U.S. Markets close in 1 hr 9 mins

Xilinx, Inc. (XLNX)

NasdaqGS - NasdaqGS Real Time Price. Currency in USD
Add to watchlist
114.59-1.79 (-1.54%)
As of 2:51PM EDT. Market open.
Sign in to post a message.
  • T
    $AMD conversation
    I'm still thinking no merger with $xlnx news till after AMD's ER. I don't think that Xilinx will state in a Q&A casually as a reply to a question that negotiations are going on to merge with AMD. It has to be a through thought and vetted press release with many details not just a casual answer as a "by the way" - it's a very fundamental move for both companies.
    So my view is a no comments on speculation standard answer will be given at both Xilinx and AMD's ERs. Actual news latter probably first or second week of November after the elections dust settles and Ryzen launching Nov 5 as well as Big Navi Oct 28 with availability Nov 5th too to enable building new systems with both GPUs and CPUs from AMD latest releases.
  • T
    $AMD conversation
    Hummm.... I wonder if $xlnx merge with AMD will only be announced after AMD's ER at a special meeting? Let's see what Xilinx replies to questions on that for a hint.
    I think both companies will want to wait for very strong AMD's ER and Q4 guidance that will ON ITS OWN julp the PPS of AMD over $100!
    To give Xilinx shares owners additional boost as the ratio of AMD's shares for Xilinx shares has been decided!
    See? If Xilinx owners get their ER this week, and it could be not good with so so outlook dropping their PPS, while AMD's ER is triple beating with strong Q4 guidance, and analysts upgrades JUST ON THIS, the Xilinx owners could see a higher "premium" and agree because of the respective pps action on their own ER results!
    But remember this is artificial because "guidance" will obviously be different forthe merged companies as I've posted before!
    So both companies will state that they do not respond to rumors but won't deny it!
    Then a week after AMD's ER, after Nov 5th Zen3 launching revealing huge overclocking at 6GHz or even well over 5GHz, sure higher than 105Watt power but no liquid nitrogen, as I've posted, it will be top standing for AMD! Then it's a good time to have a SPECIAL MEETING jointly to allow questions and discussion with analysts about the merger and its future joint new TAMs!
    Makes sense?
    Let's hear what Xilinx response to the merger questions. If it's nonsense they can easily say so! Fake news should be dismissed ASAP! But so far no such response even as analysts already have stated positive comments like Hans Mosesmann saying 50% higher TAM possible etc.
    So this is a done deal maybe not signed to avoid accusations of not revealing material financial nees, but will be signed after AMD's ER next week!
    Until signed, it's confidential negotiations AMD doesn't have to report on!
  • T
    $AMD conversation
    Understand that $xlnx trying to go after the pure datacenters AI chips market share but it only has PCIe gen 4 connection as I've posted in my previous post.
    This while nVidia's connect it's AI processing to GPUs that are needed for other analytics in the datacenters.
    Xilinx trying to offer separate chips not as FPGAs. Read this article for insight a year ago Xilinx was talking about.
    Hence why merging with AMD is a huge benefit for Xilinx too. Their AI datacenters chips revenues is tiny. Couple with the EPYC chiplets is unique and nVidia's GPUs cannot match it.

    Search for "Xilinx refines AI chips strategy: It’s not just the neural network" Yahoo won't post the link. ..
  • T
    $AMD conversation
    Once the dust settles on the $xlnx merger, and especially both companies explain to the analysts the deep technology reasons how this is huge going after new TAMs, big PPS jump to begin. We're talking way over $100. My previous post outlines just the AI machine learning synergy. There's more.
    As I've posted before on various other components of new TAMs possible, from software defined networks SDN and SWITCHES, to automotive, 5G, telcos and industrial. ..it's like Xilinx has the seeds of MULTI COMPANIES in one. Combining Mobileye, Mellanox, Altera and the nVidia's AI parts.... but ALL require tight integration with the EPYC’s datacenters ecosystem. ARM alone won't do.
    Please read previous post for reference.
    Don't be fooled. ..this is huge for BOTH companies.
  • T
    $AMD conversation
    Regarding the $xlnx merge with AMD and additional new TAMs both companies can address if unified vs separate revenues for each. Consider the networking big TAMs. Both the datacenters networking and the telcos networking. This includes various "smart" management and offloading functions as well as packets routine.
    I cannot provide a long post here on these markets, current players and TAMs but there are many companies building dedicated chips, ASICs and making software to power such all the way to cloud computing and telcos network management.
    Mellanox is one such. Cisco is another. Juniper, Broadcom, Huawei etc many others in perhaps multi $trillions TAM, yes including telco 5G and fiber backbone wired networking.
    How can AMD fit with Xilinx?
    The network operation, datacenters or telcos, has a software stack that runs on x86 and in some cases ARM cores. But, a 64 cores EPYC will be a monster network packet processing engine, routing and monitoring for security purposes, such as hacking or distributed denial of service attacks. Dual EPYC sockets SWITCH telco or datacenters ToR - top of tach or spline offer throuput rival to any current ASIC chips made by networking companies.
    This is all software defined so software running on Zen cores used to inspect packets, route them further or filter threats.
    You have seen Nokia presentation at EPYC Rome Sep 2019 launching of such.
    So what is Xilinx to provide ? Tight coupling within a socket of the circuits needed to physically interface with high bandwidth external ports to which, typically fibers optical adapters connect. There are various standards different speeds for datacenters and telcos. Datacenters need as much bandwidth as possible, moving to 400G. 10G isn't enough at all and 100G is chocking with the new powerful CPUs and GPUs hungry for data in distributed processing between servers, solid state storage and the internet.
    SDN software defined network includes software stack that personalized general purpose bare metal hardware switches depending on the network it runs in. Obviously 64 cores inside the EPYC’s CPUs communicate with each other very fast. It's a network on a chip that can route those packets as needed through the IO chiplet. That IO chiplet should not be just PCIe gen 4 or 8 memory channels. It should have DIFFERENT IO interfaces i.e. IO is Input Outlook to connect with the 400G and 800G interfaces networks use in addition to various signal processing that DSPs can perform.
    That new NETWORKING IO chiplet can come from Xilinx using existing circuits even WITHOUT any FPGAs circuits on it to be the new networking chip solution.
    Connection with the Zen3 chiplets must be very efficient high bandwidth and low latency. That's why the Infinity fabric 3 will be used to connect such between Xilinx and AMD's chiplets impossible to get with PCIe gen 4 or 5.
    This is just one example of how new TAMs emerge from tightly merger of both companies. I'm sure such and others will be discussed at the merger announcement.
    Analysts will have a frenzy PT upgrade as a result.
    I know above is technical and many won't understand but I think there are here who will. Those who deeply understand will benefit. ....
  • T
    $AMD conversation
    33m very low volume, yesterday 40m, the day before 42m and before 46m.
    This to me means no action took place really. As it has dropped 1% it could jump 10% in a day on mews.
    Don't be fooled. Technology driven future is very strong and the merger with $xlnx is strongly beneficial for both companies.
    We're waiting for the final confirmation and the details of the new TAMs joint companies will address not possible before. Analysts will upgrade PT after.
  • T
    $AMD conversation
    Before the $xlnx merger, and a bearish analyst - imagine what Hans Mosesmann will raise the PT after the merger is official:
    "Citing the strong fundamentals, RBC raises AMD's (NASDAQ:AMD) price target from $84 to $92 and maintains an Outperform rating."
  • T
    $AMD conversation
    So you think AMD cannot get a $trillion market cap being THE chips supplier for the world IT? Think again. ..
    Look at Zoom. Look at its PE and market cap way higher than IBM. This is a recent example of a company recently launched and of course stay at home life due to the corona made its business explode. And... after the corona has vaccines etc?. ..
    No the world's multiple trillions dollars a year spending won't slow. Transformation will accelerate. AMD with $xlnx will dominate the future of IT chips and solutions. Surely $1000 in 5 years PPS. ...check Zoom's PPS...
  • K
    AMD is on track with high growth to be $100 by Christmas and $300 in 2022.

    Xilinx investors should get to know AMD better to see if you should hold Xilinx through the merger. After all, if the companies are to get hitched you have the right to know AMD’s prospects.

    2018 AMD was the S&P 500 leader with an 80% gain in share price.
    2019 AMD was the S&P 500 leader with a 134% gain in share price.
    2020 AMD is up Year to date at 70% and will finish the year up 100% Plus.

    But first let’s look at what AMD is on track to do in the next 2 years without Xilinx. Then we can calculate how much a merger needs to be accretive in value.

    AMD growth rates are accelerating. But even at today’s growth rates in 2023 AMD will be reaching 40% market share of the lucrative data centers. Also PC and Graphics will enjoy greater mind share. This would enable AMD to raise ASP to reach Gross Margins of 60 % and be on par with today’s Xilinx margins.

    With those kinds of numbers AMD will broach an $8B quarter in 2022. Analysts will see this $8B quarterly revenue and project a future 12 month Revenue stream of $32B. That will give AMD the following breakdown.

    * Revenue = $32.0 billion
    * Gross Margins @ 60% = $19.2 billion
    * OpEx @ 25% = $8 billion
    * Operating Income = $11.2 billion
    * Taxes @ 20% = $2.24 billion = $8.96 Earnings
    * EPS = $8.96 billion/1.2 billion shares = $7.47
    At 40x = $298.66

    With a forward looking market these levels could be reached in 2022.

    Currently AMD Longs are holding long for 18 months expecting AMD to be $300, without Xilinx.

    Wouldn’t it be nice if your Xilinx shares turned into AMD shares without being dilutive through the merger? I’ll answer what needs to occur for that to happen next.
  • T
    $AMD conversation
    To emphasize, AMD merging with $xlnx has nothing to do with just combining revenues. It has to do with future products not possible to make as a partnership because the need to exchange deep circuits designs mix in both companies and develop future such by existing joined engineering teams.
    IBM paid $34B for Redhat - citing it's key to a bigger scope of ibm's future cloud strategy, not to continue Redhat past business. NVidia's buying Mellanox for $7B and ARM for $40B proposal isn't about combining the revenues. They want extract additional revenues not possible to get with separate companies regardless of partnership deals.
    So for AMD to consider a $30B merge, huge scale and $$$ at the level of giants, including more than Intel has paid for Altera and Mobileye combined, it's for a very compelling reason. Such REQUIRE to be told explicitly and answer questions from Analysts. Not just announcement of the terms how many AMD shares for a Xilinx share etc.
    Looking for a massive new roadmap with products not possible without the merge for BOTH companies that substentially will lead to big revenues. Bigger than the high market share gains already seen e.g. AMD's 32% revenues gain this year vs last, on EXISTING TAMs!
    Has to be made clear and that will push AMD's PPS very high. Way over $100. Upgrades will follow because the model of revenues the analysts had used, require additional new TAMs not included before!
    I hope you longs of both companies understand this perspective.
  • T
    $AMD conversation
    Me thinking the merger with $xlnx may be announced near the ER. It requires answering questions from analysts and detailed outline of the joint operation to support AMD's high performance computing vision, chiplets design, infinity fabric evolution and expanding more into machine learning, AI, heterogeneous computing and NEW MARKETS AND TAMs.
    Cannot just leave it at a PR item describing the terms and that's it.
    It's a big expansion for both companies out of fundamental key existing IP, circuit designs, expert engineering teams developing next gen products and evolution of the IT space with the global ecosystem of technologies, from the 5G broad adoption not just as higher bandwidth lower latency for smartphones and laptops, but additional functions not possible before, including autonomous cars, industrial and government and business transformation, and expanding the cloud to mini cell tower cloud, edge datacenters cloud and the big cloud integration with telco network operation and services telcos can offer.
    This is very big and requires discussions to show both Xilinx and AMD's owners why the whole new merged companies can do way more than individual sum of the parts.
    Once this is done, analysts and investors will understand this is not just combining revenues of two companies but creating different solutions not possible alone, with huge new TAMs revenues. Yes could take 3 to 5 years to get $100B revenues but the total new TAMs for both will be over $1Trillion at market share that commands exclusive solutions not possible with the traditional building boards from discrete chips connected with PCIe or other such open source standards.
    That's the chiplets ecosystem Mark Papermaster webinar reveal on July I've posted about.
    So it's not just an announcement. It requires to describe the new vision, roadmap and new TAMs use cases as well as answer questions.
    After such, I won't be surprised to see massive gains in amd's pps and Xilinx at the ratio agreed on for the merge..
    Don't be fooled! It's happening because it's coherent with all that AMD said before technically on datacenters expansion.
  • T
    $AMD conversation
    I'm reiterating my view of $100B revenues, 10% market share in a $1T total TAMs and an AMD PPS of $1000 in 5 years. While I've made this view after Mark Papermaster webinar in July talking about solutions around chiplets ecosystem inside a socket, with partners and software stack to power this, the merge with $xlnx further expands this roadmap for both companies to become the main chips supplier to the world.
    Anything less than the above is conservative bean counting perspective that fails to understand the technology deeply. Without such understanding, it's impossible to call this right.
    I've previously posted examples of TAMs and forcast of their growth by ODC in the next 5 years to 2025. There are many that today don’t even have good solutions either technically tge specs aren't yet tgere or costs too much or additional macro advancements are needed, as with the broad adoption of the 5G not only as smartphones like Apple's and Samsung"s but radio tower mini cloud stations accross the world, edge datacenters, AI and more that results from such.
    So this is my call now.
    Don't expect any analyst to cite that now as they usually have a 12 months view and even longer view doesn't include new TAMs just higher market share of existing ones.
    But think back to 2016 or 2017. Was there any analyst with a PT of $100 at that time for 2020 year? Nope...
    Don't be fooled. ..this is a huge development for both xilinx and AMD. Cannot be otherwise.
    Details and explanations from both companies in an analysts day or such to follow once signed.
    Upgrades increasing PT to $200+ will come after that.
    Big win for longs of both companies!
  • T
    $AMD conversation
    So pseudo technical experts write articles to explain why the merge between AMD and $xlnx is good more than just financials on the face value numbers. Here's one claiming that it's good for AMD but not for Xilinx. Totally wrong and shows misunderstanding of workloads and profiles of what real applications are like. It's like saying nVidia's GPUs are all that is needed and no CPUs x86 Intel or AMD means much for either games or datacenters. Remember nVidia's Ampere had to use EPYC Rome for a reason vs Intel. ..but is it just PCIe gen 4? Nope. In fact you need much more integration than PCIe provides. Having just GPUs won't work. Hooking a massive GPU with a tiny processor won't achieve gaming performance nor datacenters workloads performance. That's the key to understand. I've posted references to Mark Papermaster explaining the nature of processing required in datacenters. Even machine learning AI applications cannot work on Tensorflow hardware accelerating cores, nor can raytracing work in games just on such hardware. You need multiple types of accelerators all binded together sharing data moves very efficiently through memory and cache not PCIe gen whatever.
    So xilinx very.much needs AMD as much or even more than AMD needs xilinx to really get significant market share in their current TAMs and expand into new TAMs not possible alone.
    This author as well as otgers lack the deep technical understanding of both companies products and use cases.
    Read my previous posts and references for more details but this area is very complex I can tell you that. Only very few expers can understand this. Most analysts will not fully. But Lisa Su and Peng will explain this at the level of TAMs and use cases that could be understood by financial analysts hence big upgrades coming!

    The acquisition of Xilinx will add FPGAs to MPU and GPU sectors that AMD competes with Intel. AMD is after Xilinx' Versal ACAP solution, the next generation of technology moving AMD into 5G and Artificial Intelligence.
    The acquisition of Xilinx will add FPGAs to MPU and GPU sectors that AMD competes with Intel. AMD is after Xilinx' Versal ACAP solution, the next generation of technology moving AMD into 5G and Artificial Intelligence.
  • T
    $AMD conversation
    Longs here's an exercise in research of news. Try find the address and phone number of this research firm that published the hit piece by Futurum Research Daniel Newman on risks with the AMD and $xlnx merge.
    Let's see if you can. It's a firm created in 2016. But...why no address nor phone number with their web only providing a form to "contact"..?
    Think again about this and remember the past cases of tiny research firms with hit pieces then revealed hedge funds profited from.
    It's a great sign as someone really wants your shares...!
    Don't be fooled!
  • T
    $AMD conversation
    From what money does AMD wanna buy $XLNX all programmable??? Does victor peng wanna sell his serious business to X box and playstation chimpmaker Average micro devices??
  • D
    I have been haphazardly accumulating XLNX shares since 1995. I am also an FPGA designer with 25+ years of experience.

    I think AMD needs XLNX more than the other way around. XLNX financial metrics are stronger and products are less "commoditized". While I can definitely see potential in tightly integrating FPGAs and CPUs, it may prove a tough nut to crack. I was very optimistic about this when Intel bought Altera, but not much has come of it, so far. The integration is also less of a hardware challenge, than a software challenge (i.e. compiler).

    At a $30B valuation, likely paid with overvalued AMD shares, I will definitely be voting NO with my shares. I think XLNX has much potential on its own. And I expect there will be other suitors down the road.
  • T
    $AMD conversation
    Mark Papermaster on FPGAs - $xlnx obviously now - chiplets mentioned last July as I've posted. To give you an idea, this is not about financial numbers and current revenues ar all. TMP is very deep technical person. Read all related to FPGAs in this interview from last year - sample:

    "TPM: Where do you think FPGAs are going to play in future supercomputers?

    Mark Papermaster: Going forward in exascale machines, our thought is very straightforward from a strategic standpoint. We have invested in a CPU plus GPU solution. You need to accommodate scalar workloads, and provide the densest floating point you can for vectorized workloads, and there is a third class: very tailored workloads that you can customize for FPGAs and get a significant performance gain. That’s where FPGAs play. Our view of exascale computing is that it is going to be heterogeneous, it is going to be CPU and GPU, it is going to have massive scale, and it has to be able to accommodate tailored accelerators. The latter is done most proficiently with FPGAs because as algorithms change, you can reprogram and continually optimize and get better acceleration."

    It's clear isn't it? Way bigger TAMs and pps gains as a result.

    It is funny to think that in a certain light, AMD has Big Blue to thank for its resurgence in the datacenter. And not because IBM is not good at crafting
    It is funny to think that in a certain light, AMD has Big Blue to thank for its resurgence in the datacenter. And not because IBM is not good at crafting
  • T
    $AMD conversation
    And yet..low volume of 46m only, way under the, still declining, daily average. What does it tell you? Daily average itself declining too. It's a sure sign of ACCUMULATION of AMD's shares by institutions! With all the action last week, even as $xlnx leak news, and no panic selling other than the daily traders andva few poor soul shorts. So... I'm expecting a big PPS jump as we get closer to the ER end of this month.
  • T
    $AMD conversation
    The way to think about $xlnx merging with AMD, isn't about FPGAs market at all. Yes, those pure applications requiring FPGAs will continue revenues. But the bigger new TAMs is the Xilinx designs used tightly with AMD's Zen chiplets. Xilinx can pit a smart AI FPGA as the IO chiplet and connect with a few Zen 8 cores chiplets. The FPGAs have DDR5 memory controllers and PCIe gen 5 already. Remember FPGAs are used for prototyping of processors and other chips and are the first to have advanced circuits so developers canbuse such to prototype before a chip is made.
    If you look at all such existing circuits Xilinx has, added in a heterogeneous accelerators based EPYC ecosystem, as Mark Papermaster webinar revealed, and the many customers using Xilinx, and the software stack components existing to power such circuits for different use cases, from autonomous cars driving to AI machine learning, you understand we don't have a business as usual play here. It's enabling many new use cases and TAMs not yet in existence.
    Hence Hans Mosesmann 50% bigger TAM for AMD is not including these new use cases. Just existing ones.
    This all will be explained by AMD and Xilinx after the merger is announced in details.
    The AMD's PPS will jump high making Xilinx current shares holders very rich too.
    The ratio of how many AMD's shares current Xilinx shares holders get has been decided already, so AMD's PPS jumping give them way more $$$ in value. Obviously they will agree.... Otherwise they are going to lose a lot... AMD's shares owners win win in any case.
    I'm expecting over $100 PPS very soon.
  • T
    $AMD conversation
    So the net is, $xlnx leak by anonymous sources to the WSJ is significant and was manipulated to pretend it is bad to hurt amd's pps while pumping xilinx. But that interpretation was by Wallstreet controlled media shills to buy amd cheap while selling Xilinx after the bump.
    News must be announced asap as even with a quiet period if significant news happen, amd must reveal even while in a quiet period.
    The SEC needs investigate the leak because someone made huge money out of this dropping amd's pps and soon it's jumping on the reveal and explanations how it creates new TAMs for both companies.
    The ratio of amd's shares per xilinx shares won't change now it has been fixed.
    Read history posts to see how this is big news for both companies.
    Don't be fooled. ..!