You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!

Intel reports slight Beat & In line guide Focus is on slowing server group & lower GM

Robert Maire

Moderator
Intel reports slight Beat & In line guide Focus is on slowing server group & lower GM

Capex cut $10B to $9.5B implies no hurry on 10nm. Intel reported a slight beat coming in at $0.74 on $14.9B of revenues versus analysts at $0.63 on $14.8B - Gross Margins were 64.3%. The stock sold off after the announcement as investors were concerned about downward trajectory on GM potentially to 58% in Q1 2016. Server chips appear to be slowing as growth was lower than expected.

View attachment 16334

Linked to China and Global weakness...
We think that much of the server slowing is obviously due specifically to China weakness as well as just general global economic sluggishness. Corporate cloud spending has to have a negative impact.

Downward PC vector remains unchanged...
The downward spiral of the PC market continues at more or less the same pace as expected . At the very least the rate of decline did not increase

Capex was cut from Novembers $10B down to $9.5B...
Management characterized it as "pencil sharpening" but a further cut rather than an expected increase does not make us feel warm and fuzzy about the 10nm ramp. Although in our last report we had heard Intel was moving up some equipment spending the top line number is going down. Given the recent news from TSMC upping their capex it seems clear that we are looking at a very tight race at 10nm. Although TSMC's 10nm and Intels 10nm are not the same fruit its still a PR war that Intel may suffer from.

A 2, 2.5 or 3 year cadence?
Although management was quick to support a return to a 2 year cadence from the tick tock hiccup at 10nm that may be 2.5 or 3 years we did not feel it was definitive as it sounds more like a goal than a definitive imperative.

ASML negative impact...
Management said one of the reasons they were not sure about a return to a 2 year cadence is that they are not sure that ASML will be ready with EUV in time for 7nm. We have not heard a lot of positive news coming out of the EUV space and will get an update next month at SPIE but if ASML cant get it together for Intel at 7nm thats pretty sad indeed. Intel has process flow at 7nm without EUV and they have to start nailing things down sooner or later. If EUV isn't "in the bag" soon it will likely be shut out of Intel at 7nm. Obviously good news for dep and etch folks like Lam and AMAT not great for ASML....

Equipment reuse and longer depreciation..... Getting more blood out of an equipment stone...
Intel management did talk about extending depreciation time for equipment as they are getting longer life out of them. Obviously depreciating equipment over a longer period of time is a good financial trick to goose numbers overall. Intel is likely getting more life out of tools but the longer depreciation schedule makes compares to previous realities a bit harder. The more Intel ramps up memory the more use it has for older tools , the longer EUV is delayed the longer existing litho tools will suffice with more dep and etch.

Not a lot of 10nm chatter on the call......
We were surprised not to hear more discussion about the roll out of 10nm. If I was BK I would be crowing about how the ramp of 10nm is going to reinvigorate the product line and accelerate away from the competition ....but we didn't hear that enthusiasm. It makes us a little concerned that 10nm is tougher/slower than expected or isn't providing the "juice" that was hopped for.

Maybe we have hit an asymptotic point where customers and applications don't get excited about incremental improvements in performance and power consumption

Sounding a lot more like TSMC...
We couldn't help but come away feeling that Intel is clearly invading TSMCs home court as management talked about all the connected technology things on the planet as it "pivots" away from a PC centric world. The problem is that its not wide open greenfield as TSMC has been there for a long time which may make the "pivoting" into that space a bit tougher all while in an slower economic environment. maybe gross margin are going to face even more pressure. We'll find out....

The stocks...
We remain somewhat positive on Intel and would get aggressive if the market punished it too much. We would probably view $30 as a near term low point and probably a good entry point if it got there.

The equipment stocks which were on fire today after TSMCs capex increase will likely see a fade in enthusiasm as Intels cut somewhat offsets TSMC increase.

The comments on ASML were very negative while positive for multi patterning dep and etch, companies like LRCX and AMAT. We would expect metrology and yield management to do well from what we perceive as difficulty at 10nm, at Intel and elsewhere.

We had gotten more positive on the equipment stocks two days ago after positive comments by UCTT, AEIS & MKSI and remain positive as Q1 will likely be better than originally expected. Obviously we were well rewarded with todays performance which may be tempered tomorrow. We continue to hear positive news from our sources in the industry but everyone remains concerned about the sustainability and Intels cut probably adds to those concerns
 

Don Dingee

New member
There has been a lot of discussion lately about 10nm and 7nm throughout the SemiWiki threads. I'm sure I sound pessimistic, and it's not that - I'm fascinated by the technological advances and am rooting for breakthroughs. However, I'm a realist, and I've seen too many organizations build up only to undergo the human costs of tearing them down as plans fall short of expectations.

I think economically, things are completely unclear after 10nm. The investments in both fab and design are huge. The applications that require the technology are near flat plus or minus single digits. I know history dictates we get smaller, cheaper, faster. History also proves if you can't deliver something like a 2x-4x improvement, don't bother, because nobody will upgrade and growth will slow to a crawl.

We are seeing the first signs of deep structural changes - a slowing in the cadence, difficulty in making the advanced nodes work, keeping current nodes longer.

Intel has no choice but to plow ahead. Their hand is being forced by TSMC pushing 7nm. The question is how big will those investments actually be at the onset. This is going to be an interesting dance between piloting the technology, creating actual demand, and bringing capacity online.
 
Top