News AMD meets EPS, misses on revenue

[Hitman928]

TL;dr
{
Non-GAAP EPS of $0.08, matches expectations.

Revenue of $1.42B, $1.44B was expected. Revenue from semi-custom business and consumer GPU sales down, causing lower revenue. Strong Ryzen and Epyc sales offset much of the missing revenue as well as some data center GPU growth.

Projected high single digit revenue growth for 2019.
}

More details. . .

https://seekingalpha.com/pr/17395264-amd-reports-fourth-quarter-annual-2018-financial-results

For full year 2019, AMD expects high single digit percentage revenue growth driven by Ryzen, EPYC and Radeon datacenter GPU product sales

Developing. . .

 

[godihatework]

Because under-promising and over delivering works much better than the other way around.

That’s not how this works. The street expects accuracy, not surprises, when guidance is provided.

 

[zinfamous]

That’s not how this works. The street expects accuracy, not surprises, when guidance is provided.

Well, how do they project based on as-yet unknown dirty tricks that Intel is very well known to employ, and very likely will within the year? I think maybe that wiggle room is baked into this guidance.

 

[AtenRa]

They didn't specify what category they were leading in- could be "highest power consumption"

Actually they did,

 

[Abwx]

https://www.computerbase.de/2019-01/quartalszahlen-amd-q4-2018/

An interesting graph that raise some questions.

In principle they should have kept more or less the pace that can be observed in Q1/Q2 2018 given the state of the competition, but instead there is a sharp decline that not only accelerated in Q4 2018 but is expected to keep on going in Q1 2019.

This can be hardly explained by the graphic cards market recent oversupply and resultent excess inventories, beside the CPU sales that saw 50% growth in units sold is hardly a big number given their maketshare at the time.

 

[piesquared]

https://www.computerbase.de/2019-01/quartalszahlen-amd-q4-2018/

An interesting graph that raise some questions.

In principle they should have kept more or less the pace that can be observed in Q1/Q2 2018 given the state of the competition, but instead there is a sharp decline that not only accelerated in Q4 2018 but is expected to keep on going in Q1 2019.

This can be hardly explained by the graphic cards market recent oversupply and resultent excess inventories, beside the CPU sales that saw 50% growth in units sold is hardly a big number given their maketshare at the time.

50% is huge. The growth in 2018 was from Globalfoundries, there is only so much capacity available. The steady growth allows them to rebuild solid foundations rather than exploding growth built in a house of cards. 2018 sets them up for a strong move in 2019 with more capacity on 7nm from TSMC along with a versatile and high performance Zen 2 design.

 

[piesquared]

Actually they did,

Yep, Picasso and Raven Ridge before that are the best mobile processors available.

 

[DrMrLordX]

Excellent results and guidance from AMD. Vega seems to be doing extremely well in the data center.

RTG on the other hand is a complete dumpster fire.

How . . . wait . . . what?

That’s not how this works. The street expects accuracy, not surprises, when guidance is provided.

The street also employs analysts who are complete weasels. Actually, that's not fair . . . to the weasels.

AMD: we met our guidance!
Analysts: we came up with our own guidance, and you didn't meet our expectations. Sell!
Various investors, hedgefund managers, etc: YES GODLIKE ANALYSTS, RIGHT AWAY!
AMD: um?
Analysts: by the way, we totally aren't shorting you or anything. That would be unethical.

 

[Hans de Vries]

AMD: we met our guidance!
Analysts: we came up with our own guidance, and you didn't meet our expectations. Sell!
Various investors, hedgefund managers, etc: YES GODLIKE ANALYSTS, RIGHT AWAY!

Hmm, something changed stock up 30% now since reporting.

 

[exquisitechar]

I missed that it was said that the split between datacenter CPU and GPU revenue was roughly the same. That's actually shocking. Has AMD been taking some real marketshare from Nvidia in such a short amount of time since launching Radeon Instinct?

 

[H T C]

Hmm, something changed stock up 30% now since reporting.

Quick: report more often ...

 

[Hitman928]

I missed that it was said that the split between datacenter CPU and GPU revenue was roughly the same. That's actually shocking. Has AMD been taking some real marketshare from Nvidia in such a short amount of time since launching Radeon Instinct?

tl;dr If we can trust Lisa Su's statement at face value, it looks like AMD has reached more than 10% GPU datacenter market share in about a years worth of time since MI25 was released.

No one knows for sure because they didn't give specific numbers, but taken at face value, yes, the Radeon Pro/MI line has gained market share much more quickly than expected. AMD said that they saw most of the success in virtualization (cloud based) environments. They claim to have full hardware support for GPU virtualization which they say Nvidia currently lacks, so it would make sense if true. They also said they have had some success in machine learning builds and have started to see some HPC revenue as well.

A writer on seekingalpha did some napkin math on it: https://seekingalpha.com/article/4236795-amd-gaining-much-rapidly-nvidia-market-realizes

 

[Phynaz]

How . . . wait . . . what?

Fan Fantasy vs. AMD presentation.

Which do you think is correct?

 

[DrMrLordX]

Fan Fantasy vs. AMD presentation.

Which do you think is correct?

Hell if I know. I had suspected that Mi25, MI50, and MI60 were selling well. Well enough that AMD effectively bagged on the entire consumer dGPU market. But hey if there's something in that earnings report that shows that AMD's pro cards can't sell then feel free to tell us all what we're missing.

tl;dr If we can trust Lisa Su's statement at face value, it looks like AMD has reached more than 10% GPU datacenter market share in about a years worth of time since MI25 was released.

That's a pretty big deal.

Hmm, something changed stock up 30% now since reporting.

Market's up broadly since the Fed signaled they're gonna slow down interest rate hikes (probably). That may have tamped down any negatives from analysts scolding AMD.

Or maybe, just maybe, investors are starting to ignore analysts a bit more. If so, great!

 

[Tuna-Fish]

I missed that it was said that the split between datacenter CPU and GPU revenue was roughly the same. That's actually shocking. Has AMD been taking some real marketshare from Nvidia in such a short amount of time since launching Radeon Instinct?

One thing to consider is that they are shipping a very large supercomputer project. I wonder how exactly that deal is structured -- will all those GPUs show up in the results of a single quarter?

 

[Phynaz]

Hell if I know. I had suspected that Mi25, MI50, and MI60 were selling well. Well enough that AMD effectively bagged on the entire consumer dGPU market. But hey if there's something in that earnings report that shows that AMD's pro cards can't sell then feel free to tell us all what we're missing.

It's all in the slide deck and the 2019 expectations.

Q1’19 revenue expected to decrease ~12 q/q and 24% y/y. The q/q decrease is primarily driven by continued softness in the graphics channel and seasonality across the business. The y/y decrease is primarily driven by lower graphics sales due to excess channel inventory, the absence of blockchain related GPU revenue and lower memory sales. Semi-custom revenue is expected to be lower y/y while AMD Ryzen, EPYC and Radeon datacenter GPU sales are expected to increase.

Sales of Ryzen processors and record datacenter GPUs offset lower GPU sales
Client processor unit shipments grew by more than 50% y/y; Ryzen represented over 80%
of client sales; highest client computing revenue in more than 4 years

Ryzen sales are increasing, yet they expect $500M - half a billion dollars - less revenue in 1Q2019 than in 1Q2018.

Simple math tells you were the sales drop is.

Slide deck

 

[DrMrLordX]

It's all in the slide deck and the 2019 expectations.

Ryzen sales are increasing, yet they expect $500M - half a billion dollars - less revenue in 1Q2019 than in 1Q2018.

Simple math tells you were the sales drop is.

Slide deck

What you're missing here is that AMD is taking losses from depressed sales of consumer dGPUs - particularly Polaris (which sold like hotcakes during the mining boom). They're reporting healthy increases in datacenter GPU growth. RTG has made a conscious decision to focus on deep learning/AI/compute. That's all you're seeing here. Read your own quotes.

And honestly, why wouldn't AMD take losses from the consumer segment? All they have is Polaris and Vega. It's like they aren't even trying to compete for consumer marketshare.

 

[AtenRa]

It's all in the slide deck and the 2019 expectations.

Ryzen sales are increasing, yet they expect $500M - half a billion dollars - less revenue in 1Q2019 than in 1Q2018.

Simple math tells you were the sales drop is.

Slide deck

500M less revenue is due to both lower Blockchain AND Semi-Custom sales, not only to GPUs.

Q1 2018 revenue by segment

Computing and Graphics = 1,115M (Blockchain sales estimated to be ~10% of 1.647M of revenue)
Enterprise, Embedded and Semi-Custom inside = 532M

Q2 2018 revenue by segment

Computing and Graphics = 1,086M
Enterprise, Embedded and Semi-Custom inside = 670M

Q3 2018 revenue by segment

Computing and Graphics = 938M
Enterprise, Embedded and Semi-Custom inside = 715M

Q4 2018 revenue by segment

Computing and Graphics = 986M
Enterprise, Embedded and Semi-Custom inside = 433M

 

[Phynaz]

500M less revenue is due to both lower Blockchain AND Semi-Custom sales, not only to GPUs.

It’s due to the reasons presented by AMD, not some bs made up by you.

Jesus, did you even bother look at the presentation?

 

[AtenRa]

It’s due to the reasons presented by AMD, not some bs made up by you.

Jesus, did you even bother look at the presentation?

No , I just made it up

Lisa Su

Turning to our enterprise embedded in semicustom segments, revenue was flat from a year ago as a double-digit percentage decrease in semicustom revenue was offset by strong growth in EPYC processor sales. As expected, semicustom sales were down from a year ago based on where we are in the current console cycle. This console generation remains one of the most successful ever as Microsoft and Sony combined have now shipped well in excess of 120 million AMD powered consoles.

Devinder Kumar

Let me turn to the details of the fourth quarter results, fourth quarter revenue, gross margin, operating margin and earnings per share all improved year-over-year. Quarterly revenue of $1.42 billion was up 6% from a year ago. Strong sales of Ryzen and EPYC processors and data center GPUs more than offset lower channel GPU and semicustom sales during the quarter.

In Graphics, sales were down year-over-year due to negligible Blockchain related revenue in the fourth quarter of 2018, as well as elevated levels of Graphics inventory in the channel. Total Graphics revenue grew sequentially, driven by strong Radeon datacenter GPU sales.

From Q&A

Toshiya Hari

Lisa, I had a question on your full year 2019 outlook. You guys are guiding revenue growth in the high-single-digit range. Embedded in that outlook, what growth rates are you assuming for your core businesses computing, graphics, semicustom and server CPUs?

Lisa Su

So relative to the full year guidance of up high-single-digits, the primary growth drivers are Ryzen and EPYC or the two largest growth drivers with datacenter GPU. Also, we're expecting that to be up. And then we would expect that semicustom will be down. If you look at life cycle of semicustom will be in the seventh year of the console cycle. And so we expect semicustom would be down approximately 20% and then we would expect consumer graphics to also be down let's call it double-digits as we really burn off some of the channel inventory that we see in the first half of the year.

 

[Abwx]

Marketshare trends for mobile/DT/server in the last 5-6 quarters :

https://www.computerbase.de/2019-02/prozessoren-marktanteile-amd-desktop-notebook-server/

 

[PotatoWithEarsOnSide]

So AMD went from 12.0% to 15.8% in a year; almost a 33% increase in their own sales volumes.
Wasn't their target to get to around 30% within 5(?) years?