to come out several years ago), and how that stacks up to TSMC's best. 

    thtFileMakerFellerbaconstangwatto_cobra
  • Reply 13 of 30
    All right folks. Time to get out of the kiddie pool delusions and join the adults in reality.
    1. Intel's decline in 2022 are due to the recession and the stuff going on with China, Russia and Ukraine. We know this because it impacted iPhone sales too, as their sales were down 15% year-over-year.

    2. Even with one of the worst years in history, 286 million PCs sold in 2022. Even with one of their best years in history, Apple accounted for only 23 to 28 million of those depending on who you believe. OF the remaining 260 million or so PCs that did sell, Intel had 70% of those to AMD's 30%. Intel projects that this is going to rebound to 300 million next year. Analysts are railing (from their MacBooks and iPhones) that Intel is crazy to project this, but the 3 year replacement cycle for devices that were bought at the start of the pandemic will end in 2Q 2023. The enterprise "refresh the hardware we bought during the pandemic" will last through 2024.

    3. Intel did take a big hit in servers ... but from AMD, not ARM. Despite all the hype that you have heard - 99% of it typed on MacBooks and iPad Pros since November 2020 - ARM has only 3% of the server business. Marvell left the ARM server business because they weren't making any money, leaving Ampere as the last major player there. (And Ampere's "major" is "fewer than 1000 global employees.") Granted, Nvidia will join Ampere in a few months. But, AMD released their response to ARM servers in 2022. Intel's response won't come until 2024 but it will be mighty substantial: Sierra Forest, a Xeon with 334 efficiency cores on a 6nm process. It removes the sole advantage that ARM servers have over x86, which is better power efficiency per thread. 

    4. Apple Silicon fans do their best to evade this, but even Apple acknowledges that Intel has superior single core performance. The M1 Ultra scored 1780. The M2 Pro? 1952. Fine but the last gen Intel Core i9-12900K (1986) as well as the current gen Core i9-13900KS (2286), Core i7-13700K (2107), and Core i5-13600KF (2011) all clearly beat it. You should know that the latter is in a budget gaming desktop (16 GB RAM, 1 TB SSD, NVIDIA GeForce RTX 3060) that costs $1200. Yes, the 3nm M3 will come out this year. No, it will not beat the 10nm Core i9-13900KS. It won't beat the 14th gen Core i7 will be out by the end of the year either.

    5. This makes the "Combined they are aiming for 1.8x improvement in performance-per-watt. That only brings them to where Apple is now if they deliver" wishful thinking. Intel's 14nm chips were already beating the M2 Pro. This means that when Intel reaches 3nm, they will be able to make 28W chips (the M2 Pro Mac Mini's TDP is around 26.5 W) that have the same performance as what Apple offers as well as 75W-125W chips that will crush it. How? Simple: where Apple currently has only 1 performance core design (used for smartphones/tablets, laptops and PCs) Intel has 5: Core i3, Core i5, Core i7, Core i9/Xeon. (This is down from 7, as Intel ditched Celeron and Pentium.) Core i7 and Core i9 are bigger, meaning better performance but worse efficiency. Core i5 and Core i3? Smaller cores for better efficiency but worse performance. As Intel shrinks the die from 10nm to 3nm, they will keep the 125W base for Core i9 and Core i7 ... but reduce it for Core i5 and especially Core i3. The Intel core i3-13100F, for example, has unofficial single core scores of about 1700, or 95% of the M1 Ultra. They only need to maintain that while significantly decreasing the TDP for certain Core i3 and Core i5 versions. Granted, this will occur in the laptop versions of the chips - for desktops their competition will be AMD's Ryzen 5 and Ryzen 7, not Apple - but that doesn't matter because mini-PCs like the Intel NUC Extreme that are going to compete with the Mac Mini and Mac Studio will use laptop chips anyway. Wherever Apple is going to be in 2025 with 2nd gen TSMC N3E, Intel will be able to come up with a (likely) Core i5 to match it in 2026.  

    So long as Intel is inside 200 million Windows and Linux PCs sold each year, they aren't going anywhere. 
    williamlondondewmeFileMakerFellernubus
  • Reply 14 of 30
    blastdoorblastdoor Posts: 3,346member
    thadec said:
    All right folks. Time to get out of the kiddie pool delusions and join the adults in reality.
    1. Intel's decline in 2022 are due to the recession and the stuff going on with China, Russia and Ukraine. We know this because it impacted iPhone sales too, as their sales were down 15% year-over-year.

    2. Even with one of the worst years in history, 286 million PCs sold in 2022. Even with one of their best years in history, Apple accounted for only 23 to 28 million of those depending on who you believe. OF the remaining 260 million or so PCs that did sell, Intel had 70% of those to AMD's 30%. Intel projects that this is going to rebound to 300 million next year. Analysts are railing (from their MacBooks and iPhones) that Intel is crazy to project this, but the 3 year replacement cycle for devices that were bought at the start of the pandemic will end in 2Q 2023. The enterprise "refresh the hardware we bought during the pandemic" will last through 2024.

    3. Intel did take a big hit in servers ... but from AMD, not ARM. Despite all the hype that you have heard - 99% of it typed on MacBooks and iPad Pros since November 2020 - ARM has only 3% of the server business. Marvell left the ARM server business because they weren't making any money, leaving Ampere as the last major player there. (And Ampere's "major" is "fewer than 1000 global employees.") Granted, Nvidia will join Ampere in a few months. But, AMD released their response to ARM servers in 2022. Intel's response won't come until 2024 but it will be mighty substantial: Sierra Forest, a Xeon with 334 efficiency cores on a 6nm process. It removes the sole advantage that ARM servers have over x86, which is better power efficiency per thread. 

    4. Apple Silicon fans do their best to evade this, but even Apple acknowledges that Intel has superior single core performance. The M1 Ultra scored 1780. The M2 Pro? 1952. Fine but the last gen Intel Core i9-12900K (1986) as well as the current gen Core i9-13900KS (2286), Core i7-13700K (2107), and Core i5-13600KF (2011) all clearly beat it. You should know that the latter is in a budget gaming desktop (16 GB RAM, 1 TB SSD, NVIDIA GeForce RTX 3060) that costs $1200. Yes, the 3nm M3 will come out this year. No, it will not beat the 10nm Core i9-13900KS. It won't beat the 14th gen Core i7 will be out by the end of the year either.

    5. This makes the "Combined they are aiming for 1.8x improvement in performance-per-watt. That only brings them to where Apple is now if they deliver" wishful thinking. Intel's 14nm chips were already beating the M2 Pro. This means that when Intel reaches 3nm, they will be able to make 28W chips (the M2 Pro Mac Mini's TDP is around 26.5 W) that have the same performance as what Apple offers as well as 75W-125W chips that will crush it. How? Simple: where Apple currently has only 1 performance core design (used for smartphones/tablets, laptops and PCs) Intel has 5: Core i3, Core i5, Core i7, Core i9/Xeon. (This is down from 7, as Intel ditched Celeron and Pentium.) Core i7 and Core i9 are bigger, meaning better performance but worse efficiency. Core i5 and Core i3? Smaller cores for better efficiency but worse performance. As Intel shrinks the die from 10nm to 3nm, they will keep the 125W base for Core i9 and Core i7 ... but reduce it for Core i5 and especially Core i3. The Intel core i3-13100F, for example, has unofficial single core scores of about 1700, or 95% of the M1 Ultra. They only need to maintain that while significantly decreasing the TDP for certain Core i3 and Core i5 versions. Granted, this will occur in the laptop versions of the chips - for desktops their competition will be AMD's Ryzen 5 and Ryzen 7, not Apple - but that doesn't matter because mini-PCs like the Intel NUC Extreme that are going to compete with the Mac Mini and Mac Studio will use laptop chips anyway. Wherever Apple is going to be in 2025 with 2nd gen TSMC N3E, Intel will be able to come up with a (likely) Core i5 to match it in 2026.  

    So long as Intel is inside 200 million Windows and Linux PCs sold each year, they aren't going anywhere. 
    It's as if somebody asked ChatGPT to imitate an Intel troll. 

    Here's my best attempt at a pro-Intel perspective. 

    Since Gelsinger's return, Intel has done a fantastic job of turning their lemons into lemonade. Lots of customers care about benchmarks, Apple has committed to keeping power and heat low, so Intel realized they can win single core benchmarks by overclocking one big core for brief periods of time. That doesn't lead to the best product or user experience, but it's good marketing. It gives OEMs something to say to keep people from going to Apple. 

    Besides running one core super hot to eek out a single thread benchmark win, Intel realized that they can compete on multi-threaded benchmarks if they throw in a large number of smaller cores running at lower clocks. And so we see products like the 13900K with 8 'performance' cores but 16 'efficiency' cores. Can software take advantage of all those little cores? Some can, for sure, but more importantly -- another benchmark win. 

    In theory, Apple could respond in kind -- they could run one of their big cores super hot for brief bursts to recapture the single core crown. They could also stuff in more of their far superior efficiency cores to win on multicore. But Apple's not doing that, leaving Intel some space for a 'win.' I suspect Apple will leave Intel with that space for quite a while because from Apple's point of view, creating a chip to beat Intel at their own game wouldn't actually lead to better Macs. 

    So in some sense, everybody wins. Intel is less profitable, but still in business and surviving to hopefully get their manufacturing back on track. Intel's customers 'win' in that they get cheap PCs that look good in benchmarks. Apple's customers win because they get computers that have excellent real world performance with low heat and great battery life. Apple wins because the customers that appreciate their products also happen to be highly educated and affluent. Happy happy joy joy. 
    FileMakerFellerbaconstangAlex_Vwatto_cobrarezwitsmagman1979
  • Reply 15 of 30
    DAalsethDAalseth Posts: 2,806member
    JP234 said:
    tbornot said:
    It’s also likely that Apple sensed that Intel was a subsidiary of Microsoft, and their needs came first. “What do we have to lose?”,Apple thought as they took the M1 gamble. 
    Actually, at one point, Apple was run by the CEO of Intel, Andy Grove. And in his 500 days at the helm, he almost ran it aground.
    Um, I think you might mean Gil Amilio who came over from National Semiconductor. I just checked both the list of Apple CEOs and Andy Grove’s biography and there’s no record of him having an association with Apple. 
    FileMakerFellerthtnubuswatto_cobrarezwitsmagman1979
  • Reply 16 of 30
    dewmedewme Posts: 5,422member
    From a long term and big-picture perspective, my gut level feeling is that the health of the US economy that is powered by semiconductor underpinnings really needs Intel to be successful. They have stumbled and bled but they have learned from their missteps. I for one will always be pulling for Intel's success regardless of whether they are designing or simply sourcing chips for my favorite computing products. They could be our Plan B if the globalization thing blows up in our faces or suffers a serious hiccup.

    That said, it is a good thing that Apple, AMD, and ARM are constantly reminding Intel that they need to raise their game. I just hope they are fast learners.
    FileMakerFellerDAalsethbaconstangmuthuk_vanalingamwatto_cobra
  • Reply 17 of 30
    MarvinMarvin Posts: 15,357moderator
    thadec said:
    2. Even with one of the worst years in history, 286 million PCs sold in 2022. Even with one of their best years in history, Apple accounted for only 23 to 28 million of those depending on who you believe. OF the remaining 260 million or so PCs that did sell, Intel had 70% of those to AMD's 30%. Intel projects that this is going to rebound to 300 million next year. Analysts are railing (from their MacBooks and iPhones) that Intel is crazy to project this, but the 3 year replacement cycle for devices that were bought at the start of the pandemic will end in 2Q 2023. The enterprise "refresh the hardware we bought during the pandemic" will last through 2024.

    4. Apple Silicon fans do their best to evade this, but even Apple acknowledges that Intel has superior single core performance. The M1 Ultra scored 1780. The M2 Pro? 1952. Fine but the last gen Intel Core i9-12900K (1986) as well as the current gen Core i9-13900KS (2286), Core i7-13700K (2107), and Core i5-13600KF (2011) all clearly beat it. You should know that the latter is in a budget gaming desktop (16 GB RAM, 1 TB SSD, NVIDIA GeForce RTX 3060) that costs $1200. Yes, the 3nm M3 will come out this year. No, it will not beat the 10nm Core i9-13900KS. It won't beat the 14th gen Core i7 will be out by the end of the year either.

    5. This makes the "Combined they are aiming for 1.8x improvement in performance-per-watt. That only brings them to where Apple is now if they deliver" wishful thinking. Intel's 14nm chips were already beating the M2 Pro. This means that when Intel reaches 3nm, they will be able to make 28W chips (the M2 Pro Mac Mini's TDP is around 26.5 W) that have the same performance as what Apple offers as well as 75W-125W chips that will crush it. How? Simple: where Apple currently has only 1 performance core design (used for smartphones/tablets, laptops and PCs) Intel has 5: Core i3, Core i5, Core i7, Core i9/Xeon. (This is down from 7, as Intel ditched Celeron and Pentium.) Core i7 and Core i9 are bigger, meaning better performance but worse efficiency. Core i5 and Core i3? Smaller cores for better efficiency but worse performance. As Intel shrinks the die from 10nm to 3nm, they will keep the 125W base for Core i9 and Core i7 ... but reduce it for Core i5 and especially Core i3. The Intel core i3-13100F, for example, has unofficial single core scores of about 1700, or 95% of the M1 Ultra. They only need to maintain that while significantly decreasing the TDP for certain Core i3 and Core i5 versions. Granted, this will occur in the laptop versions of the chips - for desktops their competition will be AMD's Ryzen 5 and Ryzen 7, not Apple - but that doesn't matter because mini-PCs like the Intel NUC Extreme that are going to compete with the Mac Mini and Mac Studio will use laptop chips anyway. Wherever Apple is going to be in 2025 with 2nd gen TSMC N3E, Intel will be able to come up with a (likely) Core i5 to match it in 2026.  

    So long as Intel is inside 200 million Windows and Linux PCs sold each year, they aren't going anywhere. 
    The i9-13900KS is 150-250W boosting to 6GHz, the fact its single-core is only a fraction higher than M2 Pro (30W CPU) is shockingly bad performance-per-watt.

    This was always the strength of Apple Silicon, they had 5x better performance-per-watt than Intel when they switched. Fractionally better performance numbers are meaningless if the computer sounds like a hairdryer and lasts 60 minutes on battery.


    Wherever Apple is going to be in 2025 with 2nd gen TSMC N3E, Intel will be able to come up with a (likely) Core i5 to match it in 2026.
    Right, Intel is trying to catch up in 3 years, we're talking about why they are underperforming now. They are currently significantly behind Apple and AMD.

    Intel marketed improvements for decades, it was always a lie, especially in integrated graphics. Apple's chips aren't just CPUs, they have best-in-class integrated graphics.

    Until Intel can deliver BOTH a CPU and integrated GPU that rivals Apple on performance-per-watt, they have nothing worth buying, which means longer upgrade cycles. Apple makes more revenue from Macs than Intel from PC chip sales with 1/10th the volume for good reason.
    Alex_Vhydrogenwilliamlondonwatto_cobra
  • Reply 18 of 30
    blastdoorblastdoor Posts: 3,346member
    JP234 said:
    blastdoor said:
    thadec said:
    All right folks. Time to get out of the kiddie pool delusions and join the adults in reality.
    1. Intel's decline in 2022 are due to the recession and the stuff going on with China, Russia and Ukraine. We know this because it impacted iPhone sales too, as their sales were down 15% year-over-year.

    2. Even with one of the worst years in history, 286 million PCs sold in 2022. Even with one of their best years in history, Apple accounted for only 23 to 28 million of those depending on who you believe. OF the remaining 260 million or so PCs that did sell, Intel had 70% of those to AMD's 30%. Intel projects that this is going to rebound to 300 million next year. Analysts are railing (from their MacBooks and iPhones) that Intel is crazy to project this, but the 3 year replacement cycle for devices that were bought at the start of the pandemic will end in 2Q 2023. The enterprise "refresh the hardware we bought during the pandemic" will last through 2024.

    3. Intel did take a big hit in servers ... but from AMD, not ARM. Despite all the hype that you have heard - 99% of it typed on MacBooks and iPad Pros since November 2020 - ARM has only 3% of the server business. Marvell left the ARM server business because they weren't making any money, leaving Ampere as the last major player there. (And Ampere's "major" is "fewer than 1000 global employees.") Granted, Nvidia will join Ampere in a few months. But, AMD released their response to ARM servers in 2022. Intel's response won't come until 2024 but it will be mighty substantial: Sierra Forest, a Xeon with 334 efficiency cores on a 6nm process. It removes the sole advantage that ARM servers have over x86, which is better power efficiency per thread. 

    4. Apple Silicon fans do their best to evade this, but even Apple acknowledges that Intel has superior single core performance. The M1 Ultra scored 1780. The M2 Pro? 1952. Fine but the last gen Intel Core i9-12900K (1986) as well as the current gen Core i9-13900KS (2286), Core i7-13700K (2107), and Core i5-13600KF (2011) all clearly beat it. You should know that the latter is in a budget gaming desktop (16 GB RAM, 1 TB SSD, NVIDIA GeForce RTX 3060) that costs $1200. Yes, the 3nm M3 will come out this year. No, it will not beat the 10nm Core i9-13900KS. It won't beat the 14th gen Core i7 will be out by the end of the year either.

    5. This makes the "Combined they are aiming for 1.8x improvement in performance-per-watt. That only brings them to where Apple is now if they deliver" wishful thinking. Intel's 14nm chips were already beating the M2 Pro. This means that when Intel reaches 3nm, they will be able to make 28W chips (the M2 Pro Mac Mini's TDP is around 26.5 W) that have the same performance as what Apple offers as well as 75W-125W chips that will crush it. How? Simple: where Apple currently has only 1 performance core design (used for smartphones/tablets, laptops and PCs) Intel has 5: Core i3, Core i5, Core i7, Core i9/Xeon. (This is down from 7, as Intel ditched Celeron and Pentium.) Core i7 and Core i9 are bigger, meaning better performance but worse efficiency. Core i5 and Core i3? Smaller cores for better efficiency but worse performance. As Intel shrinks the die from 10nm to 3nm, they will keep the 125W base for Core i9 and Core i7 ... but reduce it for Core i5 and especially Core i3. The Intel core i3-13100F, for example, has unofficial single core scores of about 1700, or 95% of the M1 Ultra. They only need to maintain that while significantly decreasing the TDP for certain Core i3 and Core i5 versions. Granted, this will occur in the laptop versions of the chips - for desktops their competition will be AMD's Ryzen 5 and Ryzen 7, not Apple - but that doesn't matter because mini-PCs like the Intel NUC Extreme that are going to compete with the Mac Mini and Mac Studio will use laptop chips anyway. Wherever Apple is going to be in 2025 with 2nd gen TSMC N3E, Intel will be able to come up with a (likely) Core i5 to match it in 2026.  

    So long as Intel is inside 200 million Windows and Linux PCs sold each year, they aren't going anywhere. 
    It's as if somebody asked ChatGPT to imitate an Intel troll. 

    Here's my best attempt at a pro-Intel perspective. 

    Since Gelsinger's return, Intel has done a fantastic job of turning their lemons into lemonade. Lots of customers care about benchmarks, Apple has committed to keeping power and heat low, so Intel realized they can win single core benchmarks by overclocking one big core for brief periods of time. That doesn't lead to the best product or user experience, but it's good marketing. It gives OEMs something to say to keep people from going to Apple. 

    Besides running one core super hot to eek out a single thread benchmark win, Intel realized that they can compete on multi-threaded benchmarks if they throw in a large number of smaller cores running at lower clocks. And so we see products like the 13900K with 8 'performance' cores but 16 'efficiency' cores. Can software take advantage of all those little cores? Some can, for sure, but more importantly -- another benchmark win. 

    In theory, Apple could respond in kind -- they could run one of their big cores super hot for brief bursts to recapture the single core crown. They could also stuff in more of their far superior efficiency cores to win on multicore. But Apple's not doing that, leaving Intel some space for a 'win.' I suspect Apple will leave Intel with that space for quite a while because from Apple's point of view, creating a chip to beat Intel at their own game wouldn't actually lead to better Macs. 

    So in some sense, everybody wins. Intel is less profitable, but still in business and surviving to hopefully get their manufacturing back on track. Intel's customers 'win' in that they get cheap PCs that look good in benchmarks. Apple's customers win because they get computers that have excellent real world performance with low heat and great battery life. Apple wins because the customers that appreciate their products also happen to be highly educated and affluent. Happy happy joy joy. 
    Both of you ignore the obvious. Intel is a company in terminal decline. You're right that won't belly up, but only because it's "too big to fail." Kind of like Sears. All it's going to take is an activist CEO to break it up and sell it off for parts.
    I don’t think they are necessarily in terminal decline. I think under previous management they were living a lifestyle that was killing them. It was as if they were eating too much and not exercising — focused only on short run pleasure not long run health. They were kind of like Rocky at the start of Rocky 3, but worse (imagine Rocky 50 pounds over weight eating pizza and drinking beer all the time). 

    I think Gelsinger is trying to get this fighter back in shape and restore the eye of the tiger. There’s no guarantee he will succeed, but I think it’s at least possible that he might.

    I hope he does because it would mean we get better manufacturing nodes faster. And apple could use Intel to fab apple silicon. 
    muthuk_vanalingamwatto_cobra
  • Reply 19 of 30
    danoxdanox Posts: 2,976member
    blastdoor said:
    JP234 said:
    blastdoor said:
    thadec said:
    All right folks. Time to get out of the kiddie pool delusions and join the adults in reality.
    1. Intel's decline in 2022 are due to the recession and the stuff going on with China, Russia and Ukraine. We know this because it impacted iPhone sales too, as their sales were down 15% year-over-year.

    2. Even with one of the worst years in history, 286 million PCs sold in 2022. Even with one of their best years in history, Apple accounted for only 23 to 28 million of those depending on who you believe. OF the remaining 260 million or so PCs that did sell, Intel had 70% of those to AMD's 30%. Intel projects that this is going to rebound to 300 million next year. Analysts are railing (from their MacBooks and iPhones) that Intel is crazy to project this, but the 3 year replacement cycle for devices that were bought at the start of the pandemic will end in 2Q 2023. The enterprise "refresh the hardware we bought during the pandemic" will last through 2024.

    3. Intel did take a big hit in servers ... but from AMD, not ARM. Despite all the hype that you have heard - 99% of it typed on MacBooks and iPad Pros since November 2020 - ARM has only 3% of the server business. Marvell left the ARM server business because they weren't making any money, leaving Ampere as the last major player there. (And Ampere's "major" is "fewer than 1000 global employees.") Granted, Nvidia will join Ampere in a few months. But, AMD released their response to ARM servers in 2022. Intel's response won't come until 2024 but it will be mighty substantial: Sierra Forest, a Xeon with 334 efficiency cores on a 6nm process. It removes the sole advantage that ARM servers have over x86, which is better power efficiency per thread. 

    4. Apple Silicon fans do their best to evade this, but even Apple acknowledges that Intel has superior single core performance. The M1 Ultra scored 1780. The M2 Pro? 1952. Fine but the last gen Intel Core i9-12900K (1986) as well as the current gen Core i9-13900KS (2286), Core i7-13700K (2107), and Core i5-13600KF (2011) all clearly beat it. You should know that the latter is in a budget gaming desktop (16 GB RAM, 1 TB SSD, NVIDIA GeForce RTX 3060) that costs $1200. Yes, the 3nm M3 will come out this year. No, it will not beat the 10nm Core i9-13900KS. It won't beat the 14th gen Core i7 will be out by the end of the year either.

    5. This makes the "Combined they are aiming for 1.8x improvement in performance-per-watt. That only brings them to where Apple is now if they deliver" wishful thinking. Intel's 14nm chips were already beating the M2 Pro. This means that when Intel reaches 3nm, they will be able to make 28W chips (the M2 Pro Mac Mini's TDP is around 26.5 W) that have the same performance as what Apple offers as well as 75W-125W chips that will crush it. How? Simple: where Apple currently has only 1 performance core design (used for smartphones/tablets, laptops and PCs) Intel has 5: Core i3, Core i5, Core i7, Core i9/Xeon. (This is down from 7, as Intel ditched Celeron and Pentium.) Core i7 and Core i9 are bigger, meaning better performance but worse efficiency. Core i5 and Core i3? Smaller cores for better efficiency but worse performance. As Intel shrinks the die from 10nm to 3nm, they will keep the 125W base for Core i9 and Core i7 ... but reduce it for Core i5 and especially Core i3. The Intel core i3-13100F, for example, has unofficial single core scores of about 1700, or 95% of the M1 Ultra. They only need to maintain that while significantly decreasing the TDP for certain Core i3 and Core i5 versions. Granted, this will occur in the laptop versions of the chips - for desktops their competition will be AMD's Ryzen 5 and Ryzen 7, not Apple - but that doesn't matter because mini-PCs like the Intel NUC Extreme that are going to compete with the Mac Mini and Mac Studio will use laptop chips anyway. Wherever Apple is going to be in 2025 with 2nd gen TSMC N3E, Intel will be able to come up with a (likely) Core i5 to match it in 2026.  

    So long as Intel is inside 200 million Windows and Linux PCs sold each year, they aren't going anywhere. 
    It's as if somebody asked ChatGPT to imitate an Intel troll. 

    Here's my best attempt at a pro-Intel perspective. 

    Since Gelsinger's return, Intel has done a fantastic job of turning their lemons into lemonade. Lots of customers care about benchmarks, Apple has committed to keeping power and heat low, so Intel realized they can win single core benchmarks by overclocking one big core for brief periods of time. That doesn't lead to the best product or user experience, but it's good marketing. It gives OEMs something to say to keep people from going to Apple. 

    Besides running one core super hot to eek out a single thread benchmark win, Intel realized that they can compete on multi-threaded benchmarks if they throw in a large number of smaller cores running at lower clocks. And so we see products like the 13900K with 8 'performance' cores but 16 'efficiency' cores. Can software take advantage of all those little cores? Some can, for sure, but more importantly -- another benchmark win. 

    In theory, Apple could respond in kind -- they could run one of their big cores super hot for brief bursts to recapture the single core crown. They could also stuff in more of their far superior efficiency cores to win on multicore. But Apple's not doing that, leaving Intel some space for a 'win.' I suspect Apple will leave Intel with that space for quite a while because from Apple's point of view, creating a chip to beat Intel at their own game wouldn't actually lead to better Macs. 

    So in some sense, everybody wins. Intel is less profitable, but still in business and surviving to hopefully get their manufacturing back on track. Intel's customers 'win' in that they get cheap PCs that look good in benchmarks. Apple's customers win because they get computers that have excellent real world performance with low heat and great battery life. Apple wins because the customers that appreciate their products also happen to be highly educated and affluent. Happy happy joy joy. 
    Both of you ignore the obvious. Intel is a company in terminal decline. You're right that won't belly up, but only because it's "too big to fail." Kind of like Sears. All it's going to take is an activist CEO to break it up and sell it off for parts.
    I don’t think they are necessarily in terminal decline. I think under previous management they were living a lifestyle that was killing them. It was as if they were eating too much and not exercising — focused only on short run pleasure not long run health. They were kind of like Rocky at the start of Rocky 3, but worse (imagine Rocky 50 pounds over weight eating pizza and drinking beer all the time). 

    I think Gelsinger is trying to get this fighter back in shape and restore the eye of the tiger. There’s no guarantee he will succeed, but I think it’s at least possible that he might.

    I hope he does because it would mean we get better manufacturing nodes faster. And apple could use Intel to fab apple silicon. 

    Apple has already lived through the Three Stooges, Motorola, IBM and Intel. They don’t need to live through a fourth on a rebound.
    edited January 2023 watto_cobra
  • Reply 20 of 30
    ;) danox said:
    blastdoor said:
    JP234 said:
    blastdoor said:
    thadec said:
    All right folks. Time to get out of the kiddie pool delusions and join the adults in reality.
    1. Intel's decline in 2022 are due to the recession and the stuff going on with China, Russia and Ukraine. We know this because it impacted iPhone sales too, as their sales were down 15% year-over-year.

    2. Even with one of the worst years in history, 286 million PCs sold in 2022. Even with one of their best years in history, Apple accounted for only 23 to 28 million of those depending on who you believe. OF the remaining 260 million or so PCs that did sell, Intel had 70% of those to AMD's 30%. Intel projects that this is going to rebound to 300 million next year. Analysts are railing (from their MacBooks and iPhones) that Intel is crazy to project this, but the 3 year replacement cycle for devices that were bought at the start of the pandemic will end in 2Q 2023. The enterprise "refresh the hardware we bought during the pandemic" will last through 2024.

    3. Intel did take a big hit in servers ... but from AMD, not ARM. Despite all the hype that you have heard - 99% of it typed on MacBooks and iPad Pros since November 2020 - ARM has only 3% of the server business. Marvell left the ARM server business because they weren't making any money, leaving Ampere as the last major player there. (And Ampere's "major" is "fewer than 1000 global employees.") Granted, Nvidia will join Ampere in a few months. But, AMD released their response to ARM servers in 2022. Intel's response won't come until 2024 but it will be mighty substantial: Sierra Forest, a Xeon with 334 efficiency cores on a 6nm process. It removes the sole advantage that ARM servers have over x86, which is better power efficiency per thread. 

    4. Apple Silicon fans do their best to evade this, but even Apple acknowledges that Intel has superior single core performance. The M1 Ultra scored 1780. The M2 Pro? 1952. Fine but the last gen Intel Core i9-12900K (1986) as well as the current gen Core i9-13900KS (2286), Core i7-13700K (2107), and Core i5-13600KF (2011) all clearly beat it. You should know that the latter is in a budget gaming desktop (16 GB RAM, 1 TB SSD, NVIDIA GeForce RTX 3060) that costs $1200. Yes, the 3nm M3 will come out this year. No, it will not beat the 10nm Core i9-13900KS. It won't beat the 14th gen Core i7 will be out by the end of the year either.

    5. This makes the "Combined they are aiming for 1.8x improvement in performance-per-watt. That only brings them to where Apple is now if they deliver" wishful thinking. Intel's 14nm chips were already beating the M2 Pro. This means that when Intel reaches 3nm, they will be able to make 28W chips (the M2 Pro Mac Mini's TDP is around 26.5 W) that have the same performance as what Apple offers as well as 75W-125W chips that will crush it. How? Simple: where Apple currently has only 1 performance core design (used for smartphones/tablets, laptops and PCs) Intel has 5: Core i3, Core i5, Core i7, Core i9/Xeon. (This is down from 7, as Intel ditched Celeron and Pentium.) Core i7 and Core i9 are bigger, meaning better performance but worse efficiency. Core i5 and Core i3? Smaller cores for better efficiency but worse performance. As Intel shrinks the die from 10nm to 3nm, they will keep the 125W base for Core i9 and Core i7 ... but reduce it for Core i5 and especially Core i3. The Intel core i3-13100F, for example, has unofficial single core scores of about 1700, or 95% of the M1 Ultra. They only need to maintain that while significantly decreasing the TDP for certain Core i3 and Core i5 versions. Granted, this will occur in the laptop versions of the chips - for desktops their competition will be AMD's Ryzen 5 and Ryzen 7, not Apple - but that doesn't matter because mini-PCs like the Intel NUC Extreme that are going to compete with the Mac Mini and Mac Studio will use laptop chips anyway. Wherever Apple is going to be in 2025 with 2nd gen TSMC N3E, Intel will be able to come up with a (likely) Core i5 to match it in 2026.  

    So long as Intel is inside 200 million Windows and Linux PCs sold each year, they aren't going anywhere. 
    It's as if somebody asked ChatGPT to imitate an Intel troll. 

    Here's my best attempt at a pro-Intel perspective. 

    Since Gelsinger's return, Intel has done a fantastic job of turning their lemons into lemonade. Lots of customers care about benchmarks, Apple has committed to keeping power and heat low, so Intel realized they can win single core benchmarks by overclocking one big core for brief periods of time. That doesn't lead to the best product or user experience, but it's good marketing. It gives OEMs something to say to keep people from going to Apple. 

    Besides running one core super hot to eek out a single thread benchmark win, Intel realized that they can compete on multi-threaded benchmarks if they throw in a large number of smaller cores running at lower clocks. And so we see products like the 13900K with 8 'performance' cores but 16 'efficiency' cores. Can software take advantage of all those little cores? Some can, for sure, but more importantly -- another benchmark win. 

    In theory, Apple could respond in kind -- they could run one of their big cores super hot for brief bursts to recapture the single core crown. They could also stuff in more of their far superior efficiency cores to win on multicore. But Apple's not doing that, leaving Intel some space for a 'win.' I suspect Apple will leave Intel with that space for quite a while because from Apple's point of view, creating a chip to beat Intel at their own game wouldn't actually lead to better Macs. 

    So in some sense, everybody wins. Intel is less profitable, but still in business and surviving to hopefully get their manufacturing back on track. Intel's customers 'win' in that they get cheap PCs that look good in benchmarks. Apple's customers win because they get computers that have excellent real world performance with low heat and great battery life. Apple wins because the customers that appreciate their products also happen to be highly educated and affluent. Happy happy joy joy. 
    Both of you ignore the obvious. Intel is a company in terminal decline. You're right that won't belly up, but only because it's "too big to fail." Kind of like Sears. All it's going to take is an activist CEO to break it up and sell it off for parts.
    I don’t think they are necessarily in terminal decline. I think under previous management they were living a lifestyle that was killing them. It was as if they were eating too much and not exercising — focused only on short run pleasure not long run health. They were kind of like Rocky at the start of Rocky 3, but worse (imagine Rocky 50 pounds over weight eating pizza and drinking beer all the time). 

    I think Gelsinger is trying to get this fighter back in shape and restore the eye of the tiger. There’s no guarantee he will succeed, but I think it’s at least possible that he might.

    I hope he does because it would mean we get better manufacturing nodes faster. And apple could use Intel to fab apple silicon. 

    Apple has already lived through the Three Stooges, Motorola, IBM and Intel. They don’t need to live through a fourth on a rebound.
    The fourth is Apple Within the rebounder!  o:)
    watto_cobra
Sign In or Register to comment.