Review Index:

Intel Releases Q1 2018 Results: The Cracks Appear

Subject: Editorial
Manufacturer: Intel

2018: A banner year

Intel has a long history of generating tremendous amounts of revenue and income. This latest quarter is no exception. Intel has announced record Q1 revenues for this year and they look to continue that trend throughout 2018. AMD released their very positive results yesterday, but their finances are dwarfed by what Intel has brought to market. The company had revenue of $16.1 billion with a net income of $4.5 billion. Compare this to AMD’s $1.625B revenue and $81M net income we see that the massive gulf between these two companies will not be bridged anytime soon with either Intel falling or AMD gaining.

View Full Size

Intel has put its money to good use with a wide variety of products that stretch between the PC market and datacenters. While their low power and ultra-mobile strategies have been scaled back and cancelled in some cases, their core markets are unaffected and they continue to make money hand over fist. The company has always been fundamentally sound in terms of finances and they do not typically spend money recklessly. They continue to feature market leading IPC with their product lines and can address multiple markets with the x86 products they have.

Click here to continue reading about Intel's Q1 results!

Perhaps the two biggest pieces of news from Intel in the positive column is that they are retaining revenue from the PC market that has been weakening as of late, even with the increased competition from AMD. That group achieved about a 2% growth year on year. The datacenter side is almost matching the PC side in terms of revenue which has long term advantages for the company as marketshare and demand changes are much more stable than the seasonally oriented PC market. Not to mention that it is much harder for competitors to take away significant marketshare from the datacenter.

These are certainly positive signs from the company that they will continue along their path throughout 2018 without major bumps or disruptions. Their 14nm process and derivatives are still performing well and yields are apparently excellent. It has been an excellent moneymaker for them as it has only been relatively recently that the 3rd party foundries have process nodes available that match Intel in terms of transistor performance and dimensions.

Having put those positive things out there, we also know that not all is well at Intel. The elephant in the room has been the 10nm process that Intel has been developing for years now. It is still not up to spec in terms of consistency and yields. Intel claims to be shipping limited amounts of product on it, but those truly are limited and I doubt the average user can get their hands on it. While Samsung, GLOBALFOUNDRIES, and TSMC are busily producing their 14/16nm products to their customers, they are also moving ahead with advanced process nodes that will not only match Intel’s 10nm, but also potentially surpass it.

View Full Size

Intel’s only major competitor in the x86 business has recently released their 12nm based Zen+ processors. While there are no real dimensional improvements from GF’s 14nm to 12nm processes, what has improved is transistor switching speed, power characteristics, and manufacturability. Faster, more efficient parts that have improved yields going along with a mild redesign/optimization has made the new Ryzen 2000 series of CPUs a true competitor for the latest Intel CPUs on the desktop and mobile markets.

10nm for Intel has been a disaster. The company has essentially lost its process advantage over the competition. They are now on what is essentially an even playing field when it comes to manufacturing products throughout the rest of this year. We have seen AMD’s desktop lineup (or so we have been lead to believe) and know what is coming up in terms of the mobile space. The products, from top to bottom, are very competitive between Intel and AMD. Previously Intel could expect several months of near parity in terms of process performance before they release the next generation of product on a new process node and again achieve a tremendous advantage over their competition in terms of performance and die size. The competition is then forced to lower prices to make their products more attractive to buyers all the while Intel reaps the benefits of increased performance, smaller die sizes, and greater efficiency. Now Intel does not have that luxury for anytime in the near future.

During the conference call Intel announced that it would not be reaching volume manufacturing on 10nm until 2019. When pressed further it was mentioned that it would likely be in 2H 2019. This is unknown territory for Intel. In the past when they have encountered real competition from AMD, they still had a good 18+ months of process superiority, sometimes even longer than that. Now Intel needs to rely on their improved 14nm processes that actually give up some die space in return for better power and switching properties.

View Full Size

The foundries are currently shipping some power efficient 10nm parts and are working on getting 7nm out the door. TSMC is in full production with 7nm which can be compared quite favorably to Intel’s 10nm in terms of dimensions and transistor performance. This production is still very new, but parts are flowing out of the fab. TSMC will most likely be relegating much of the line space to smaller customers with smaller ASICs, but will be transitioning to larger designs in greater volume. We already know that AMD has received 7nm Vega based silicon back, but that part is not meant for general consumer usage. AMD apparently is aiming for more compute oriented markets with 7nm Vega and will release NAVI based parts early next year on 7nm as well.


Video News

April 27, 2018 | 03:40 PM - Posted by Disclose (not verified)

Shouldn't PCPer be disclosing the consulting affiliations with companies on its news posts, as well? Some of this reads like an ad for a company known to fund PCPer's adjacent consulting agency.

April 27, 2018 | 04:10 PM - Posted by Josh Walrath

Which company? Because there are about equal positives and negatives for both here when looking at this year vs. next.

As the author, I have no vested interest in either of the companies. I own no shares of Intel or AMD, nor any of the foundry companies listed here.

April 27, 2018 | 04:20 PM - Posted by RatherRotundAndComicalCommentard (not verified)

My Only Investment is a rare Steve Allen Pog! So I most definitely do not have any vested interest in any company's stocks or bonds.

April 27, 2018 | 06:54 PM - Posted by Jeremy Hellstrom

The real question is if you own a mobile goal post, methinks.

April 27, 2018 | 05:01 PM - Posted by Mike Hunt (not verified)

I can fully disclose that Intel didn't learn its lesson in 2004, and therefore repeating its same mistake.

April 27, 2018 | 04:04 PM - Posted by EpycWillBeMovenOnUpToTheEastSideOfThatServerMarketShortly (not verified)

Oh those non GAAP gross marging figures look as worrysome as the Non-GAAP figures with that all important gross margin figures moving downward slightly but still holding its own over time. And look at that Intel margin history(1) going back to when AMD's Opteron had AMD a larger slice of that server/HPC market share pie. That TTM Net Margin chart between 9/30/2017 and 12/31/2017 is interesting also.

Now wait a few more business quarters as the 2h of 2018 runs into 2019 and beyond!


"Intel (INTC) Profit Margin History"

April 27, 2018 | 04:55 PM - Posted by Bill Murray (not verified)

For datacenters, AMD still bit player unlike Intel. Already 3Qs passed and Epyc has not changed AMD's EESC profitability and margins. Too little volume to even influence the server sector.

"Advanced Micro Devices (AMD) Profit Margin History"

April 27, 2018 | 09:21 PM - Posted by YourLipidToGraymaterRatioIsSkewedTowardsTheFatEnd (not verified)

Well AMD has nowhere to go but up in server market share and Intel is the incumbent with more to lose! So AMD currently being a "bit player" or not does not matter in the historical persective. Those gross margin basis points that are fed into the Wall Street Quants computerized algorithms are only looking to gauge one company's margin trends.

AMD is much less extended in the Server CPU makrest currently so Intel has the most to lose as Intel being the market share leader has to keep that market share in the face of greater competition by lowering its high margins. AMD's margins are too low but just high enough with AMD's low overhead to begin showing some excess/profits remaining after all the quarterly charges are tallied.

I do not think that you are understanding what that Intel margin history is telling what with those historical Intel margin lows directly related to AMD's Opteron lower margin pressures on Intel were at that period of time. So move on to currently and AMD's Epyc line of Zen/Server CPUs are hitting more of the performance CPU and MB feature bells against Intel than Opteron ever could have hoped to do.

You are definitly not able to use inductive logical thought processes in looking at Intel's margin history at the Opterion High market share High water mark and seeing the affect at the time that Opteron had on forcing Intel to have to lower its server CPU margins/markups in order to keep AMD's Opteron server marrket share from getting any higher. Your deductive logical skills are lacking if you think that its smiply a matter of market cap size when it comes to maintaining gross margin percentages business quarter to business quarter and how much weight that Wall Street attatches to those gross margin basis points.

You are mostly using Circular reasoning and basing that on market cap/revenue size and failing to see Intel's finincial position as it relates to its gross margin situation in the presence of any Server Competition and not only AMD's x86 based competition.

This is not a who has a bigger member contest this is a How Wall Street reacts to any lowering gross margin/competative market pressures in a historical perspective and how that's mathematically modeled and used to infer outcomes that the hedge funds utilize to determine long term strategy with respect to investing in Intel's stocks/bonds.

Any attempts at extrapolating fully any future AMD Epyc server market trends more exactily will have to wait for the first business reportng quarter into 2019 before the most accurate measure of AMD's server market share growth trends can be more fully established. The Hedge Funds managers have to look at the past as well as wait more fully into the future. But Hedge Funds have to at least keep 2 positions on how things may work out for Intel's high margins for the better or worse in the coming months after Epyc has started being deployed in server rooms. And that process has just begun to happen after Epyc was vetted and certified fully for production server workloads by those larger cloud server interests.

You are really unable to think properly in an investor aware manner and that shows. This is not a pissing contest with regards to gross revenues is more about how extended any market share leader has become with regards to having to defend at all cost the highest gross margins possible to maintain current share values. And it's how any sorts of competition can affect those High Margins that a company's stock holders have become accustomed to.

Stock Holders/Investors will move their investments to other companies at the slightest indication of any long term gross margin declines once that process begins and that can really scare off investment as any share price begins to show signs of falling at an unacceptable rate. The higher the share value/market cap the bigger the losses in billions that can occur for those very large market cap companies like Intel and Nvidia if their margins/markups have to be lowered as a result of more competition.

April 27, 2018 | 09:42 PM - Posted by YourLipidToGraymaterRatioIsSkewedTowardsTheFatEnd (not verified)

P.S. Your understanding of how long any new server CPU server SKU rollout takes after that product was first released to market is lacking. And that process takes more than a few quarters after release to beging happening as Epyc's deployment has only just started to in earnest be deployed. Those Epyc CPU SKUs have been undergoing the usual longer amount of vetting time by the server interests. New CPUs can not be qucikly moved to server production workload status as that workload production process can not afford to have any interruptions or millions can be lost.

Stability of production workloads is first and formost on these clouds server companies minds where any disruptions can result in millions in losses for the cloud providors and their clients.

Bill, whatever your real name is, you are not very informed as it relates to any real world business/financial matters!

April 28, 2018 | 12:56 AM - Posted by Bill Murray (not verified)

I see nothing but runaway ramblings of a daydreamer. Look how long it took Intel's new Xeon server chips to dominate from launch. Just less than 1Q only with many big customers already deploying shiny Xeon powered new servers. But after 3Qs Epyc still pose too little competition to become any significant player. EESC results reflects this. Peel away conslow chip revenues and immediately will show how little Epyc's contribution for the past Qs since launch. Also shows how miniscule Epyc's volume shipped still is.

April 28, 2018 | 09:09 AM - Posted by BlindBillTheIntelShillDoesNotSeeTheMarginsAsWallStreetDoes (not verified)

Well look at AMD's Opteron and Intel's high margins took a hit as Opteron sales/market share increased! So every precentage point of lost gross margins for a company as big as Intel with its sales volume, well that's millions in lost revenues and share value.

Watch as the Epyc sales and market share increases whittle away at Intel Gross margin figures, even Intel in thir guidence lists increasing competition starting in 2H of 2018 with an expectation that more competition is appearing from both AMD/Epyc And IBM/Nvidia with OpenPower Power9/Nvidia Volta both beginning to change the market in the comimg business quarters.

You Keep talking about AMD's Epyc revenues and that's not the issue it's Epyc lower markups for that very performant Epyc ecosystem and its affect on Intel margins that's really going to do a number on Intel's Fat Fat margin server busines. So Intel will be forced to lower its Fat margins on those server CPUs in order to keep AMD's Epyc market share from getting too large. The Intel margin history chart does not lie if you care to look at that margin historical chart and match Intel's low margin point with AMD's Opteron high water mark that had AMD's server market share around 23% and AMD's share price of around $45.

AMD's Epyc chip revenues are not what matters as the analysis is about Intel's gross margins taking a big hit over time as AMD gets back its Server Market Share. And Opteron is no where near as performant and feature packed as the Epyc/SP3 CPU/MB ecosystem is. Look at the Epyc SP3/MBs and that 128 PCIe lanes and 8 memory channels per socket and that right there has Intel running to respond.

Intel will be forced to lower its traditionally high margin markups on its server kit, and Wall Street is already looking at that Opteron server market share history with is affects on Intel's margins and taking note of the probablities for those Intel Gross Margin Basis Points to begin to go in the wrong direction.

It's not about AMD's Epyc volumes it's about Intel's gross margin shrinkage potential in order for Intel to stave off AMD's Epyc market share gains. And every Percentage Point of lost gross margins for a big company like Intel repersents millions to billions in lost share value if Wall Street begins to lower Intel's ratings.

The bigger you are the harder your share value will fall in that millions of dollers figure for each percentage point of lost gross margins. AMD is so low currently that it only sees skys above while Intel is so high that it can only see that ground rushing up quickly if Wall Street takes notice after Intel is forced to lower is margins in order to keep Epyc from taking even more server market share.

April 28, 2018 | 03:01 PM - Posted by Bill Murray (not verified)

More quotes from Anandtech "The Data Center Group had revenues of $5.2 billion, which is up 24% year-over-year. As we saw in AMD’s earnings, EPYC hasn’t really made an impact on their earnings yet, but Intel continues to dominate in this segment."

April 28, 2018 | 07:50 PM - Posted by LowMarginDietsStarveChipzillasToExtension (not verified)

IT'S Not about earnings it's about Intel's gross margins having to fall once Epyc starts to take server market share! Intel can only stop that by giving up on its higher per unit markups and that's going to give Intel's High Gross Margins a swift kick in the groin!

AMD's Gross Margins are only Around 36% and yet AMD is such a low overhead and lean business operation currently that AMD just posted per share earnings/profits of 11 cents a share.

Intel on the other hand is such a overweight beast that it has to maintain those gross margins at in 63% range or things will start to get dicey. 5.2 billion in revenues does not mean anything stock value wise for a company with as large of a market cap as Intel and it's great to have revenues and revenue growth! But Intel's gross margins must remain higher than AMD's or Intel will not show any earnings per share.

Intel has a fat high overhead business operation with a lot of vary expansive chip fabs that cost billions to maintain and those 10nm fab node R&D costs are in the billions also and will be even more costly for Intel to get fixed, what with the 10nm issues being much worse that Intel's 14nm issues ever where. Intel has no other foundry customers in large enough amounts for Intel to be spread that massive chip fab upkeep/fab process node development cost across any third party customer base like GF, TSMC, Samsung/others can!

AMD has no Fabs eating away at any earnings if they are not kept at as close to full production as is possible and even then that's still costly for upkeep. AMD only pays for wafers and nothing else, any other expesnses are AMD's fab partners' problems!

Intel can not make earnings per share on any low margin sales and hope to cover its massive monthly expenses that chip fabs require, in addition to other expenses of running such a large market cap company with so many employees and middle management. Intel is a huge hungry beast of a company that requires billions more per month to operate than AMD. Intel will never be able to suck it up and match AMD's Epyc Price Latitude and look at that Intel Share price history even when those gross margins have been in the 49%-55% range.

Look at Intel share price/gross margins figures history at around the time that AMD's Opteron had about 23% server market share as see that even with gross margins in a higher range than AMD currently has that Intel can never go that low with is unit markups and risk any margins below the 45%-50% range. Intel has expenses that dwarfs AMD's expenses on a daily/monthly/yearly basis and that requires higher margins/gross margins to be maintained on Intel's share price will tank!

Chipzillas require quite a bit more daily calorie intake than that small AMD mammal! And that Epyc Asteroid is about to knock down any high margin food sources in server world!

April 29, 2018 | 01:52 PM - Posted by Bill Murray (not verified)

You forgot that AMD still has to pay the third party fab's service fees which is much higher than having own fab with own fabbing expenses. Because of those costs, WSA problems crops up and had to negotiate all the the time to bring costs down. Luckily GF owned by same shareholders as with AMD. But still the fab has to make a profit, hence penalties despite being "flexible". Using any other fab than GF would costs AMD much more. This is why Intel have always able to maintain high gross margins on its own chips, plus control ramp production. But AMD cannot do such things, so Epyc's slow ramp is an Epyc fail.

April 29, 2018 | 08:30 PM - Posted by EpycEconomiesOfScaleIntelsHighMarginsHeadingTowardHell (not verified)

No you are clueless! GF spreads that cost of Chip fab upkeep across its entire foundry customer base and if GF wants to charge AMD too much they can not until after 2024. AMD is saving billions on not having to pay for expensive fab equipment or expensive Fab bulidng upkeep and related foundry process node engineering expenses. AMD makes use of TSMC also and even with that renegotiated Wafer agreement with GF, AMD saves millions.

There is a reason that the automakers switched to using third party parts suppliers decades ago and got out from under that expense. And the automakers saved billions and billions by all geting their respective parts made by the same parts supplier companies and the automakers had less plants/overhead to worry about.

Intel alone has to cover all those plant upkeep costs and chip fab related R&D costs and is mostly unable to spread those billions of cost to any other business entity. And Chip Fabs cost in the 10s of billions of dollars to build and equip and if those chip fabs/equipment can not be kept producing at as close as 100% caoacity as possible they bleed money for the owner.

There is an economy of scale in useing those third party fabs with the fabs able to spread those fab upkeep and process node development costs across an enire industry of fabless chip designers. Apple, Nvidia, AMD, and even IBM/most others are now all fabless and they all make use of the same third party chip fabs with those chip fab businesses spreading the fab upkeep/process node development across the fabless CPU/GPU/SOC/other industry that includes GPUs and CPU of all different ISAs.

Intel has not even got its 10nm out on schedule and TSMC and GF are already moving ahead with smaller nodes 10nm(TSMC) or are already going towards 7nm(GF, TSMC) while Intel has to extend its 14nm uasge even longer because Intel is having more issues with is 10nm yields. Apple's A11 is already at 10nm on that TSMC 10nm process node.

AMD can now since that new wafer agreement make use of TSMC so no problems there and really AMD is still saving money even if it still has to pay some extra to GF just to use TSMC/others. And in the current demend environment for third party chip fab capacity AMD having its wafer cost fixed by contract until 2024 assures AMD of space at GF even though there may be others willing to pay GF more and AMD making use of TSMC/others also assures AMD of a second source if GF can not keep up with demand or GF screws up with its 7nm timeline.

Intel can only have to spend more paying for all that expensinve fab equipment on its own and that also includes chip plant building maintainence and all the other billions of dollars of costs even if those plants sit idle. Intel is stuck with those fixed chip plant costs no matter what and Intel can not lower its prices to compete with AMD's Epyc as Intel does not have the latitude to let its gross margins slip below around the middle 50% mark before Intel will lose money and share value.

You obviously did not bother taking a look at Intel's Gross margin history and see how AMD's Opteron affected Intel's gross margins and what that did to big high overhead Intel's stock value when Intel was forced to forgo those high gross margins for more than a few business quarters. And Keller, Raja, or anyone else can not snap their fingers and make Intel a low overhead operation overnight, no way no how will that happen!

Having control over any costly fab to ramp production is not going to do any good if your competition can and does charge so much less for its server SKUs like AMD can and does with its Epyc line of products. And AMD's only charging close to half the cost of what Intel is charging and Zen/Epyc is nearly as performant at IPC as Intel's Xeon! And this is with Epyc better at some workloads than Intel's offerings at a lower Epyc/MB cost then that Epyc Price/Performance metric is what the server clients will look at.

Like I said AMD just reported 11 cents a share earnings with gross margins only around 36% and do you even care to guess what amount of losses Intel would show if its gross margins were only 36%! And Intel would be in serious doo-doo if its margins fell to the 50% level what with all of Intel's High Overhead fixed costs that do not change without so much pain of layoffs and plant closings.

Watch for Intel to very soon begin to trim back in numbers of CPU/SOC offerings in the consumer market before too long as the total Epyc server market share figures begin coming on over the next 3 business quarters. AND Intel has more than AMD/Epyc to worry about as there is that IBM/Nvidia OpenPower Power9/Volta juggernaut coming in broadside from the other direction. Intel will trim is low margin consumer busines down to the bones before it risks any cutting on the high margin side of its product portfolio. For Intel cutting margins means reducing fixed expenses and that will begin to become apparent as the next several Business Quarters go by.

Look at all the IOT and other non high gross margin products and projects already shut down at Intel and just you wait and see what Epyc will do to force Intel to lower its Sky High server markups and margins.

April 30, 2018 | 05:58 PM - Posted by Bill Murray (not verified)

Watching you squirm out meaningless upon meaningless ramblings is such joy to watch. Despite all those dumb excuses made you made, Intel still churning out huge profits and high margins every Q even when their own fabs are being upgraded costing billions. Just shows that you are always out of touch with reality.

May 1, 2018 | 10:47 AM - Posted by BlindsidedByMarginHistoryYouWillBe (not verified)

For each Epyc system that is purchased instead of a Xeon that's twice as much revenue lost on a per unit basis for Intel and some more revenue gained for AMD. And AMD is already showing a profit at 36% gross margins whereas Intel will lose money if its gross margins drop below the 55% metric.

Huge profits are the result of huge revenues thanks to huge margins. But huge margins in the face of stiff competition can see those huge profits swing quickly over to huge losses as those huge margins have to shrink hugely to compete. Intel is a huge high overhead operation with chip fabs, fat middle management pay/perks, and other expenses. AMD is so lean that they are profitable with a margin rate that would place Intel in recievership if Intel had to survive on only 36% gross margins.

Reality is that Intel will have to begin to lower its margins to compete with Epyc's steller price/performance metrics. And Epyc's price/performance metrics are way better than Opteron's price performance metrics ever were and AMD had 23% server market share on its Opteron server market share hay days. And Intel's margin history at that same time period tells it all with Intel's margins at the lowest marks, much lower than Intel's curremt 63% gross margin rates.

You can not see the history for your fanboy blindness but you will shortly get a chance to relive that very sort of experience once Epyc begins to take back that server market share.

Tha Horror! The Horror, of lower gross margins for Chipzilla!

May 1, 2018 | 12:17 PM - Posted by Bill Murray (not verified)

Squirm some more for me. Your ramblings are so hillarious. Epyc fail from the beginning right after launch. Until today Epyc did not make any dent in the enterprise market let alone contribute to AMD's low gross margins. Those 36% gross margins were a result of price hike for GPUs during the mining craze and better ASPs for Ryzen desktop CPUs. But nothing much from Epyc at all in the EESC division. Once the mining craze is over, watch AMD's gross margins drop once more.

May 1, 2018 | 03:05 PM - Posted by TheWhittlingOfLipidsBegins (not verified)

The Fat trimming begins now(1) and accelerates over the next several business quarters as the Rotund Chipzilla goes on a crash costs cutting diet. AMD's Epyc will be doing a price/performance number on that Dino's high gross margin food source, so its more lettuce and less of that delicious and very expensive Japanese Wagyu for the Fat dino's intake!

Can you hear the sounds of that dino's lettuce crunching as Intel's uber high gross margins fall back to earth. Oh that expression Of 絶望(Zetsubō) on Chipailla's face as he chomps down on that bitter lettuce!


"Intel Kicks Kaby Lake-X CPUs to the Curb"

April 28, 2018 | 01:05 AM - Posted by Bill Murray (not verified)

You are still rambling away without much logic nor much rationale. Still stuck at 11 since yesterday. Wall Street stock players not that confident with AMD's turbulent earnings history. 3Qs has passed and Epyc still not doing much in EESC. But less than 1Q new Xeons after launched already made big strides while maintaining gross margins in their usual 60% range including this Q. That is a big difference. As for your character, as I have once quoted "Yes it's true. This man has no dick."

April 28, 2018 | 09:18 AM - Posted by BlindBillTheIntelShillDoesNotSeeTheMarginsAsWallStreetDoes (not verified)

Go back some years on that Intel Margin history chart and its bigger than a side of the biggest barn what AMD's Opteron sale did to Intel's Gross margin history. Yes Intel is maintaining currently but Wall Street and ther Quants have to make use of those past Intel psat hsitory of gross margin shrinkage in the face of stiff competition figures. And Margins that are high currently can not remain high in the face of any potent new Epyc competition.

You are not going to be seen as anything but the negative spin pecker that you in fact are and talking in circles is not going to help that at all!

You failure to take account of Intel's gross margin history will be your undoing!

April 28, 2018 | 03:00 PM - Posted by Bill Murray (not verified)

Dear rambling dickless lunatic, some quotes from Anandtech "Intel is always a company built on strong margins, and although they were down 1.3% from last year, at 60.6% they are still quite strong."

April 28, 2018 | 08:12 PM - Posted by LowMarginDietsStarveChipzillasToExtension (not verified)

That Fat high overhead Chipzilla has some very expensive Chip Fab/Middle Management/employee tape worms that eat up billions. And Intel can not show any earning per share the way AMD does with only 36% gross margins.

Intel will be in serious trouble if its margins get to 55% and below range with its current levels of expenses and Intel can never get its margins down to 36% and still have anything but massive losses that forces massive layoffs.

AMD is such a small low overhead operation with no chip fab upkeep costs of its own that AMD has had per share profits/earning for several business quarters now with gross margins in only the around 36% level. So even small 1%-3% server market share gains will bring in millions to billions in revenues that will mostly go towards per share earnings now that AMD's Epyc SKUs are being deployed for actual production workloads. AMD will be reinvesting a good part of those earnings in newer products and will be able to fund more of what needs to recieve more resources to bulid for AMD's future.

April 29, 2018 | 01:42 PM - Posted by Bill Murray (not verified)

I don't think AMD can even managed 1% market share yet this year at the painfully slow rate of their Epyc ramp. Speaking of Intel's high gross margins, billions lost in contra revenues didn't dent it either. Intel has lots of other high margin business also.

April 29, 2018 | 09:05 PM - Posted by EpycEconomiesOfScaleIntelsHighMarginsHeadingTowardHell (not verified)

AMD's Opteron, and even less performant line of Server SKUs than Epyc, netted AMD around a 23% server market share and look what Opteron did to Intel's high margin server markups and to Intel's quarterly gross margin figures.

You are completely Blinded by size and what high overheads via high Fixed cost does to Intel's latitude in even being able to lower its server market margins/markups.

Contra Revenue billions lost over 3 or 4 years is nothing compared to billions in lost revenues each and every business quarter as AMD's Epyc begins to take back more sever market share. And Intel's Fixed costs will very quickly have to be reduced before Intel can even think about how much it can even afford to lower its higher server SKU markups/margins to compete with AMD's Epyc in Intel's holy cash cow server/HPC market business.

Intel's other high margin producing business unit revenues are dwarfed by Intel's server/HPC market business unit revenues. Intel's high gross margins are the direct result of its server/HPC unit's revenues/markups that are Intel's very bread and butter! And That Intel Holy Cow that will need to be saved at all costs or Intel's share value and market cap will drop rapidly as Wall Street begins to get a better picture of what Epyc's market share numbers will be doing to force Intel to lower its high gross margins.

April 30, 2018 | 06:04 PM - Posted by Bill Murray (not verified)

Still not happening, Intel maintained high gross margins despite Epyc failed rollout. Wall Street not kind to AMD either, now falling from 11 back to 10 and soon back to single digits because of Epyc fail from slow ramp. Back K8 era, AMD owned the fabs so they have complete control and reduced cost per chip made unlike relying on 3rd party fabs. But ever since AMD start making losses they cannot maintain and upgrade the fabs anymore so they had to offload them to rich middle east investors. Sad fate indeed.

May 1, 2018 | 11:21 AM - Posted by BlindsidedByMarginHistoryYouWillBe (not verified)

Fabs cost money even when the difussion lines are idle, and 22nm and lower nm/Node chip fabs cost double what any 32nm above chip fabs cost. Expensive chip fabs can quickly become an albatross around that fab owner's neck.

GF and TSMC/other third party chip fab companies have the best economy of scale with all these third party chip fab's clients sharing in the much lower better amortized costs spread across an entire industry of fabless chip designer clients.

Intel has to foot the bill for all of that by itself with no ability to transfer costs to any other entity but Intel's expense accounts.

Rich investors will move their investments at the first indications of any downward margin pressures forced on Intel by AMD's Epyc competition! Rich investors are the ones with the money to afford the top level Quants and those Quants have those margins basis points feeding into the computers on a regular basis. Those Quants can even rely on historial gross margin basis point data in which to extrapolate how Epyc sales will force Intel to lower its margins over time. The Quants will be running whole different scenarios for best case to worse case on Inrel goss margins basis points falls in responce to Epyc's, Power9's and ARM's, competitive pricing pressures on Intel high markups/margins.

AMD got smart and got out from under that Fab expense as quickly as its could, and even IBM is fabless. Chip Fabs eat money and whether thay are producing or not they consume billions and cost billions to finance with those loan payments due regardless of if the Chip Fab physical plant and equipment is producing or idle! And Chip Fabs lose money if their poduction falls below as close to full capacity as possible, such is the high cost related to chip fab construction and equipment wear and tear and financing!

Wait a few more business quarters and see what will be!

May 1, 2018 | 12:11 PM - Posted by Bill Murray (not verified)

Still yapping away incoherently. So many quarters have gone by and yet Intel remains profitabke with high margins despite maintaining multiple fabs in different countries and continents. Guess you did not know that Intel has a foundry business for customers wanting to use its fabs. That is where all those excess capacity of older process nodes were used to make more money. Meanwhile AMD without its own fab means cost per chip made has increased.

May 1, 2018 | 04:03 PM - Posted by TheWhittlingOfLipidsBegins (not verified)

You are in the delusional state now with your grasping at having to deal with some microeconomics that deal with third party chip fab economy of scale. But that Chipzilla fat will have to be trimmed as each new business quarter after Epyc's full usage on server production workloads begins in earnest(Now) and begins to ramp up from this point on!

Intel's foundries have no large customers and some smaller customers that were are no longer there as the TSMC has already got a 10nm process up and working for some time now. And any clients that may come well that's some time off to be of any help to Intel's high fixed costs that will not allow Intel to price it's third party fab services as low as TSMC, GF/Samsung, and others.

The look of 絶望(Zetsubō) on your face will become apparent as the facts of the market begin to take place! Oh the terrrible despair(Zetsubō) as those Intel high margins fall back towards the ground! Oh your pain, your terrible sense of dispair that so overcomes you that you break down and assume the fetal position and the salty tears begin to form their own little Amazon Rivers across your sunken cheeks. Those black rings around your sunken eyes will make you appear as a sad sad Panda that's lost its fur! Oh the pain the metaphysical angst that surrounds you in a dark cloud of mist. Oh will the Gods of High Margins so abandon your Intel(Mother) as the painful Fat trimming begins to eat away at that high margin Intel Morale as its falls down to the very depths of Share Value Hell!

April 30, 2018 | 01:01 PM - Posted by Anonymously Anonymous (not verified)

Please don't give that other guy fuel to post his great-wall-o-txts.

Honestly, surprised he/she hasn't eaten up storage space with the amount he types out.

Also, if that other guy see this... have you considered a career as a stenographer or court reporter? Another thought, what is your career that allows you to take so much time out of a work day to type of these novelletes? makes one wonder, if you do any work at all, because this is probably not the only website you cause mass-eye-rolling with text-walls-so-large-they-acually-have-real-weight!

May 1, 2018 | 12:34 PM - Posted by Bill Murray (not verified)

Just increasing his anger and hate so he can be seduced into joining the dark side which forever will dominate his life. He's already half way there. His great-wall-o-txts now has with many "!".

May 1, 2018 | 04:13 PM - Posted by TheWhittlingOfLipidsBegins (not verified)

Enjoy those great Walls-O-Text as they so piss you and your sockpuppet above OFF with more Walls-O-Text. That is until that siggle enfeebled brain cell in that vast sea of lipids becomes so Triggered that it sets off a gigaton conversion of fat elements directly into pure E = MC**2.

Oh the Fall of The House Of High Margins and the fat that must be painfully hacked off!

April 28, 2018 | 07:12 AM - Posted by WhyMe (not verified)

I can't dispute that AMD are a bit player in datacenters but that's changing, it's difficult to get a full picture from just the numbers of EESC group because it includes things like the consoles, other semi-custom customers, and embedded so it's difficult to tease out the enterprise from all that.

However Lisa Su said on a conference call with Wall Street analysts this week that...

Server revenue increased in double digit percentage sequentially, across mega datacenter, OEM, and channel customers, Epyc processor unit shipments nearly doubled from the previous quarter. We continued to grow our datacenter momentum with dozens of new design wins across key workloads, including HPC, storage, virtualization, and cloud applications.

Now obviously she could be telling porky pies but i don't see any reason for that, let alone the legal implications of misleading investors, so to say Epyc has not changed AMD's EESC profitability and margins is probably less than acurate.

April 28, 2018 | 09:30 AM - Posted by BlindBillTheIntelShillDoesNotSeeTheMarginsAsWallStreetDoes (not verified)

It's not Fibbing as Epyc sales/revenue figures are just getting started and going from nothing to very small can and does sometimes represent double digit percentage gains.
Epyc has only just this quarter begun to be used in actual production server workloads after the nominal server client vetting and certification process and that takes some business quarters to occur after Epyc was released to market. It's going to take until Q1 of 2019 to suss out fully Epyc's average quarterly server market share gains so until then it's a wait and see mode.

Looking at AMD's past Opterion market share figures is a good indicator of how Epyc may do at a minimum going forward! And Epyc/SP3 is a bit more performant and feature packed than that the Opreron/MB ecosystem ever was.

April 28, 2018 | 03:16 PM - Posted by Bill Murray (not verified)

At only 2.63% with operating income versus revenue, the overall picture rather telling. The other businesses within EESC is taking away the margins or making losses. AMD may as well got 0.1% market share for the Q and double that growth is 0.2% which is still insignificant. AMD could split out the datacenter business from EESC but that would expose their still weak position in server sector. It took nearly 2Qs after launch before very few design wins for Epyc to surface. Q1 2018 does not seem to get much from Epyc. Probably why they keep telling investors about 2019. But by then the possibility of a better competing product from its large rival swamping the market again could put a big dent once more.

April 29, 2018 | 05:01 AM - Posted by WhyMe (not verified)

It's probably a little misleading to make judgments about the enterprise sector on a per quarter basis, you really need to be looking at year on year figures as that's the sort of time frame enterprises work on.

Personally i find it surprising AMD have seen this sort of year on year growth in the enterprise sectors so quickly as it can takes ages for some enterprise customers to make decisions.

April 27, 2018 | 05:10 PM - Posted by willmore

Josh, good article. Informative and insightful. Any similar data on Samsung's fab status as that's relevant to the overall market. Please?

Oh, and the captcha for logged in users is bizarre. Why? I though the whole point of creating an verifying an account was to get rid of such things? Of course, you might never see this post as it looks like the little spinny on the captcha is going to go around and around forever. *sigh*

Edited to add: I had to reload three times to get the captcha to work.

April 27, 2018 | 06:56 PM - Posted by Jeremy Hellstrom

odd ... do you normally log out of the site and only log in during when making a comment?

April 29, 2018 | 07:51 PM - Posted by willmore

Nope, I'm logged in almost all the time. I see the "Welcome willmore!" greeting up there above the "logout" button. And it just gave me a captcha.

April 28, 2018 | 10:42 PM - Posted by PivotEnabled (not verified)

AMD has an inherent advantage in that it can grow (immensely) by simply taking market share away from Intel. PC market growth has been slow to stagnant. This is a problem for Intel.

Look at what happened in the late 80s-90s when WorldCom->MCI->Verizon became a massive company by taking market share from AT&T. A few quarters of success can start to flip market share pretty quickly.

The advantage that Intel has is momentum and its own fabs. As Zen continues to prove itself that will affect Intel's momentum and with the slow growth of the PC market, having your own very expensive fabs that may not stay busy, is not such an advantage. OTOH AMD will need to meet demand to take share away.

April 30, 2018 | 08:46 AM - Posted by braganza (not verified)

hard to believe NVdia is valued at half intels value. (130 vs 255) maybe this whole AI thing is more than a flash in the pan.

April 30, 2018 | 09:59 AM - Posted by TheMarketForAIonlyNowIsJustStartigToGrow (not verified)

What does either of Intel's or Nvidia's market caps have to do with AI being considered a "flash in the pan".

More than half of Nvidia's revenues still comes from consumer/gaming currently and Intel gets the majority of its revenies from the server/HPC market.

Nvidia's non consumer gaming oriened GPU accelerator revenues(AI included) and automotive/other(SOC/GPU) revenues are growing at a faster pace for Nvidia than its gaming GPU revenues. So at some point that is not to far off Nvidia will be making more from its non gaming market sales/revenues than its gaming market sales/revenues.

Nvidia has to compete with AMD for that largest of consumer gaming market sub-classes the mainstream GPU sub-class where the majorty of revenue potential can be had. Nvidia currently has the flagship gaming crown but the overall flagship GPU revenues are smaller compared to the mainstream GPU revenues.

Intel's CPU market/revenue reach is larger than any GPU only market reach and so will be AMD's CPU market/revenue reach once AMD's Epyc SKUs begning to take server market share back from Intel. AMD's Epyc/Server and Radeon Pro WX and Radeon Instinct professional Compute/AI revenues will soon dwarf any of AMD's consumer/Gaming CPU/GPU revenues as the higher markups/margins will come more from AMD's professional market compute/AI focused CPU and GPU lines of products.

Intel's maket cap is larger exactily becaues of Intel's professional market markups/margins from its server/HPC sales. And AMD's Epyc SKUs will force Intel to have to lower its markups drastically in order to compete with AMD's Epyc Price/performance metrics, so look for Intel's Market cap to shrink some because of the lost sales revenues.

Remeber that Intel's higher per unit price/markups on its Xeon products are almost twice in some cases what AMD charges for its Epyc unit price/markups. So For Intel each Xeon Unit sale lost to an Epyc unit sale costs Intel almost twice as much in lost revenues simply because of Intel's higher per unit markups.

AMD is actually earing 11 cents a share on Markup/Margins of just 36% while Intel's higher margins/markups are around the 63% rate. So Intel will lose almost twice as much revenues on a lost per unit sale for every Epyc server SKU that is purchased from AMD instead of a Xeon from Intel.

The bigger you are the more food(Revenues/Gross margins) you require to survive, and that's true for Nvidia also on its mainstream revenues when someone goes with AMD instead of Nvidia. Nvidia has been lucky this time around, just as AMD has been, with that mining demand soaking up all the mainstream Vega offerings and inflating Vega's retail proices.

GPU Mining sales demend Helped both Nvidia and AMD by keeping AMD's Vega GPU's from taking more mainstream gaming sales away from Nvidia while also keeping GPU prices unnaturally inflated. Nvidia was never forced to have to lower is mainstream 1080/1070 markups/prices because the miners bought up all of AMD's Vega SKUs, they even bought up Polaris also.

Nvidia's gross margins would have been much worse if there where no mining demand forcing Vega prices to remain so far above MSRP for so long a time period. Mining helped both AMD and Nvidia gain revenues rather than have to more directly compete at the cost of having to lower their GPU's prices to compete directly with each other. Nvidia never had to lower its markups/margins this time around to compete with any lower cost AMD Vega offerings, as mining kept Vega prices even higher than Nvidia's gmaing GPUs' prices, how unusual is that compared to past GPU sales history before mining was a factor.

May 1, 2018 | 09:59 AM - Posted by Bill Murray (not verified)

More cracks appearing at AMD again. Chris Hook left AMD to join Intel as head of discrete GPU marketing. So more top ex-AMD execs moving to both Nvidia and Intel. Like rats abandoning a sinking ship.

May 1, 2018 | 11:37 AM - Posted by BlindsidedByMarginHistoryYouWillBe (not verified)

Where Employees choose to go are those employes business. Epyc SKUs are fully to market doing production workloads for the clients who purchased them. And it is those clients that will decide based on Epyc's Price/Performance and TCO(Total Cost of ownership). TCO costs have to fiqure in inital equipment perchase costs and the their Financing/Loan cost over time. And equipment that costs half as much but has performance almost the same or better in most cases than equipment that cost double will defnitely cost less to finance over time with loans and total amortized interest payments halved over time also.

Price/Perfirmance and TCO is what the server clients will look at and that TCO includes power usage also.

Oh The Margin Presures on Intel from Epyc's competition and those high margins will have to fall and the profit bleeding will start in earnest. Oh My Margins My Margins! where did my High Margins go, screamed the Rather Rotund Chipzilla! Who will feed me Now!

May 1, 2018 | 12:21 PM - Posted by Bill Murray (not verified)

Daydreaming again as usual. The margin pressures not from Epyc but from other segments like desktop CPUs. Epyc still too little too late again. Yet more news on Nvidia this time, Google TPU Cloud now all going Voltas. Looks like Radeons were shunned again, the usual story.

May 1, 2018 | 02:07 PM - Posted by WhyMe (not verified)

It's not nice to mock the afflicted. ;-)

May 1, 2018 | 04:21 PM - Posted by TheWhittlingOfLipidsBegins (not verified)

Watch as the Intel Sky High margins fall and it's big fat layoffs for one and for all! They'll be starting from middile management and moving on down as the big fat High Fixed Costs are brought to the ground!

May 13, 2018 | 04:53 PM - Posted by Alb (not verified)

Who say AMD server 7nm parts will are out in H1/2019???

Who say GloFo Process is on track?? Who say it runs?? Who say it clocks high enough??

No informations available and without working silicon available no article is a better choice IMO

May 24, 2018 | 04:26 PM - Posted by John S (not verified)

Intel will eventually face a hit in sales with these hardware flaws. The fixes slow a PC and do not completely mitigate the threat. The next firmware appears to be slowing PC's another 1-8% by Intel's own tests. Combined with current firmware losses this could be a big deal to lower end and older models. My thought already is looking at AMD as maybe a better option right now. They appear much less affected by these architectural flaws then Intel. Already I have seen HP and Acer introduce new models with Ryzen as their main CPU then Intel. Some of the reasoning would be performance, but clearly it probably also addresses the Intel issues as well.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

This question is for testing whether you are a human visitor and to prevent automated spam submissions.