Just finished "Only the Paranoid Survive", and have a few thoughts about current day Intel

[DanX]

Moore's Law truly is dead when it comes to wafer price.

At least not today.
I am in bitcoin .
The cost of transistor is cheaper for every new generation.
OK , maybe not half price.

 

[BruceA]

The strategy is AI which is obvious enough at this point.

The competitive moat and barriers in AI are huge as are the ones in Foundry for Intel.

Also in datacenter as well as client AMD is more than a formidable competitor. This IMs the AMD of a decade ago. If anything AMD is the leader.

Intel is second and actually not a close second except in x86 and it isn’t like they have cash and scale to catch up.

Actually Intel needs an unanticipated or unexpected thing to enable a catch up. AMD to bungle a design cycle, TSMC to do an Intel 10nm disaster. Catching anyone in AI is not going to happen unless the whole market disappears.

Pretty ugly for Intel from any and every angle

 

[BruceA]

Here's a view of how Intel was thinking about their successes in new markets across three decades. But they missed the next couple of huge market opportunities - Mobile and AI.

View attachment 3316

This is why they were so blinded and failed and missed every opportunity that presented in the last twenty years

 

[XYang2023]

The competitive moat and barriers in AI are huge as are the ones in Foundry for Intel.

Also in datacenter as well as client AMD is more than a formidable competitor. This IMs the AMD of a decade ago. If anything AMD is the leader.

Intel is second and actually not a close second except in x86 and it isn’t like they have cash and scale to catch up.

Actually Intel needs an unanticipated or unexpected thing to enable a catch up. AMD to bungle a design cycle, TSMC to do an Intel 10nm disaster. Catching anyone in AI is not going to happen unless the whole market disappears.

Pretty ugly for Intel from any and every angle

If AMD can catchup with Nvidia, it just shows that there is no moat in this space. Everyone can compete. Plus GPU is not ideal for inference anyways.

Intel should focus on AI front and centre simply because this space is large enough to invest in it and get a reward.

For Intel, the objective is first to turn around financially.

 

[XYang2023]

You cannot hide the fact that the current GPU based AI is very inefficient.

The road to commercial success for neuromorphic technologies - Nature Communications

Neuromorphic processors for commercial success require facing challenges in methods for programming neuromorphic applications and deployment at scale. Here, the authors discuss the pathways towards widespread consumer adoption of neuromorphic technology in relation to academia and industry.

www.nature.com

 

[BruceA]

If AMD can catchup with Nvidia, it just shows that there is no moat in this space. Everyone can compete. Plus GPU is not ideal for inference anyways.

Intel should focus on AI front and centre simply because this space is large enough to invest in it and get a reward.

For Intel, the objective is first to turn around financially.

And has caught up with Nvidia

Let’s check I. What 6months or 6 years ?

 

[XYang2023]

And has caught up with Nvidia

Let’s check I. What 6months or 6 years ?

Not just AMD but also Amazon:

https://www.cnbc.com/2025/06/17/aws-chips-nvidia-ai.html