4 ms·
I am old enough to remember when AMD was nowhere, and Intel was the only game in town for both consumer and server chips, laptop and desktop. I think it was 201
by acyou 2y ago
I am old enough to remember when AMD was nowhere, and Intel was the only game in town for both consumer and server chips, laptop and desktop. I think it was 2011.
Then 2013 rolled around and I built a desktop with an AMD processor, because it was dirt cheap. I remember people saying that they were excited and looking forward to the upcoming AMD architectures and roadmap. I couldn't understand how AMD could possibly have come from behind to beat Intel.
Realistically, you just need to come up with a good roadmap, get good people, get the resources, execute.
Roadmap is good, people - not so sure about this one, resources - seem a little low, execution - seems hampered by too many cooks. If Intel broke up into smaller companies and their new leading edge chip design group had 500 good people, I honestly think they would have a better shot. So I think we are seeing what makes the most sense, Intel must die and let the phoenixes rise from the ashes.
Fast forward to 2024/2025, and remember, anything is possible if AMD beat 2011 Intel.
- thijson 2y agoAMD doesn't have the burden of capital expense that a fab is. It did take several years for AMD to capitalize on the freedom this enables though. As far as I know Intel is the only company left that is vertically integrated. This might be the last domino to fall. Arrow Lake uses TSMC heavily. Maybe that's a sign of things to come. Andy Grove said only the paranoid survive. It didn't seem like Intel management was that paranoid. They didn't invent 64bit x86, had to license it from AMD. Spent so much money on stock buy backs, when it should have been funneled back into the fab, getting yields up on the next process node. It's easy for me to be an armchair quarterback though, I'm not aware of everything that happened behind the scenes. I saw D1X being built, and the future seemed bright.