Profile - To The Moon

Seb195

Posts · 2Posts · 2
Following · 0Following · 0
Followers · 0Followers · 0

- Seb195Seb195
  ·2021-06-09
  Wow
  Intel: AMD Threat Is Finished
  Summary Although competition from Arm is increasing, AMD remains Intel’s biggest competitor, as con
  Intel: AMD Threat Is Finished
  16 Comment
  Like
  Share
  Report
- Seb195Seb195
  ·2021-06-09
  ?
  Sorry, this post has been deleted
  4 Comment
  Like
  Share
  Report

Most Discussed

Terms and Conditions·Privacy Policy

Company: TTMF Limited. Tech supported by Xiangshang Yixin.

Email：uservice@ttm.financial

{"i18n":{"language":"en_US"},"isCurrentUser":false,"userPageInfo":{"id":"3582678388280889","uuid":"3582678388280889","gmtCreate":1619585844410,"gmtModify":1619585844410,"name":"Seb195","pinyin":"seb195","introduction":"","introductionEn":null,"signature":"","avatar":"https://static.laohu8.com/default-avatar.jpg","hat":null,"hatId":null,"hatName":null,"vip":1,"status":2,"fanSize":0,"headSize":6,"tweetSize":2,"questionSize":0,"limitLevel":999,"accountStatus":4,"level":{"id":0,"name":"","nameTw":"","represent":"","factor":"","iconColor":"","bgColor":""},"themeCounts":0,"badgeCounts":0,"badges":[],"moderator":false,"superModerator":false,"manageSymbols":null,"badgeLevel":null,"boolIsFan":false,"boolIsHead":false,"favoriteSize":0,"symbols":null,"coverImage":null,"realNameVerified":"success","userBadges":[{"badgeId":"1026c425416b44e0aac28c11a0848493-2","templateUuid":"1026c425416b44e0aac28c11a0848493","name":"Senior Tiger","description":"Join the tiger community for 1000 days","bigImgUrl":"https://static.tigerbbs.com/0063fb68ea29c9ae6858c58630e182d5","smallImgUrl":"https://static.tigerbbs.com/96c699a93be4214d4b49aea6a5a5d1a4","grayImgUrl":"https://static.tigerbbs.com/35b0e542a9ff77046ed69ef602bc105d","redirectLinkEnabled":0,"redirectLink":null,"hasAllocated":1,"isWearing":0,"stamp":null,"stampPosition":0,"hasStamp":0,"allocationCount":1,"allocatedDate":"2024.01.27","exceedPercentage":null,"individualDisplayEnabled":0,"backgroundColor":null,"fontColor":null,"individualDisplaySort":0,"categoryType":1001},{"badgeId":"7a9f168ff73447fe856ed6c938b61789-1","templateUuid":"7a9f168ff73447fe856ed6c938b61789","name":"Knowledgeable Investor","description":"Traded more than 10 stocks","bigImgUrl":"https://static.tigerbbs.com/e74cc24115c4fbae6154ec1b1041bf47","smallImgUrl":"https://static.tigerbbs.com/d48265cbfd97c57f9048db29f22227b0","grayImgUrl":"https://static.tigerbbs.com/76c6d6898b073c77e1c537ebe9ac1c57","redirectLinkEnabled":0,"redirectLink":null,"hasAllocated":1,"isWearing":0,"stamp":null,"stampPosition":0,"hasStamp":0,"allocationCount":1,"allocatedDate":"2021.12.21","exceedPercentage":null,"individualDisplayEnabled":0,"backgroundColor":null,"fontColor":null,"individualDisplaySort":0,"categoryType":1102},{"badgeId":"a83d7582f45846ffbccbce770ce65d84-1","templateUuid":"a83d7582f45846ffbccbce770ce65d84","name":"Real Trader","description":"Completed a transaction","bigImgUrl":"https://static.tigerbbs.com/2e08a1cc2087a1de93402c2c290fa65b","smallImgUrl":"https://static.tigerbbs.com/4504a6397ce1137932d56e5f4ce27166","grayImgUrl":"https://static.tigerbbs.com/4b22c79415b4cd6e3d8ebc4a0fa32604","redirectLinkEnabled":0,"redirectLink":null,"hasAllocated":1,"isWearing":0,"stamp":null,"stampPosition":0,"hasStamp":0,"allocationCount":1,"allocatedDate":"2021.12.21","exceedPercentage":null,"individualDisplayEnabled":0,"backgroundColor":null,"fontColor":null,"individualDisplaySort":0,"categoryType":1100}],"userBadgeCount":3,"currentWearingBadge":null,"individualDisplayBadges":null,"crmLevel":2,"crmLevelSwitch":0,"location":null,"starInvestorFollowerNum":0,"starInvestorFlag":false,"starInvestorOrderShareNum":0,"subscribeStarInvestorNum":0,"ror":null,"winRationPercentage":null,"showRor":false,"investmentPhilosophy":null,"starInvestorSubscribeFlag":false},"page":1,"watchlist":null,"tweetList":[{"id":180964790,"gmtCreate":1623170052710,"gmtModify":1704197675610,"author":{"id":"3582678388280889","authorId":"3582678388280889","name":"Seb195","avatar":"https://static.laohu8.com/default-avatar.jpg","crmLevel":2,"crmLevelSwitch":0,"followedFlag":false,"authorIdStr":"3582678388280889","idStr":"3582678388280889"},"themes":[],"htmlText":"Wow","listText":"Wow","text":"Wow","images":[],"top":1,"highlighted":1,"essential":1,"paper":1,"likeSize":0,"commentSize":0,"repostSize":0,"link":"https://ttm.financial/post/180964790","repostId":"1124688970","repostType":4,"repost":{"id":"1124688970","kind":"news","pubTimestamp":1623164900,"share":"https://ttm.financial/m/news/1124688970?lang=&edition=fundamental","pubTime":"2021-06-08 23:08","market":"us","language":"en","title":"Intel: AMD Threat Is Finished","url":"https://stock-news.laohu8.com/highlight/detail?id=1124688970","media":"seekingalpha","summary":"Summary\n\nAlthough competition from Arm is increasing, AMD remains Intel’s biggest competitor, as con","content":"<p><b>Summary</b></p>\n<ul>\n <li>Although competition from Arm is increasing, AMD remains Intel’s biggest competitor, as concerns of losing market share weigh on Intel’s valuation.</li>\n <li>AMD's short-lived laptop competitiveness is already waning. Intel will further crush AMD with its (up to) 16-core Alder Lake: going from half the core count, to double in one generation.</li>\n <li>Intel is also re-investing in the (high-end) desktop, could leapfrog AMD in the data center, and seems to be overtaking AMD-Xilinx for FPGA leadership.</li>\n <li>AMD is slow to transition to the leading edge in process technology. For example, AMD will not launch 5nm laptop CPUs until 2023, when Intel might have outsourced (TSMC) 3nm.</li>\n <li>Given all the above, the Intel bear thesis of AMD benefiting from Intel's stumbles, gaining a large tech advantage and taking much market share, is finally finished.</li>\n</ul>\n<p><b>Investment Thesis</b></p>\n<p>While Arm previously tried (and failed) to enter the data center about a decade ago, in recent years, there has been a more credible resurgence in Arm competition, led by the Apple(NASDAQ:AAPL)Silicon transition and Amazon (AMZN) Graviton chips. Reportedly, Microsoft (MSFT) is also responding to Amazon with its own Arm chips, and with Ampere, there is also a merchant vendor.</p>\n<p>Nevertheless, still the biggest bear thesis for Intel (INTC) is AMD (AMD), which has rolled out a competitive Zen-based product portfolio (while, at the same time, Intel was stumbling with 10nm), and hence, Intel would be at risk of losing a major amount of market share.</p>\n<p>However, as I see it, AMD’s window of opportunity, which it had due to Intel’s multi-year 10nm delays, is closing rapidly as Intel’s next-gen (much more competitive) chips are entering the market. While AMD will still be a credible alternative supplier going forward, I don’t foresee it holding a significant lead, if any lead at all. I will illustrate in several areas.</p>\n<p>That means the story of AMD profiting from Intel’s delays to take loads of market share (almost for free) is finished. It simply didn't happen when it had to.</p>\n<p>1. Waning laptop competitiveness</p>\n<p>One of the main reasons AMD has attracted much attention is because the fastest chips are, not surprisingly, found in the desktop market, where the power budget is much higher. To that end, desktop is the segment where AMD’s flagship technology debuts first.</p>\n<p>However, one just has to look at Intel’s earnings to note that the most important segment (financially) is actually the laptop one, which account for ~70% of the market. And given that AMD’s laptop chips have generally launched six months or more after the desktop segment, this means Intel has actually experienced much less pressure from AMD than many would probably have expected (at least financially in its PC business).</p>\n<p>This changed, though, with the launch of AMD Renoir in 2020. This chip packed eight 7nm Zen 2 cores. In the 15W thin-and-light segment, this meant AMD had twice as many cores as Intel, while in the 35-45W range, Intel hadn't transitioned to 10nm yet, which meant it was competing with its old 14nm Skylake-based IP. Hence, Renoir posed AMD’s first significant threat to Intel’s PC business (even more so than Zen 3, AMD’s other 2020 launch).</p>\n<p>Intel, however, has alreadyanswered Renoir with Tiger Lake:</p>\n<ul>\n <li>In the 15-28W segment, Tiger Lake has a major performance advantage in per-core and graphics performance, as well as a generally superior platform with integrated Wi-Fi 6, AI acceleration, Thunderbolt 4, etc. Tiger Lake is able to compete against 6-core Renoir chips despite having only four cores, which means, only at the very high-end (and low volume), Intel loses in multi-threaded performance.</li>\n <li>More recently, Intel hasbrought Tiger Lake to 35-45W with eight coresas well. While AMD, for its part, has transitioned to Zen 3 in laptops, benchmarks show the two CPUs are roughly equal.</li>\n</ul>\n<p>Looking forward, and it seems, AMD’s competitiveness will fade further rather quickly.</p>\n<p>Intel will launch Alder Lake in the second half of 2021. Alder Lake will implement big.Little, which has been used for many years in the smartphone space to improve power efficiency. Hence, by combining high performance and high frequency cores, Intel will be able to deliver an unmatched capability, since AMD only has one architecture. According to leaks, Alder Lake will come in 2+6 up to 6+8 configurations of Core + Atom cores (and even 8+8 for 55W laptops). Altogether, leaks have indicated Intel is expecting Alder Lake to double in performance.</p>\n<p>Hence, as I see it, Intel's hybrid designs will be a big blow for AMD's laptop competitiveness going forward. Indeed: quite recently, there have actually been somerumors of AMD developing its own hybrid designcalled Strix Point. It would consist of eight high-performance Zen 5 cores and four low power \"Zen 4D\" cores, all on 3nm.</p>\n<p>Meanwhile, looking even further out, into 2022 and beyond, AMD’s roadmap is even bleaker than Intel’s. It is highly unlikely AMD will launch a 5nm part before 2023. Given the publicity Intel received from its 7nm delay (from 2022 to 2023), that should raise concerns. While much has been discussed about Intel’s loss of process leadership, this loss only means something if its competitors make use of that advantage.</p>\n<p>In this case, AMD is not making use of it. Even though 5nm launched in the market in late 2020, AMD’s 2022 laptop roadmap consists of “Zen3+”, which means a refresh of its 2021 laptop line-up. This also means there will be no 5nm laptops from AMD until some time in 2023. That, in turn, means Intel might actually launch its 7nm CPUs before AMD launches 5nm. Never mind if Intel also charges forward with TSMC-based (TSM) 3nm CPU in 2023. The Strix Point CPU from AMD (on 3nm) is also rumored for 2024, while Intel has talked about outsourcing for 2023 already.</p>\n<p>As a last indication, Intel also took CPU connectivity leadership in laptops with PCIe 4.0, whereas AMD stayed with 3.0. This also means AMD will still be on 3.0 when Intel launches Alder Lake, which will be further upgraded to PCIe 5.0.</p>\n<p>2. Re-investing in the desktop</p>\n<p>Besides defending its laptop stronghold, Intel is also re-investing in the desktop. The desktop is one of the main victims of Intel’s 10nm delays, as Intel has yet to launch its first 10nm desktop CPUs.</p>\n<p>This will change with Alder Lake in H2’21, as Intel will bring this CPU also to the desktop. That means the desktop will (finally) return to parity with Intel’s laptop segment, in terms of technology. That should substantially improve Intel’s competitiveness. Here, likewise, leaks haveindicatedIntel is expecting 2x performance. This would put Intel on performance parity (or even a slight leadership) against Zen 3.</p>\n<p>Since recentrumors indicate that Zen 4 will launch in Q4'22, this means Intel could be more or less on parity with AMD for at least a full year (if the 2x performance claim holds true across the board).</p>\n<p>As described, though, for AMD, the desktop represents its flagship segment, whereas Intel has most vigorously defended the (much bigger) laptop space. Hence, I do not foresee Intel necessarily vigorously overtaking AMD. Still, given the seemingly late 2022 launch for Zen 4, it's a bit of pity that Meteor Lake has been delayed. Nevertheless, based on the large jump Intel is making with Alder Lake, the gap should close substantially, especially for all but those who need the highest core counts.</p>\n<p>3. Re-entering high-end desktop (HEDT)</p>\n<p>Another segment that Intel has basically ignored for the last few years is the high-end desktop. Once proud of its $1700 10-core CPU, these chips immediately became obsolete once AMD launched its Threadripper line. Even with many price cuts, Intel hasn't really had a compelling offering for this segment for years already.</p>\n<p>Reports indicate, however, that Intel is outright skipping the Ice Lake-X generation and will move straight to Sapphire Rapids-X. Since AMD lately also hasn't given its Threadripper line the most aggressive roadmap, Intel could bring some serious competition back to this market if Sapphire Rapids-X would launch in 2022.</p>\n<p>4. Overtaking AMD in the data center?</p>\n<p>Besides the desktop, the data center has been the other segment where Intel had fallen substantially behind due to its 10nm delays. Frankly, ever since AMD launched its 7nm Rome CPUs with 64 cores, it has been surprising that Intel has not lost more market share, given that its own offering consisted of 28-core CPUs on 14nm for a long time.</p>\n<p>More recently, acomprehensive benchmark effort by Phoronixhas indicated that Intel is actually surprisingly competing against these 64-core Milan CPUs with its own 40-core Ice Lake-SP on 10nm.</p>\n<p>Intel's competitiveness will further improve withSapphire Rapids. It will move to Intel’s latest technology, with the same architecture and 10nm Enhanced SuperFin process as the upcoming Alder Lake. It will (almost) close to the gap in core count, with a boost to 56 cores.</p>\n<p>In fact, in tech forums, enthusiasts continue to debate whether Sapphire Rapids will top out at 56 or 72 cores, as there have also been rumors of the latter variant. In that case, Intel's chances of unambiguously overtaking AMD would be greatly increased.</p>\n<p>Additionally, Sapphire Rapids, in any case, will take a substantial lead in I/O, with PCIe 5.0 and CXL, as well as DDR5 and HBM support. It also has an integrated data engine (Data Streaming Accelerator), and it will move to Intel’s chiplet design with four EMIB-connected tiles. This means each chiplet will have 14 (or 18) cores.</p>\n<p>Lastly, Sapphire Rapids will also substantially widen Intel’s already vast lead in AI performance, with the inclusion of Intel’s version of Nvidia’s (NVDA) Tensor Cores. In aninterview with AnandTechearlier this year, Intel said that AVX-512 (which Intel's current DLBoost is based on) is one of the largest factors for customers choosing to adopt Intel over AMD. So, to that end, Intel expects Sapphire Rapids to improve AI performance by a further 4-8x.</p>\n<p>To be sure, given the delays of at least several quarters, I do not expect Intel to take an unsurmountable leadership position. For example, in 2019, Intel said that the next-gen 7nm Granite Rapids would launch when Sapphire Rapids will actually launch: in early 2022. This means AMD will transition to 5nm before Intel transitions to its 7nm Granite Rapids CPUs, which gives AMD a chance to one-up Intel.</p>\n<p>Nevertheless, for investors, the key point is that I do not foresee that, at any point going forward, AMD will hold a substantial advantage, and for a substantial amount of time, anymore. Even with the 7nm delays, I do not foresee a repeat of the 28-core vs. 64-core situation described above.</p>\n<p>As a case in point, remember that enterprises do not care so much about who has leadership at any given time, as much as that they demand a long-term roadmap. Customers buy into roadmaps rather than single point products. Intel has such a competitive roadmap at an annual cadence: Ice Lake early 2021, Sapphire Rapids early 2022, Granite Rapids early to mid 2023, Diamond Rapids in 2014, etc.</p>\n<p>What this means is that performance will remain contested: Sapphire Rapids will likely overtake AMD, but AMD will respond with Genoa. Then, Intel will respond with Granite Rapids, etc. This raises the rhetorical question: will enterprises bother to switch to AMD if, half a year later, Intel may launch a faster CPU, etc.? The pure performance benchmarks also neglect less quantifiable advantages such as Intel’s vastly larger sales force, etc.</p>\n<p>In summary, AMD did not even manage to achieve 10% market share while it had over twice the core count (and hence a substantial leadership across the board). That advantage now seems gone for at least the next few generations. AMD simply didn't capitalize when it had the once-in-a-century opportunity.</p>\n<p>5. Challenging Xilinx for FPGA leadership</p>\n<p>I will describe FPGAs rather briefly, as this could be its own topic. As a preliminary note, one should be more cautious here since FPGAs are more esoteric technology in nature.</p>\n<p>For example, in light of AMD’s acquisition, some remarked that Intel’s Altera acquisition supposedly would be a failed one. If any arguments were given at all (to substantiate that claim), it would supposedly be because Intel has not launched an FPGA integrated with its Xeon CPUs, or because of its lackluster financial performance. However, the FPGA integration argument goes against the industry trend, which is to position the FPGA as an accelerator, just like GPUs which in the data aren't integrated directly into the CPU either. In the future, FPGAs will be connected through the open CXL interconnect, which was developed by Intel, and has also been backed by Xilinx(NASDAQ:XLNX), Arm and even AMD.</p>\n<p>Acquisition issues aside, with regards to actual FPGA leadership, here as well Intel has made much progress to catch up and even surpass Xilinx.</p>\n<p>Prior to the acquisition, Altera had delays with its 20nm generation, which led to it being one year behind Xilinx to the 16/14nm generation. However, almost literally the first day after the acquisition, Intel invested in a second, parallel design team for the 10nm generation. This allowed Intel to catch up and achieve parity to the 10nm/7nm generation, as both FPGAs started sampling around mid-2019, and have recently begun ramping more broadly.</p>\n<p>In fact, as part of the quite recentIce Lake-SP launch, Intelclaimedthat its 10nm FPGAs achieve up to 2x higher performance/watt compared to Xilinx' 7nm Versal FPGAs. So, arguably, Intel has not just got back to parity, but has in fact leapfrogged Xilinx.</p>\n<p>There are other aspects as well that demonstrate Intel’s FPGA leadership, including its pioneering use of chiplets (and in the future even 3D stacking), as well as Intel’s transceiver leadership (and indeed those transceivers are separate chiplets): Intel was first to 58G and 116G speed, and first to demo 224G in 2020.</p>\n<p>Even at 14nm, despite being later to initial launch (as described), Intel still managed to launch the first 14/16nm FPGAs with integrated HBM, integrated Arm cores and even PCIe 4.0.</p>\n<p>6. Regaining process leadership (process technology decreasing in importance)</p>\n<p>As discussed in the first point, having a process technology leadership only means something if the fabless foundry customer makes use of it. In the case of AMD, it explicitly does not, as it will launch a “Zen3+” refresh in 2022, instead of 5nm Zen 4, in laptops at least.</p>\n<p>Furthermore, combining this with the outsourcing rumors, and Intel may actually return to process leadership, as instead Intel may launch 3nm CPUs in 2023, leapfrogging AMD’s 5nm ones. In fact, it seems highly unlikely that AMD will launch any 3nm CPUs in2023 at all, as for example indicated by the Strix Point rumor for 2024.</p>\n<p>As Bob Swan said in anearly 2021 interview, it would only adopt foundries if it got preferential treatment. Hence, AMD bulls may have underestimated Intel’s position as world’s largest semiconductor company when they perhaps assumed TSMC would be dismissive of Intel’s potential multi-billion wafer orders.</p>\n<p>The larger point, though, is that Moore’s Law is likely to decrease in importance. For example, TSMC’s 3nm will deliver a real-world shrink of about 1.5x at a relatively slow 2.5 year cadence. This shows Moore’s Law is slowing down. So, even if TSMC continues to have a leadership position, it is unlikely it will have enormous advantage. Pat Gelsinger, for its part, claimed that Intel is already back on track for leadership anyway.</p>\n<p>Additionally, there are many advances beyond the base process technology, such as chiplets and even 3D stacking. If anything, Intel is actually ahead with those technologies.</p>\n<p><b>Crunching the numbers</b></p>\n<p>The proof is in the pudding. Intel took back share from AMD for the first time in three years, in Q4'20. This comprehensive article covers the details:Intel Claws Back Desktop PC and Notebook Market Share From AMD, First Time in Three Years. Following article contains somemore recent numbers.</p>\n<p>This seems to prove exactly the point of this article: Intel has more or less stopped AMD's momentum with the ramp of its 10nm products. AMD's market share in data centers is still well below 10% (estimated at ~$0.5B quarterly revenue), and if the PC numbers are any indication, AMD's momentum might slow there as well with Intel's 10nm data center CPUs.</p>\n<p><b>How Intel could leapfrog AMD in 2023</b></p>\n<p>By 2023, with Meteor Lake Intel will have a \"breakthrough\" (as Intel called it) CPU architecture that might leapfrog AMD, perhaps reaching Intel's goal of \"unquestioned leadership\". Built on TSMC's 3nm and its own 7nm, it will be about half node to a full node ahead of AMD's 5nm portfolio.</p>\n<p>In other words, from being a year behind in 2019, Intel could actually be a year ahead in 2023.</p>\n<p>Officially, Pat Gelsinger has promised investors only such a leadership by 2024-2025, so if Intel reaches an unmatched leadership position faster (largely because of slow execution by AMD, offsetting Intel's 7nm delays as described), that would obviously be quite bullish.</p>\n<p>In reality, though, it will likely take Intel varying amounts of time to obtain leadership in different categories. For example, as described Alder Lake may already deliver unquestioned leadership in laptops later this year.</p>\n<p><b>Risks</b></p>\n<p>In laptops, Intel's main risk is its product cadence. While according to Pat, Intel has made tremendous progress on 7nm since mid-2020, Meteor Lake has been delayed from late 2022 to somewhere in 2023. Additionally, the 2024 AMD Strix Point product does pose a clear response to Intel's hybrid designs by combining both its Atom and Core architectures, which I called an unmatched capability.</p>\n<p>In desktops, many enthusiasts have taken a stance of waiting for Intel to prove that such a hybrid design also works in this segment. While Zen 4 seems to launch later than many had expected, it also remains unclear how Intel will respond to further core count increases by AMD: will Intel scale only the number of big cores, only the Atom cores, or both?</p>\n<p>In the data center and high-end desktop, the main issue remains Intel's ability (or willingness) to compete on core count. Even if Intel is already competitive (in some areas) with a lower core count, some Arm competitors are already talking about triple digit core counts.</p>\n<p>In FPGAs, despite Intel's vastly improved position in the last few years, this isn't showing in this group's financial and market share. Additionally, both Intel and Xilinx also have a bit of a different strategy, as Xilinx, for example, prefers to call its 7nm FPGAs \"ACAPs\", referring to their various integrated accelerators for things such as 5G.</p>\n<p>Lastly, in process technology, despite Intel's \"full embrace of EUV\", the track record of execution remains on TSMC's side. Additionally, given ASML's (ASML)supply constraints, some have remarked that Intel might not be able to obtain enough tools to ramp 7nm. However, there is no real evidence (such as indications by either ASML or Intel management) that there are any such concerns.</p>\n<p><b>Investment thesis revisited</b></p>\n<p>This article is in part an indirect response to thethesis of another SA contributor, who claimed that Intel, instead,is finished. For example, the author notes that Intel is releasing 10nm chips, compared to AMD’s 7nm and Apple’s 5nm, and hence Intel would be behind. In doing so, the authorfalls in the nanometer marketing games trap, as Intel’s 10nm process is objectively actually (slightly) superior to TSMC’s 7nm. One should be cautious of investment theses based solely on the name of the process technology (“7nm”, “5nm”, etc.), as those names are incomparable across foundries.</p>\n<p>I also differ with regards to the author’s analysis of Intel’s outsourcing strategy. Intel has only said that its 7nm node has encountered issues. Nothing is known about Intel’s 5nm. Hence, outsourcing could be an effective strategy to address the delays in one generation, investing instead more heavily in the next generation.</p>\n<p><b>Investor takeaway</b></p>\n<p>AMD, to me, looks finished. Specifically, the bear thesis of Intel losing loads of market share to AMD by the latter profiting from Intel’s 10nm delays, is finished. In fact, Intel even took some share back since Q4'20.</p>\n<p>Going forward, while AMD certainly will continue to be competitive, I do not foresee AMD to ever again attain such a meaningful tech advantage that would act as a catalyst for broad adoption of AMD, like what had been the case when AMD moved to 7nm, while Intel had to continue to rely on 14nm. I showed this in six areas:</p>\n<ol>\n <li>In laptops, Tiger Lake is already the overall superior overall platform when considering graphics and integrated Wi-Fi 6. Going forward, with Alder Lake, Intel will deliver an unmatched capability with its Hybrid Technology, also catching up on (or even surpassing in) core count and hence likely multi-threaded performance. Meanwhile, AMD won’t move to 5nm until 2023, when Intel may actually leverage TSMC’s 3nm besides its 7nm.</li>\n <li>Intel is also re-investing in the desktop, with Alder Lake, significantly improving competitiveness. Since the desktop remains AMD’s flagship platform, AMD will likely continue to have the upper hand, though.</li>\n <li>Intel is also re-entering the high-end desktop segment with Sapphire Rapids-X, skipping a hypothetical Ice Lake-X.</li>\n <li>Intel frankly is lucky that AMD hasn’t managed to take more market share in the data center. However, going forward with Sapphire Rapids and beyond, Intel will catch up: the performance crown will likely change sides with various product introductions. But that is the point: enterprises likely aren’t going to switch vendors with each new CPU release. Hence, just being competitive with an annual roadmap should suffice to largely stop the threat of severe market share losses. That is what Intel will deliver.</li>\n <li>Going into the acquisition half a decade ago, Intel-Altera was one year behind to the 16/14nm generation. However, Intel caught up and achieved parity (or even leadership) at the 10/7nm generation.</li>\n <li>The general outlook is that, going forward, process technology will matter less as Moore’s Law is slowing, and Intel and the industry is moving towards outsourcing, chiplets and even 3D stacking. More specifically, AMD is failing to transition to 5nm timely, making Intel's 7nm delays less relevant.</li>\n</ol>\n<p>In short, AMD’s golden age started with the (coincidental) confluence of the launch of its Zen architecture and Intel’s multi-year 10nm delays. However, with 10nm now ramping, Intel is quickly regaining competitiveness. Furthermore, Intel’s willingness to leverage TSCM’s most leading edge processes, such as 3nm, even before AMD adopts those, further strengthens the point.</p>\n<p><b>Bottom line</b></p>\n<p>With that, the bearish thesis of Intel losing loads of market share to AMD, to prevent Intel from capitalizing on its many growth opportunities from cloud to 5G and IoT, is arguably definitely finished. However, Intel is still largely valued as if does not have these growth prospects. Hence, if the Street would reconsider this valuation, there may be upside.</p>","source":"seekingalpha","collect":0,"html":"<!DOCTYPE html>\n<html>\n<head>\n<meta http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\" />\n<meta name=\"viewport\" content=\"width=device-width,initial-scale=1.0,minimum-scale=1.0,maximum-scale=1.0,user-scalable=no\"/>\n<meta name=\"format-detection\" content=\"telephone=no,email=no,address=no\" />\n<title>Intel: AMD Threat Is Finished</title>\n<style type=\"text/css\">\na,abbr,acronym,address,applet,article,aside,audio,b,big,blockquote,body,canvas,caption,center,cite,code,dd,del,details,dfn,div,dl,dt,\nem,embed,fieldset,figcaption,figure,footer,form,h1,h2,h3,h4,h5,h6,header,hgroup,html,i,iframe,img,ins,kbd,label,legend,li,mark,menu,nav,\nobject,ol,output,p,pre,q,ruby,s,samp,section,small,span,strike,strong,sub,summary,sup,table,tbody,td,tfoot,th,thead,time,tr,tt,u,ul,var,video{ font:inherit;margin:0;padding:0;vertical-align:baseline;border:0 }\nbody{ font-size:16px; line-height:1.5; color:#999; background:transparent; }\n.wrapper{ overflow:hidden;word-break:break-all;padding:10px; }\nh1,h2{ font-weight:normal; line-height:1.35; margin-bottom:.6em; }\nh3,h4,h5,h6{ line-height:1.35; margin-bottom:1em; }\nh1{ font-size:24px; }\nh2{ font-size:20px; }\nh3{ font-size:18px; }\nh4{ font-size:16px; }\nh5{ font-size:14px; }\nh6{ font-size:12px; }\np,ul,ol,blockquote,dl,table{ margin:1.2em 0; }\nul,ol{ margin-left:2em; }\nul{ list-style:disc; }\nol{ list-style:decimal; }\nli,li p{ margin:10px 0;}\nimg{ max-width:100%;display:block;margin:0 auto 1em; }\nblockquote{ color:#B5B2B1; border-left:3px solid #aaa; padding:1em; }\nstrong,b{font-weight:bold;}\nem,i{font-style:italic;}\ntable{ width:100%;border-collapse:collapse;border-spacing:1px;margin:1em 0;font-size:.9em; }\nth,td{ padding:5px;text-align:left;border:1px solid #aaa; }\nth{ font-weight:bold;background:#5d5d5d; }\n.symbol-link{font-weight:bold;}\n/* header{ border-bottom:1px solid #494756; } */\n.title{ margin:0 0 8px;line-height:1.3;color:#ddd; }\n.meta {color:#5e5c6d;font-size:13px;margin:0 0 .5em; }\na{text-decoration:none; color:#2a4b87;}\n.meta .head { display: inline-block; overflow: hidden}\n.head .h-thumb { width: 30px; height: 30px; margin: 0; padding: 0; border-radius: 50%; float: left;}\n.head .h-content { margin: 0; padding: 0 0 0 9px; float: left;}\n.head .h-name {font-size: 13px; color: #eee; margin: 0;}\n.head .h-time {font-size: 11px; color: #7E829C; margin: 0;line-height: 11px;}\n.small {font-size: 12.5px; display: inline-block; transform: scale(0.9); -webkit-transform: scale(0.9); transform-origin: left; -webkit-transform-origin: left;}\n.smaller {font-size: 12.5px; display: inline-block; transform: scale(0.8); -webkit-transform: scale(0.8); transform-origin: left; -webkit-transform-origin: left;}\n.bt-text {font-size: 12px;margin: 1.5em 0 0 0}\n.bt-text p {margin: 0}\n</style>\n</head>\n<body>\n<div class=\"wrapper\">\n<header>\n<h2 class=\"title\">\nIntel: AMD Threat Is Finished\n</h2>\n\n<h4 class=\"meta\">\n\n\n2021-06-08 23:08 GMT+8&nbsp;&nbsp;&nbsp;<a href=https://seekingalpha.com/article/4433617-intel-amd-threat-is-finished><strong>seekingalpha</strong></a>\n\n\n</h4>\n\n</header>\n<article>\n<div>\n<p>Summary\n\nAlthough competition from Arm is increasing, AMD remains Intel’s biggest competitor, as concerns of losing market share weigh on Intel’s valuation.\nAMD's short-lived laptop competitiveness is...</p>\n\n<a href=\"https://seekingalpha.com/article/4433617-intel-amd-threat-is-finished\">Web Link</a>\n\n</div>\n\n\n</article>\n</div>\n</body>\n</html>\n","type":0,"thumbnail":"","relate_stocks":{"INTC":"英特尔"},"source_url":"https://seekingalpha.com/article/4433617-intel-amd-threat-is-finished","is_english":true,"share_image_url":"https://static.laohu8.com/5a36db9d73b4222bc376d24ccc48c8a4","article_id":"1124688970","content_text":"Summary\n\nAlthough competition from Arm is increasing, AMD remains Intel’s biggest competitor, as concerns of losing market share weigh on Intel’s valuation.\nAMD's short-lived laptop competitiveness is already waning. Intel will further crush AMD with its (up to) 16-core Alder Lake: going from half the core count, to double in one generation.\nIntel is also re-investing in the (high-end) desktop, could leapfrog AMD in the data center, and seems to be overtaking AMD-Xilinx for FPGA leadership.\nAMD is slow to transition to the leading edge in process technology. For example, AMD will not launch 5nm laptop CPUs until 2023, when Intel might have outsourced (TSMC) 3nm.\nGiven all the above, the Intel bear thesis of AMD benefiting from Intel's stumbles, gaining a large tech advantage and taking much market share, is finally finished.\n\nInvestment Thesis\nWhile Arm previously tried (and failed) to enter the data center about a decade ago, in recent years, there has been a more credible resurgence in Arm competition, led by the Apple(NASDAQ:AAPL)Silicon transition and Amazon (AMZN) Graviton chips. Reportedly, Microsoft (MSFT) is also responding to Amazon with its own Arm chips, and with Ampere, there is also a merchant vendor.\nNevertheless, still the biggest bear thesis for Intel (INTC) is AMD (AMD), which has rolled out a competitive Zen-based product portfolio (while, at the same time, Intel was stumbling with 10nm), and hence, Intel would be at risk of losing a major amount of market share.\nHowever, as I see it, AMD’s window of opportunity, which it had due to Intel’s multi-year 10nm delays, is closing rapidly as Intel’s next-gen (much more competitive) chips are entering the market. While AMD will still be a credible alternative supplier going forward, I don’t foresee it holding a significant lead, if any lead at all. I will illustrate in several areas.\nThat means the story of AMD profiting from Intel’s delays to take loads of market share (almost for free) is finished. It simply didn't happen when it had to.\n1. Waning laptop competitiveness\nOne of the main reasons AMD has attracted much attention is because the fastest chips are, not surprisingly, found in the desktop market, where the power budget is much higher. To that end, desktop is the segment where AMD’s flagship technology debuts first.\nHowever, one just has to look at Intel’s earnings to note that the most important segment (financially) is actually the laptop one, which account for ~70% of the market. And given that AMD’s laptop chips have generally launched six months or more after the desktop segment, this means Intel has actually experienced much less pressure from AMD than many would probably have expected (at least financially in its PC business).\nThis changed, though, with the launch of AMD Renoir in 2020. This chip packed eight 7nm Zen 2 cores. In the 15W thin-and-light segment, this meant AMD had twice as many cores as Intel, while in the 35-45W range, Intel hadn't transitioned to 10nm yet, which meant it was competing with its old 14nm Skylake-based IP. Hence, Renoir posed AMD’s first significant threat to Intel’s PC business (even more so than Zen 3, AMD’s other 2020 launch).\nIntel, however, has alreadyanswered Renoir with Tiger Lake:\n\nIn the 15-28W segment, Tiger Lake has a major performance advantage in per-core and graphics performance, as well as a generally superior platform with integrated Wi-Fi 6, AI acceleration, Thunderbolt 4, etc. Tiger Lake is able to compete against 6-core Renoir chips despite having only four cores, which means, only at the very high-end (and low volume), Intel loses in multi-threaded performance.\nMore recently, Intel hasbrought Tiger Lake to 35-45W with eight coresas well. While AMD, for its part, has transitioned to Zen 3 in laptops, benchmarks show the two CPUs are roughly equal.\n\nLooking forward, and it seems, AMD’s competitiveness will fade further rather quickly.\nIntel will launch Alder Lake in the second half of 2021. Alder Lake will implement big.Little, which has been used for many years in the smartphone space to improve power efficiency. Hence, by combining high performance and high frequency cores, Intel will be able to deliver an unmatched capability, since AMD only has one architecture. According to leaks, Alder Lake will come in 2+6 up to 6+8 configurations of Core + Atom cores (and even 8+8 for 55W laptops). Altogether, leaks have indicated Intel is expecting Alder Lake to double in performance.\nHence, as I see it, Intel's hybrid designs will be a big blow for AMD's laptop competitiveness going forward. Indeed: quite recently, there have actually been somerumors of AMD developing its own hybrid designcalled Strix Point. It would consist of eight high-performance Zen 5 cores and four low power \"Zen 4D\" cores, all on 3nm.\nMeanwhile, looking even further out, into 2022 and beyond, AMD’s roadmap is even bleaker than Intel’s. It is highly unlikely AMD will launch a 5nm part before 2023. Given the publicity Intel received from its 7nm delay (from 2022 to 2023), that should raise concerns. While much has been discussed about Intel’s loss of process leadership, this loss only means something if its competitors make use of that advantage.\nIn this case, AMD is not making use of it. Even though 5nm launched in the market in late 2020, AMD’s 2022 laptop roadmap consists of “Zen3+”, which means a refresh of its 2021 laptop line-up. This also means there will be no 5nm laptops from AMD until some time in 2023. That, in turn, means Intel might actually launch its 7nm CPUs before AMD launches 5nm. Never mind if Intel also charges forward with TSMC-based (TSM) 3nm CPU in 2023. The Strix Point CPU from AMD (on 3nm) is also rumored for 2024, while Intel has talked about outsourcing for 2023 already.\nAs a last indication, Intel also took CPU connectivity leadership in laptops with PCIe 4.0, whereas AMD stayed with 3.0. This also means AMD will still be on 3.0 when Intel launches Alder Lake, which will be further upgraded to PCIe 5.0.\n2. Re-investing in the desktop\nBesides defending its laptop stronghold, Intel is also re-investing in the desktop. The desktop is one of the main victims of Intel’s 10nm delays, as Intel has yet to launch its first 10nm desktop CPUs.\nThis will change with Alder Lake in H2’21, as Intel will bring this CPU also to the desktop. That means the desktop will (finally) return to parity with Intel’s laptop segment, in terms of technology. That should substantially improve Intel’s competitiveness. Here, likewise, leaks haveindicatedIntel is expecting 2x performance. This would put Intel on performance parity (or even a slight leadership) against Zen 3.\nSince recentrumors indicate that Zen 4 will launch in Q4'22, this means Intel could be more or less on parity with AMD for at least a full year (if the 2x performance claim holds true across the board).\nAs described, though, for AMD, the desktop represents its flagship segment, whereas Intel has most vigorously defended the (much bigger) laptop space. Hence, I do not foresee Intel necessarily vigorously overtaking AMD. Still, given the seemingly late 2022 launch for Zen 4, it's a bit of pity that Meteor Lake has been delayed. Nevertheless, based on the large jump Intel is making with Alder Lake, the gap should close substantially, especially for all but those who need the highest core counts.\n3. Re-entering high-end desktop (HEDT)\nAnother segment that Intel has basically ignored for the last few years is the high-end desktop. Once proud of its $1700 10-core CPU, these chips immediately became obsolete once AMD launched its Threadripper line. Even with many price cuts, Intel hasn't really had a compelling offering for this segment for years already.\nReports indicate, however, that Intel is outright skipping the Ice Lake-X generation and will move straight to Sapphire Rapids-X. Since AMD lately also hasn't given its Threadripper line the most aggressive roadmap, Intel could bring some serious competition back to this market if Sapphire Rapids-X would launch in 2022.\n4. Overtaking AMD in the data center?\nBesides the desktop, the data center has been the other segment where Intel had fallen substantially behind due to its 10nm delays. Frankly, ever since AMD launched its 7nm Rome CPUs with 64 cores, it has been surprising that Intel has not lost more market share, given that its own offering consisted of 28-core CPUs on 14nm for a long time.\nMore recently, acomprehensive benchmark effort by Phoronixhas indicated that Intel is actually surprisingly competing against these 64-core Milan CPUs with its own 40-core Ice Lake-SP on 10nm.\nIntel's competitiveness will further improve withSapphire Rapids. It will move to Intel’s latest technology, with the same architecture and 10nm Enhanced SuperFin process as the upcoming Alder Lake. It will (almost) close to the gap in core count, with a boost to 56 cores.\nIn fact, in tech forums, enthusiasts continue to debate whether Sapphire Rapids will top out at 56 or 72 cores, as there have also been rumors of the latter variant. In that case, Intel's chances of unambiguously overtaking AMD would be greatly increased.\nAdditionally, Sapphire Rapids, in any case, will take a substantial lead in I/O, with PCIe 5.0 and CXL, as well as DDR5 and HBM support. It also has an integrated data engine (Data Streaming Accelerator), and it will move to Intel’s chiplet design with four EMIB-connected tiles. This means each chiplet will have 14 (or 18) cores.\nLastly, Sapphire Rapids will also substantially widen Intel’s already vast lead in AI performance, with the inclusion of Intel’s version of Nvidia’s (NVDA) Tensor Cores. In aninterview with AnandTechearlier this year, Intel said that AVX-512 (which Intel's current DLBoost is based on) is one of the largest factors for customers choosing to adopt Intel over AMD. So, to that end, Intel expects Sapphire Rapids to improve AI performance by a further 4-8x.\nTo be sure, given the delays of at least several quarters, I do not expect Intel to take an unsurmountable leadership position. For example, in 2019, Intel said that the next-gen 7nm Granite Rapids would launch when Sapphire Rapids will actually launch: in early 2022. This means AMD will transition to 5nm before Intel transitions to its 7nm Granite Rapids CPUs, which gives AMD a chance to one-up Intel.\nNevertheless, for investors, the key point is that I do not foresee that, at any point going forward, AMD will hold a substantial advantage, and for a substantial amount of time, anymore. Even with the 7nm delays, I do not foresee a repeat of the 28-core vs. 64-core situation described above.\nAs a case in point, remember that enterprises do not care so much about who has leadership at any given time, as much as that they demand a long-term roadmap. Customers buy into roadmaps rather than single point products. Intel has such a competitive roadmap at an annual cadence: Ice Lake early 2021, Sapphire Rapids early 2022, Granite Rapids early to mid 2023, Diamond Rapids in 2014, etc.\nWhat this means is that performance will remain contested: Sapphire Rapids will likely overtake AMD, but AMD will respond with Genoa. Then, Intel will respond with Granite Rapids, etc. This raises the rhetorical question: will enterprises bother to switch to AMD if, half a year later, Intel may launch a faster CPU, etc.? The pure performance benchmarks also neglect less quantifiable advantages such as Intel’s vastly larger sales force, etc.\nIn summary, AMD did not even manage to achieve 10% market share while it had over twice the core count (and hence a substantial leadership across the board). That advantage now seems gone for at least the next few generations. AMD simply didn't capitalize when it had the once-in-a-century opportunity.\n5. Challenging Xilinx for FPGA leadership\nI will describe FPGAs rather briefly, as this could be its own topic. As a preliminary note, one should be more cautious here since FPGAs are more esoteric technology in nature.\nFor example, in light of AMD’s acquisition, some remarked that Intel’s Altera acquisition supposedly would be a failed one. If any arguments were given at all (to substantiate that claim), it would supposedly be because Intel has not launched an FPGA integrated with its Xeon CPUs, or because of its lackluster financial performance. However, the FPGA integration argument goes against the industry trend, which is to position the FPGA as an accelerator, just like GPUs which in the data aren't integrated directly into the CPU either. In the future, FPGAs will be connected through the open CXL interconnect, which was developed by Intel, and has also been backed by Xilinx(NASDAQ:XLNX), Arm and even AMD.\nAcquisition issues aside, with regards to actual FPGA leadership, here as well Intel has made much progress to catch up and even surpass Xilinx.\nPrior to the acquisition, Altera had delays with its 20nm generation, which led to it being one year behind Xilinx to the 16/14nm generation. However, almost literally the first day after the acquisition, Intel invested in a second, parallel design team for the 10nm generation. This allowed Intel to catch up and achieve parity to the 10nm/7nm generation, as both FPGAs started sampling around mid-2019, and have recently begun ramping more broadly.\nIn fact, as part of the quite recentIce Lake-SP launch, Intelclaimedthat its 10nm FPGAs achieve up to 2x higher performance/watt compared to Xilinx' 7nm Versal FPGAs. So, arguably, Intel has not just got back to parity, but has in fact leapfrogged Xilinx.\nThere are other aspects as well that demonstrate Intel’s FPGA leadership, including its pioneering use of chiplets (and in the future even 3D stacking), as well as Intel’s transceiver leadership (and indeed those transceivers are separate chiplets): Intel was first to 58G and 116G speed, and first to demo 224G in 2020.\nEven at 14nm, despite being later to initial launch (as described), Intel still managed to launch the first 14/16nm FPGAs with integrated HBM, integrated Arm cores and even PCIe 4.0.\n6. Regaining process leadership (process technology decreasing in importance)\nAs discussed in the first point, having a process technology leadership only means something if the fabless foundry customer makes use of it. In the case of AMD, it explicitly does not, as it will launch a “Zen3+” refresh in 2022, instead of 5nm Zen 4, in laptops at least.\nFurthermore, combining this with the outsourcing rumors, and Intel may actually return to process leadership, as instead Intel may launch 3nm CPUs in 2023, leapfrogging AMD’s 5nm ones. In fact, it seems highly unlikely that AMD will launch any 3nm CPUs in2023 at all, as for example indicated by the Strix Point rumor for 2024.\nAs Bob Swan said in anearly 2021 interview, it would only adopt foundries if it got preferential treatment. Hence, AMD bulls may have underestimated Intel’s position as world’s largest semiconductor company when they perhaps assumed TSMC would be dismissive of Intel’s potential multi-billion wafer orders.\nThe larger point, though, is that Moore’s Law is likely to decrease in importance. For example, TSMC’s 3nm will deliver a real-world shrink of about 1.5x at a relatively slow 2.5 year cadence. This shows Moore’s Law is slowing down. So, even if TSMC continues to have a leadership position, it is unlikely it will have enormous advantage. Pat Gelsinger, for its part, claimed that Intel is already back on track for leadership anyway.\nAdditionally, there are many advances beyond the base process technology, such as chiplets and even 3D stacking. If anything, Intel is actually ahead with those technologies.\nCrunching the numbers\nThe proof is in the pudding. Intel took back share from AMD for the first time in three years, in Q4'20. This comprehensive article covers the details:Intel Claws Back Desktop PC and Notebook Market Share From AMD, First Time in Three Years. Following article contains somemore recent numbers.\nThis seems to prove exactly the point of this article: Intel has more or less stopped AMD's momentum with the ramp of its 10nm products. AMD's market share in data centers is still well below 10% (estimated at ~$0.5B quarterly revenue), and if the PC numbers are any indication, AMD's momentum might slow there as well with Intel's 10nm data center CPUs.\nHow Intel could leapfrog AMD in 2023\nBy 2023, with Meteor Lake Intel will have a \"breakthrough\" (as Intel called it) CPU architecture that might leapfrog AMD, perhaps reaching Intel's goal of \"unquestioned leadership\". Built on TSMC's 3nm and its own 7nm, it will be about half node to a full node ahead of AMD's 5nm portfolio.\nIn other words, from being a year behind in 2019, Intel could actually be a year ahead in 2023.\nOfficially, Pat Gelsinger has promised investors only such a leadership by 2024-2025, so if Intel reaches an unmatched leadership position faster (largely because of slow execution by AMD, offsetting Intel's 7nm delays as described), that would obviously be quite bullish.\nIn reality, though, it will likely take Intel varying amounts of time to obtain leadership in different categories. For example, as described Alder Lake may already deliver unquestioned leadership in laptops later this year.\nRisks\nIn laptops, Intel's main risk is its product cadence. While according to Pat, Intel has made tremendous progress on 7nm since mid-2020, Meteor Lake has been delayed from late 2022 to somewhere in 2023. Additionally, the 2024 AMD Strix Point product does pose a clear response to Intel's hybrid designs by combining both its Atom and Core architectures, which I called an unmatched capability.\nIn desktops, many enthusiasts have taken a stance of waiting for Intel to prove that such a hybrid design also works in this segment. While Zen 4 seems to launch later than many had expected, it also remains unclear how Intel will respond to further core count increases by AMD: will Intel scale only the number of big cores, only the Atom cores, or both?\nIn the data center and high-end desktop, the main issue remains Intel's ability (or willingness) to compete on core count. Even if Intel is already competitive (in some areas) with a lower core count, some Arm competitors are already talking about triple digit core counts.\nIn FPGAs, despite Intel's vastly improved position in the last few years, this isn't showing in this group's financial and market share. Additionally, both Intel and Xilinx also have a bit of a different strategy, as Xilinx, for example, prefers to call its 7nm FPGAs \"ACAPs\", referring to their various integrated accelerators for things such as 5G.\nLastly, in process technology, despite Intel's \"full embrace of EUV\", the track record of execution remains on TSMC's side. Additionally, given ASML's (ASML)supply constraints, some have remarked that Intel might not be able to obtain enough tools to ramp 7nm. However, there is no real evidence (such as indications by either ASML or Intel management) that there are any such concerns.\nInvestment thesis revisited\nThis article is in part an indirect response to thethesis of another SA contributor, who claimed that Intel, instead,is finished. For example, the author notes that Intel is releasing 10nm chips, compared to AMD’s 7nm and Apple’s 5nm, and hence Intel would be behind. In doing so, the authorfalls in the nanometer marketing games trap, as Intel’s 10nm process is objectively actually (slightly) superior to TSMC’s 7nm. One should be cautious of investment theses based solely on the name of the process technology (“7nm”, “5nm”, etc.), as those names are incomparable across foundries.\nI also differ with regards to the author’s analysis of Intel’s outsourcing strategy. Intel has only said that its 7nm node has encountered issues. Nothing is known about Intel’s 5nm. Hence, outsourcing could be an effective strategy to address the delays in one generation, investing instead more heavily in the next generation.\nInvestor takeaway\nAMD, to me, looks finished. Specifically, the bear thesis of Intel losing loads of market share to AMD by the latter profiting from Intel’s 10nm delays, is finished. In fact, Intel even took some share back since Q4'20.\nGoing forward, while AMD certainly will continue to be competitive, I do not foresee AMD to ever again attain such a meaningful tech advantage that would act as a catalyst for broad adoption of AMD, like what had been the case when AMD moved to 7nm, while Intel had to continue to rely on 14nm. I showed this in six areas:\n\nIn laptops, Tiger Lake is already the overall superior overall platform when considering graphics and integrated Wi-Fi 6. Going forward, with Alder Lake, Intel will deliver an unmatched capability with its Hybrid Technology, also catching up on (or even surpassing in) core count and hence likely multi-threaded performance. Meanwhile, AMD won’t move to 5nm until 2023, when Intel may actually leverage TSMC’s 3nm besides its 7nm.\nIntel is also re-investing in the desktop, with Alder Lake, significantly improving competitiveness. Since the desktop remains AMD’s flagship platform, AMD will likely continue to have the upper hand, though.\nIntel is also re-entering the high-end desktop segment with Sapphire Rapids-X, skipping a hypothetical Ice Lake-X.\nIntel frankly is lucky that AMD hasn’t managed to take more market share in the data center. However, going forward with Sapphire Rapids and beyond, Intel will catch up: the performance crown will likely change sides with various product introductions. But that is the point: enterprises likely aren’t going to switch vendors with each new CPU release. Hence, just being competitive with an annual roadmap should suffice to largely stop the threat of severe market share losses. That is what Intel will deliver.\nGoing into the acquisition half a decade ago, Intel-Altera was one year behind to the 16/14nm generation. However, Intel caught up and achieved parity (or even leadership) at the 10/7nm generation.\nThe general outlook is that, going forward, process technology will matter less as Moore’s Law is slowing, and Intel and the industry is moving towards outsourcing, chiplets and even 3D stacking. More specifically, AMD is failing to transition to 5nm timely, making Intel's 7nm delays less relevant.\n\nIn short, AMD’s golden age started with the (coincidental) confluence of the launch of its Zen architecture and Intel’s multi-year 10nm delays. However, with 10nm now ramping, Intel is quickly regaining competitiveness. Furthermore, Intel’s willingness to leverage TSCM’s most leading edge processes, such as 3nm, even before AMD adopts those, further strengthens the point.\nBottom line\nWith that, the bearish thesis of Intel losing loads of market share to AMD, to prevent Intel from capitalizing on its many growth opportunities from cloud to 5G and IoT, is arguably definitely finished. However, Intel is still largely valued as if does not have these growth prospects. Hence, if the Street would reconsider this valuation, there may be upside.","news_type":1},"isVote":1,"tweetType":1,"viewCount":16,"authorTweetTopStatus":1,"verified":2,"comments":[],"imageCount":0,"langContent":"EN","totalScore":0},{"id":180964005,"gmtCreate":1623169950807,"gmtModify":1704197674634,"author":{"id":"3582678388280889","authorId":"3582678388280889","name":"Seb195","avatar":"https://static.laohu8.com/default-avatar.jpg","crmLevel":2,"crmLevelSwitch":0,"followedFlag":false,"authorIdStr":"3582678388280889","idStr":"3582678388280889"},"themes":[],"htmlText":"?","listText":"?","text":"?","images":[],"top":1,"highlighted":1,"essential":1,"paper":1,"likeSize":0,"commentSize":0,"repostSize":0,"link":"https://ttm.financial/post/180964005","repostId":"1109039533","repostType":4,"isVote":1,"tweetType":1,"viewCount":4,"authorTweetTopStatus":1,"verified":2,"comments":[],"imageCount":0,"langContent":"EN","totalScore":0}],"defaultTab":"posts","isTTM":true}