Was so lucky to attend this talk and meet Pat after. He's incredibly considerate and personable (and obviously very smart), definitely what Intel needs now after decades of poor leadership.
Pat is what intel needs, the right person at the right time. You can tell he's going to make important advancements and make a big difference. You already see a big positive shift happening and people at intel all have positive things to say about him. Also, that enthousiasm is very inspiring. Great talk!
I am glad I watched this presentation, very happy I was able to listen. I am very proud of Intel and have been for 40 years but today gave me reason to be proud.
After Pat left Intel, the Intel Larabee GPGPU and it's follow-up projects were put to low priority and even zero-budget. How a bad decision. If there is enough resource allocated to GPGPU, then Intel vs. Nvidia would have been another competition situation.
I'm excited. I'm currently a merchant mariner but I'm going back to school to get an electrical engineering degree. It would be a pleasure and an honor to work for Intel. Although very unlikely, I'd love to be a part of what they do and assist in that.
Really happy with the strategy and leadership Pat is brought to Intel. You can see his decades of engineering experience and talent at work. Pat is not only securing fab production for America, but all of the West, with fabs in Ireland, soon to be Germany and France. And Intel and many analysts expect that Intel will once again surpass TSMC and have the best nodes in 2024, as that's when Intel starts production of 20A and 18A, nodes which surpass TSMC's N3. Really excited for the future of Intel.
Those experts who have had years of fab experiences in the field, wouldn’t agree with you on that. The success of TSMC has to do with dedicated teamwork, innovation and business models. Strategies and CEO’s leadership are important but wouldn’t guarantee on future success. Intel is at least now 3-7 years in production behind
According to known sources, the reason the semiconductor industry in the United States began to move overseas was due to industrial accidents. At that time, the hazards of semiconductor manufacturing processes were not well understood. However, as workers started dying or becoming seriously ill, the United States allowed its allies to produce semiconductors. The processes today are no longer harmful. Many improvements have been made. Moreover, it has become a geopolitical asset surpassing oil. Therefore, the United States has started to build semiconductor factories again within its own borders. "I THINK"
he was right to focus on process because intel was on top for many years because of their process supremacy. once they started to faulter in that arena, so did their innovation. alder lake, for example, was supposed to happen in 2018. could you imagine what would have happened to the semiconductor industry had intel not failed to execute?
I did not understand what he meant in 33:15 when he said 'some ether of software.' I think he is implying that the software is inferior and that it really gives power to the semiconductor. Meanwhile, he mentioned that TSMC has more automation and AI technologies than we do, and we need to be more like them ! i think it's contradictory ?!!
I wish Pat all the best. Not only Intel itself is in a tough period, it seems like tough times are coming for everybody. I do wish Intel sorts out its problems. It is really an amazing company. Not sure whether it is possible to keep the fab business. It needs to be re-imagined. I'd say Intel needs to focus more on innovative products, not on chasing cheap and efficient mass production. But they know best what their situation is.
To be honest doing more in US/Europe is going to make them lagger in terms of cost. They had better have superior technologies and next gen technologies ready if they are doing this otherwise they are just going to be walloped by the rest. TSMC does a 5x better job on foundries than Intel any day running 100s of complicated products something Intel will "never" be able to do. They should just scale down and become a design house.
The CEO should address the issue of productivity. My point is there are too many employees but designing the right products compared to Nvidia and AMD.
Nvidia did, actually, get pretty dang lucky. They bet big on FP compute and it paid off in graphics, and happened to hit it big in crypto-hash and now generative AI models. They were hesitant to implement CUDA which is their development framework, but now without that they wouldn't have a fraction of the capability they have now. Any sane person would rather be lucky than good; right place, right time.
NVidia planned to support developers from day one. Listen to Jenson on the Acquired podcast. In regards to their GPU stack, Intel treats developers as pests. No luck. This was such a stupid thing to say.
Intel only has 1 small fab in Israel making legacy nodes, Fab28. The vast majority of their production is done in America and Ireland, and a future megafab in Germany. Intel is the safest chip producer due to the locations they choose, while TSMC is the riskiest.
Pat, I better see 2nm Intel CPU/GPU in 2024 - *AS YOU PROMISED* (Your words, NOT MINE) or else I'll be upgrading to Zen5 & RTX 5090 next year on their 3nm node! I want to see REALLY GREAT GPU from Intel. NO EXCUSES.
Intel is trying its part to advance the chips but has lost track of market share way back when they created mobile pc chips and failed . They can not compete with nvidia in AI , simple because they only focus on PC Chips which is declining market . Not an easy problem to solve
It is amazing that some people have no shame and self-reflection. I work with a few Intel engineers and they always amaze me! It is the culture BS! Even Intel hires only the MIT/Stanford would not save the company. I bet he would get fired in 5 years. He just could not keep what he promised (Intel roadmap). I regret to be in the semiconductor industry. Any sane MIT/Stanford engineer will not pick semiconductor as major. It is hard major with little financial reward compared to software or finance. I graduated at top 5 university with BSEE and MSEE. Many of my professors are from MIT/Stanford and other top universities. The FAB job is horrible. Many are overqualified, underpaid and work long hours. Those jobs are for foreigners (H1B). I was on H1B. Fortunately, I invest my money in the market and do well!
I put a tons of money to invest in Intel during Covid and this guy completely missed multiple rides consoles, fabs, mining, now AI they had so many chances to pull back into the game and they chose to rest and vest. They got beaten by NVDA and also letting AMD to leap over them. Such a shame
Now Semiconductors Taiwan manufacturer in USA mostly Taiwan have 72 years independence no war country you can very crystal clearly where have war in this world now
The *BIGGEST* problem for Intel is TRUST as a foundry service. Who in his right mind would give the blueprints away to a competitor? Intel is now competing with Apple, AMD, NVIDIA in all categories (Ultra Mobile/CPU/GPU/AI) and it only gets hotter from here on. It takes years and billions and billions of R&D for a state-of-the-art silicon to come to life. One needs to send extremely detailed "schematics" to a foundry service company to make that happen. It's not exactly stealing but YOUR COMPETITOR can EASILY "glean" secret source information while making your chip. You have to be CRAZY to do that. It'd be OKAY if a product is non-competitive (appliance for example) and it's not like they trust TSMC more. It's that TSMC DOESN'T MAKE ANYTHING other than chips. There's ZERO chance of TSMC competing with you in the future. What IF, Intel's chips "suddenly" got so good that it can compete with Apple/AMD/NVIDIA after making chips for them? Yea...it's too late. See you in court!!!!!!
There is truth to this, however the industry is unhappy with TSMC's dominance. Having 2 or 3 leading edge capable suppliers (intel and samsung) is something everybody wishes for, to reduce the dependence on TSMC
That's not the trust problem Intel has very much. Schematics aren't shared with foundries, layouts are. And figuring out RTL from that is a humungous task even at a few hundred transistors, let alone billions The bigger trust problem the foundry customers have is whether Intel foundry can deliver anything at all on time So far it is all hat no cattle
Intel doesnt have any good products, yet they spend tons of money on ads. Engineers have ppts, not yield or performance. VPs have age but not expertise. Intel has legacy but not future.
Intel should start producing risc v and every semiconductor. Get rid of directors and executives who are not engineers or scientists for a start. Hire only mbas with working engineering backgrounds. Get rid of accountants without any working engineering experience
@@touni97He's talking about the transistors, packaging technology, and process nodes. Not the individual chip designs. Intel's stated goal is to separate design from fabrication so other entities can contract with intel to build their chips.... for something posted on the MIT UA-cam channel there sure are a boat load of low IQ comments.
Was so lucky to attend this talk and meet Pat after. He's incredibly considerate and personable (and obviously very smart), definitely what Intel needs now after decades of poor leadership.
Agree
At least his two predecessors were worse than phony
Pat is what intel needs, the right person at the right time.
You can tell he's going to make important advancements and make a big difference.
You already see a big positive shift happening and people at intel all have positive things to say about him.
Also, that enthousiasm is very inspiring.
Great talk!
😂
comment aged like spoiled milk. RIP grandma.
Just bought AMD Ryzen 78003D for my game PC, use less power than Intel counter part. My first AMD CPU, really happy about it.
I am glad I watched this presentation, very happy I was able to listen. I am very proud of Intel and have been for 40 years but today gave me reason to be proud.
After Pat left Intel, the Intel Larabee GPGPU and it's follow-up projects were put to low priority and even zero-budget. How a bad decision. If there is enough resource allocated to GPGPU, then Intel vs. Nvidia would have been another competition situation.
Bean counter decision I'm sure. Short sighted
Coulda woulda shoulda
I'm excited. I'm currently a merchant mariner but I'm going back to school to get an electrical engineering degree. It would be a pleasure and an honor to work for Intel. Although very unlikely, I'd love to be a part of what they do and assist in that.
Really happy with the strategy and leadership Pat is brought to Intel. You can see his decades of engineering experience and talent at work. Pat is not only securing fab production for America, but all of the West, with fabs in Ireland, soon to be Germany and France. And Intel and many analysts expect that Intel will once again surpass TSMC and have the best nodes in 2024, as that's when Intel starts production of 20A and 18A, nodes which surpass TSMC's N3. Really excited for the future of Intel.
Pat is a showman. He is a pathological liar. Intel is buying all its chips from Tsmc.
Those experts who have had years of fab experiences in the field, wouldn’t agree with you on that. The success of TSMC has to do with dedicated teamwork, innovation and business models. Strategies and CEO’s leadership are important but wouldn’t guarantee on future success. Intel is at least now 3-7 years in production behind
Sounds like a wish based on a fart from what is frankly a market laggard.
You are having a pipe dream..
Paid sponsor of Intel
According to known sources, the reason the semiconductor industry in the United States began to move overseas was due to industrial accidents. At that time, the hazards of semiconductor manufacturing processes were not well understood. However, as workers started dying or becoming seriously ill, the United States allowed its allies to produce semiconductors. The processes today are no longer harmful. Many improvements have been made. Moreover, it has become a geopolitical asset surpassing oil. Therefore, the United States has started to build semiconductor factories again within its own borders. "I THINK"
he was right to focus on process because intel was on top for many years because of their process supremacy. once they started to faulter in that arena, so did their innovation. alder lake, for example, was supposed to happen in 2018. could you imagine what would have happened to the semiconductor industry had intel not failed to execute?
I did not understand what he meant in 33:15 when he said 'some ether of software.' I think he is implying that the software is inferior and that it really gives power to the semiconductor. Meanwhile, he mentioned that TSMC has more automation and AI technologies than we do, and we need to be more like them ! i think it's contradictory ?!!
Nice man. He sounds like Andrew Grove. All the best.♥️
He went through Andy’s school
훌륭한 강연이었다고 생각합니다
I wish Pat all the best. Not only Intel itself is in a tough period, it seems like tough times are coming for everybody. I do wish Intel sorts out its problems. It is really an amazing company. Not sure whether it is possible to keep the fab business. It needs to be re-imagined. I'd say Intel needs to focus more on innovative products, not on chasing cheap and efficient mass production. But they know best what their situation is.
To be honest doing more in US/Europe is going to make them lagger in terms of cost. They had better have superior technologies and next gen technologies ready if they are doing this otherwise they are just going to be walloped by the rest. TSMC does a 5x better job on foundries than Intel any day running 100s of complicated products something Intel will "never" be able to do. They should just scale down and become a design house.
Pat is the CEO and a good marketer, but he is a hard core engineer
That is the kind of guys that must stand at the top, pushing forward boarders
Pat, many people have underestimated him, but you'll see he will MIGA, Make Intel Great Again, sooner rather than later.
A smaller version of Trump? MIGA
Intel the will be the leading global foundry of the next decade.
The CEO should address the issue of productivity. My point is there are too many employees but designing the right products compared to Nvidia and AMD.
Pat: "Nvidia got incredibly lucky". Lol!! What an incredible defense to Intel's incompetence!!!
Nvidia did, actually, get pretty dang lucky. They bet big on FP compute and it paid off in graphics, and happened to hit it big in crypto-hash and now generative AI models. They were hesitant to implement CUDA which is their development framework, but now without that they wouldn't have a fraction of the capability they have now. Any sane person would rather be lucky than good; right place, right time.
@@Amaeyth Implementing CUDA to graphics was their plan all along. They deserve kudos for their vision. Why didnt intel think about GPUs before Nvidia?
NVidia planned to support developers from day one. Listen to Jenson on the Acquired podcast. In regards to their GPU stack, Intel treats developers as pests.
No luck. This was such a stupid thing to say.
@originalmianos And thats why Intel has become a company with zero credibility.
When did he say that?? I'm seeing that quote taken way out of context. That was Jensen admitting TO PAT HIMSELF that Nvidia got very lucky. BRUH.
Pat's a great speaker.
He is a pathological liar. Dont fall into his trap
@@ArpanKundu-k9x He's a great speaker, needs to say what's necessary to pump his company
Pat Gelsinger: Taiwan is a dangerous place.
Pat Gelsinger: No need to worry about our Israel Fabs.
Intel only has 1 small fab in Israel making legacy nodes, Fab28. The vast majority of their production is done in America and Ireland, and a future megafab in Germany. Intel is the safest chip producer due to the locations they choose, while TSMC is the riskiest.
@@__aceofspades pretty sure they have a 10 nanometer fab in Israel
And Meteor Lake is mostly made by TSMC. How dangerous is Taiwan?
amazing!!!!!!!!!!
WOW!
Pat, I better see 2nm Intel CPU/GPU in 2024 - *AS YOU PROMISED* (Your words, NOT MINE) or else I'll be upgrading to Zen5 & RTX 5090 next year on their 3nm node!
I want to see REALLY GREAT GPU from Intel. NO EXCUSES.
Not gonna happen. Their second gen discrete gpu is still going to have driver kinks to work out.
Intel farming their manufacturing to TSMC lmao
GPU ain't happening lmao
CPU... we'll see with Arrow Lake.
❤️❤️❤️❤️❤️
Intel is trying its part to advance the chips but has lost track of market share way back when they created mobile pc chips and failed . They can not compete with nvidia in AI , simple because they only focus on PC Chips which is declining market . Not an easy problem to solve
It is amazing that some people have no shame and self-reflection. I work with a few Intel engineers and they always amaze me! It is the culture BS!
Even Intel hires only the MIT/Stanford would not save the company. I bet he would get fired in 5 years. He just could not keep what he promised (Intel roadmap).
I regret to be in the semiconductor industry. Any sane MIT/Stanford engineer will not pick semiconductor as major. It is hard major with little financial reward compared to software or finance.
I graduated at top 5 university with BSEE and MSEE. Many of my professors are from MIT/Stanford and other top universities. The FAB job is horrible. Many are overqualified, underpaid and work long hours. Those jobs are for foreigners (H1B). I was on H1B. Fortunately, I invest my money in the market and do well!
Well said!! Unfortunately these topics are not discussed in the paid media!!
Pat, I felt, you shortsold the Taiwanese.
⭐
I put a tons of money to invest in Intel during Covid and this guy completely missed multiple rides consoles, fabs, mining, now AI they had so many chances to pull back into the game and they chose to rest and vest. They got beaten by NVDA and also letting AMD to leap over them. Such a shame
As an ex investor of Intel, he wasted my money and time. Very disappointed with him and entire Intel guys
Now Semiconductors Taiwan manufacturer in USA mostly Taiwan have 72 years independence no war country you can very crystal clearly where have war in this world now
ゲルシンガーマジでジーニアスやな😮
Intel is like an extinct dinosaur.
Pat is an enigma. He says all the right things, he knows so much, but can't make the right the decisions with his own knowledge.
The *BIGGEST* problem for Intel is TRUST as a foundry service. Who in his right mind would give the blueprints away to a competitor? Intel is now competing with Apple, AMD, NVIDIA in all categories (Ultra Mobile/CPU/GPU/AI) and it only gets hotter from here on. It takes years and billions and billions of R&D for a state-of-the-art silicon to come to life. One needs to send extremely detailed "schematics" to a foundry service company to make that happen. It's not exactly stealing but YOUR COMPETITOR can EASILY "glean" secret source information while making your chip. You have to be CRAZY to do that. It'd be OKAY if a product is non-competitive (appliance for example) and it's not like they trust TSMC more. It's that TSMC DOESN'T MAKE ANYTHING other than chips. There's ZERO chance of TSMC competing with you in the future. What IF, Intel's chips "suddenly" got so good that it can compete with Apple/AMD/NVIDIA after making chips for them? Yea...it's too late. See you in court!!!!!!
It's protected as IP. It's protected by laws.
There are zero options-unless you go to Asia. Intel is only game until TSMC have facilities in Arizona
Made in US - thats INTEL
There is truth to this, however the industry is unhappy with TSMC's dominance. Having 2 or 3 leading edge capable suppliers (intel and samsung) is something everybody wishes for, to reduce the dependence on TSMC
That's not the trust problem Intel has very much. Schematics aren't shared with foundries, layouts are. And figuring out RTL from that is a humungous task even at a few hundred transistors, let alone billions
The bigger trust problem the foundry customers have is whether Intel foundry can deliver anything at all on time
So far it is all hat no cattle
A really disappointingly arrogant style - that puts his sacking in a whole new light for me 😢
Intel doesnt have any good products, yet they spend tons of money on ads. Engineers have ppts, not yield or performance. VPs have age but not expertise. Intel has legacy but not future.
You're delusional Intel have the best cpus out there and good price arc GPUs
Your opinions contradict the constant upgrades done by large institutional investors on INTC stock. No wonder it’s almost doubled this year.
This is something tsmc pumper would say
@@thenavigator5889 I am not a TSMC pumper. I am telling truth bcoz I used to work there. Intel is rotten inside
That was before Pat took over..
Gelsinger is notorious for exaggeration in speaking but never produce real product. Just listen , don’t be serious.
Hi
Intel lost my respect when kept adding "+"s to skylake and reselling them every year. This guy is a 🤡
What an irrational comment. Pat wasn't even at the company when 14nm+++++++++++++++++++ was going on.
This guy can only make empty promises. Liar liar pants on fire!
West vs the East , this is what pat is saying . 🤣
@@Nvidia-Lover only large foundry left here is INTEL I believe
Hi hello:;
This guy doesn't stand a chance against Jensen, they need a software guy in there.
Hindu inside
X86 is obsolete. The only reason of x86 existence is legacy.
People have been saying that for 30+ years.. ironically back when it didn't have any legacy 💀💀
Intel should start producing risc v and every semiconductor. Get rid of directors and executives who are not engineers or scientists for a start. Hire only mbas with working engineering backgrounds. Get rid of accountants without any working engineering experience
🫡💪
Clown full of tricks!
bill gates long distance cousin
Pat is cringe as fuck. I wont be putting my money in Intel for a long time and my next cpu is a ryzen
Sees 10 years forward LMAO I'd be ashamed saying that in front of a room full of people that know it's clearly untrue about AI chipss....
@@touni97He's talking about the transistors, packaging technology, and process nodes. Not the individual chip designs. Intel's stated goal is to separate design from fabrication so other entities can contract with intel to build their chips.... for something posted on the MIT UA-cam channel there sure are a boat load of low IQ comments.
Pat is a clown!
Who buys Intel products?? No one 😂😂
bro got beef with nvidia
Intel is 3rd place behind Apple and amd. In the Olympics that would be good but this isn't that 😅
Intel stock seems like a high risk bet, Im all in!!🦾