Will Intel's transition to data-centric business work?

[Subramaniam]

Intel reported 34% jump in 3Q profits, primarily due to strong growth in its data-centric businesses. Even though the PC segment, which brings in lion's share of total revenue, revenue was flat, it was able to post revenues of $16Bil, which is up 2%. 3Q results show that $INTC's transition from PC market to data-centric business is bearing fruit.

3Q data-centric businesses grew double digits (IoT revenues jumped 23%). As the PC shipments doesn't show any sign of improvement, its PC segment growth remains flat. It now expects to see high growth in IoT, AI, autonomous driving and cloud markets. $INTC expects FY17 EPS to be $3.25 and revenue of $62Bil, up $700MM from prior forecast.

Will Intel be able to take on AMD and still make sure it is maintaining its market share?

Intel's 3Q Highlights

 

[Daniel Nenni]

I think that will be a difficult transition especially when they have the same Intel management/culture. Intel has tried to pivot many times and have failed. They are very lucky the general purpose CPU business has been a constant over the years with little or no competition. GPUs and AI chips may change that, AMD as well, but I still do not see Intel as being able to pivot without a serious management AND culture change. Just my opinion.

 

[raghu78]

Intel reported 34% jump in 3Q profits, primarily due to strong growth in its data-centric businesses. Even though the PC segment, which brings in lion's share of total revenue, revenue was flat, it was able to post revenues of $16Bil, which is up 2%. 3Q results show that $INTC's transition from PC market to data-centric business is bearing fruit.

3Q data-centric businesses grew double digits (IoT revenues jumped 23%). As the PC shipments doesn't show any sign of improvement, its PC segment growth remains flat. It now expects to see high growth in IoT, AI, autonomous driving and cloud markets. $INTC expects FY17 EPS to be $3.25 and revenue of $62Bil, up $700MM from prior forecast.

Will Intel be able to take on AMD and still make sure it is maintaining its market share?

Intel's 3Q Highlights

Intel is at close to 100% x86 server market share. There is no way thats sustainable given that AMD has very competitive EPYC server chips. EPYC server platforms from OEMs are arrving in Q4 2017 but the real server volume ramp happens in 2018. AMD has a competitive CPU architecture and has just started introducing desktop, notebook and server products. These products will ramp in 2018 and take significant market share. I think Raven Ridge is a very impressive notebook chip. imo AMD should be able to get 25% desktop share , 20% notebook share and 10% server market share by end of 2018. The real test for AMD is if they can follow up with impressive 7nm chips in 2019. If they do that I think a tipping point will be reached and Intel will not be able to stop further significant revenue and profitability reduction.

 

[count]

Here is the problem: The large volumes of the PC business is necessary for Intel to keep the lead in the data center business. Without that volume to spread the costs over things like building leading edge foundries, investments in architecture, R&D spending, are not economic. I think cracks will start to emerge as the foundries overtake Intel at the leading edge over the next couple of years.

What Intel is doing is retreating upmarket where margins are higher, but this is what exposes companies to disruption in Christensens definition of the term.

 

[Daniel Nenni]

Nvidia is one of the best pivoting companies in semiconductor. Can you imagine if Intel bought Nvidia and installed Jensen as CEO instead of BK back in the day?

 

[hist78]

Nvidia is one of the best pivoting companies in semiconductor. Can you imagine if Intel bought Nvidia and installed Jensen as CEO instead of BK back in the day?

Almost impossible. Because it's Intel and often it doesn't like to do thing logically.

 

[count]

Nvidia is one of the best pivoting companies in semiconductor. Can you imagine if Intel bought Nvidia and installed Jensen as CEO instead of BK back in the day?

I think Otellini was more responsible than BK for Intel problems today, although BK did not help at all. Never let an MBA run a tech company, they are all sheep. In MBA in my mind means "I am risk adverse and money focused." That's great for running a bank, not a company that needs to constantly stay at the leading edge and fight off disruption.

 

[SimonYoung]

imo AMD should be able to get 25% desktop share , 20% notebook share and 10% server market share by end of 2018.

What are you smoking, crack?

Simple thought experiment for you. How many chips in each class does Intel ship in one year? How many of those contracts are up for renewal during 2018, and how many are multi-year, and would require re-negotiating? What does that mean for the number of a) contracts, and b) chips, that AMD would need to close.

Intel is in many things slow. It's an unwieldy beast. It's made many strategy and execution errors over the years. These are much easier to see in hindsight, when you're looking into the past with 20-20 vision.

 

[raghu78]

What are you smoking, crack?

Simple thought experiment for you. How many chips in each class does Intel ship in one year? How many of those contracts are up for renewal during 2018, and how many are multi-year, and would require re-negotiating? What does that mean for the number of a) contracts, and b) chips, that AMD would need to close.

Intel is in many things slow. It's an unwieldy beast. It's made many strategy and execution errors over the years. These are much easier to see in hindsight, when you're looking into the past with 20-20 vision.

Lets check back a year from now on how AMD fares in market share in each segment.

 

[benb]

Intel reported 34% jump in 3Q profits, primarily due to strong growth in its data-centric businesses. Even though the PC segment, which brings in lion's share of total revenue, revenue was flat, it was able to post revenues of $16Bil, which is up 2%. 3Q results show that $INTC's transition from PC market to data-centric business is bearing fruit.

3Q data-centric businesses grew double digits (IoT revenues jumped 23%). As the PC shipments doesn't show any sign of improvement, its PC segment growth remains flat. It now expects to see high growth in IoT, AI, autonomous driving and cloud markets. $INTC expects FY17 EPS to be $3.25 and revenue of $62Bil, up $700MM from prior forecast.

Will Intel be able to take on AMD and still make sure it is maintaining its market share?

Intel's 3Q Highlights

Intel remains the dominant R&D spender in semiconductors, which helps (but does not guarantee) their economic moat.

I think their ownership of X86 is a significant moat. Qualcomm can't move into Intel's core market without a license, and Intel probably wants royalties on the device value (turnabout is such a bitch Qualcomm). On a $50,000 server, that would be quite a royalty. Eventually, Qualcomm is going to have to pay, just like AMD pays.

Speaking of pivots: IM Flash is now producing their very competitive 3D NAND (64 layer) and Intel will reap 49% of that bonanza. Memory is the new logic.

Management quality: Intel CEO isn't facing 5 years in jail. Intel doesn't face the retirement of a 30+ year leader who is basically irreplaceable. Intel isn't being swarmed by anti-trust judgments. And Intel isn't owned by a frontier market sovereign wealth fund. When you consider the other horses in this race, Intel management doesn't seem so bad.

 

[Daniel Nenni]

Intel remains the dominant R&D spender in semiconductors, which helps (but does not guarantee) their economic moat.

I think their ownership of X86 is a significant moat. Qualcomm can't move into Intel's core market without a license, and Intel probably wants royalties on the device value (turnabout is such a bitch Qualcomm). On a $50,000 server, that would be quite a royalty. Eventually, Qualcomm is going to have to pay, just like AMD pays.

Speaking of pivots: IM Flash is now producing their very competitive 3D NAND (64 layer) and Intel will reap 49% of that bonanza. Memory is the new logic.

Management quality: Intel CEO isn't facing 5 years in jail. Intel doesn't face the retirement of a 30+ year leader who is basically irreplaceable. Intel isn't being swarmed by anti-trust judgments. And Intel isn't owned by a frontier market sovereign wealth fund. When you consider the other horses in this race, Intel management doesn't seem so bad.

Interesting note from a friend:

ARM saw net sales rise to JPY46.6 billion from JPY14.4 billion in Q2 2016. The gain was due mostly to the 2016 results reflecting only the period from 6 September to 30 September, while 2017 results covered the full three-month period.

In 2016, ARM/SoftBank has made: 14.4B in 24 days

In 2017, it was 46.6B in 90 days

But, in 2016 it was 14.4/24 =0.6B per day, or 54B in Q2

46.6B is 15% less than 54B !!

The trend is bad…