I think that this is a necessary move from Google, to ultimately control its own silicon, but the guy they really wanted wasn't available. Still, I'm not sure that its necessary to have a Jim Keller at the current stage of ARM development.
I agree with Gatorguy that this wasn't about the pixel per se; just a necessary move to own an important piece of the hardware stack for any projects under development. I don't think the competition is necessary, but I would argue that 2 of 3 of the world's consumer operating systems are fully committed to ARM architecture development, something not lost on the x86/x64 world or the booming fab business.
In post 32 I tred to clarify something I wrote earlier, but chose poor wording again. I said this poaching "closes the gap" between Google and Apple with regard to chip development. I should have said "narrows the gap." Apple obviously still has a significant head start and this one hire won't put Google on equal footing, but it gets Google closer than they were before.
Do you seriously believe Google is "serious" about this venture, what with their 1million pixel sales over 8 months????
I think they're serious about it, but not just for smartphones. You're right that if was only about a Pixel processor it might not be terribly serious.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Also, has Google announced it's intention to develop custom SoCs? As others have rightly pointed out, just one high profile hiring does not make you a company capable of making custom SoCs. Isn't this (custom SoC development part, not the hiring part) still a rumor?
In post 32 I tred to clarify something I wrote earlier, but chose poor wording again. I said this poaching "closes the gap" between Google and Apple with regard to chip development. I should have said "narrows the gap." Apple obviously still has a significant head start and this one hire won't put Google on equal footing, but it gets Google closer than they were before.
Do you seriously believe Google is "serious" about this venture, what with their 1million pixel sales over 8 months????
I think they're serious about it, but not just for smartphones. You're right that if was only about a Pixel processor it might not be terribly serious.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Really tone deaf on your part. This isn't just about SoC's for one device and gosh, even I, an avowed Apple users and supporter from the era of the original Mac can comprehend that, yes, Google is in it for the long run.
We converse frequently here of the performance of ARM, and how, in some cases it outperforms Intel designs for specific applications. We speak of an ARM Mac like it's a sure thing, but it is as well easy to argue that Google will build a Chromebook followup at some point that will compete directly with a segment of Mac Books and Pros and Surface devices; Google already does that to some extent. More to the point, as Apple and others have demonstrated, their is a need for purpose built satellite processors for accessories like AirPods and Alexa devices, or VR headsets.
In post 32 I tred to clarify something I wrote earlier, but chose poor wording again. I said this poaching "closes the gap" between Google and Apple with regard to chip development. I should have said "narrows the gap." Apple obviously still has a significant head start and this one hire won't put Google on equal footing, but it gets Google closer than they were before.
Do you seriously believe Google is "serious" about this venture, what with their 1million pixel sales over 8 months????
I think they're serious about it, but not just for smartphones. You're right that if was only about a Pixel processor it might not be terribly serious.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Hmmm... Let's think... What use would Google have for processors and/or other chips? Servers? ...that's a biggie since Google buys more server chips than almost anyone on the planet including companies that actually sell servers.
Perhaps Smarthome devices? Automotive? VR/AR headsets and gear? Phones? Chromebooks? An upcoming line of devices that hasn't been publicly mentioned yet running a new OS?
FWIW Google has already designed their own chips, in service now. This looks like they plan to expand those efforts and believe Mr Gulati's expertise is needed.
In post 32 I tred to clarify something I wrote earlier, but chose poor wording again. I said this poaching "closes the gap" between Google and Apple with regard to chip development. I should have said "narrows the gap." Apple obviously still has a significant head start and this one hire won't put Google on equal footing, but it gets Google closer than they were before.
Do you seriously believe Google is "serious" about this venture, what with their 1million pixel sales over 8 months????
I think they're serious about it, but not just for smartphones. You're right that if was only about a Pixel processor it might not be terribly serious.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Really tone deaf on your part. This isn't just about SoC's for one device and gosh, even I, an avowed Apple users and supporter from the era of the original Mac can comprehend that, yes, Google is in it for the long run.
We converse frequently here of the performance of ARM, and how, in some cases it outperforms Intel designs for specific applications. We speak of an ARM Mac like it's a sure thing, but it is as well easy to argue that Google will build a Chromebook followup at some point that will compete directly with a segment of Mac Books and Pros and Surface devices; Google already does that to some extent. More to the point, as Apple and others have demonstrated, their is a need for purpose built satellite processors for accessories like AirPods and Alexa devices, or VR headsets.
Google is just being prudent,
Now, I get what you are saying. But there are so many other question marks. Apple has always been a hardware & software company, having expertise in both for decades!!! Google has been a software company (a damn good one at that), with MANY half hearted attempts at hardware, all ending up failures. Even the "Pixels" were touted as Google taking hardware seriously, guess where it ended up? Worser than the earlier Nexus phones which had better success due to right pricing. The natural question then arises - Do they take hardware seriously? Do they really have any expertise of whatsoever to be successful at making their own hardware? Wouldn't they be better of just making software and leaving the hardware development to their partners? If they really want to succeed at hardware, what would be the relationship with their hardware partners, whom Google needs for their 'ad' business?
In post 32 I tred to clarify something I wrote earlier, but chose poor wording again. I said this poaching "closes the gap" between Google and Apple with regard to chip development. I should have said "narrows the gap." Apple obviously still has a significant head start and this one hire won't put Google on equal footing, but it gets Google closer than they were before.
Do you seriously believe Google is "serious" about this venture, what with their 1million pixel sales over 8 months????
I think they're serious about it, but not just for smartphones. You're right that if was only about a Pixel processor it might not be terribly serious.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Hmmm... Let's think... What use would Google have for processors and/or other chips? Servers? ...that's a biggie since Google buys more server chips than almost anyone on the planet including companies that actually sell servers.
Perhaps Smarthome devices? Automotive? VR/AR headsets and gear? Phones? Chromebooks? An upcoming line of devices that hasn't been publicly mentioned yet running a new OS?
FWIW Google has already designed their own chips, in service now. This looks like Google plans to expand those efforts.
Thanks and Tmay posted something similar as well. In essence, Google trying to become an "Apple"??? I would say "good luck" with that.
In post 32 I tred to clarify something I wrote earlier, but chose poor wording again. I said this poaching "closes the gap" between Google and Apple with regard to chip development. I should have said "narrows the gap." Apple obviously still has a significant head start and this one hire won't put Google on equal footing, but it gets Google closer than they were before.
Do you seriously believe Google is "serious" about this venture, what with their 1million pixel sales over 8 months????
I think they're serious about it, but not just for smartphones. You're right that if was only about a Pixel processor it might not be terribly serious.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Hmmm... Let's think... What use would Google have for processors and/or other chips? Servers? ...that's a biggie since Google buys more server chips than almost anyone on the planet including companies that actually sell servers.
Perhaps Smarthome devices? Automotive? VR/AR headsets and gear? Phones? Chromebooks? An upcoming line of devices that hasn't been publicly mentioned yet running a new OS?
FWIW Google has already designed their own chips, in service now. This looks like Google plans to expand those efforts.
Thanks and Tmay posted something similar as well. In essence, Google trying to become an "Apple"??? I would say "good luck" with that.
LOL... Hardly, tho who wouldn't like to be a little Apple.
Spreading their wings a bit to rely a little less on services/advertising is just as smart as Apple spreading their wings a bit more to depend a tad less heavily on hardware and increase services revenue.
In post 32 I tred to clarify something I wrote earlier, but chose poor wording again. I said this poaching "closes the gap" between Google and Apple with regard to chip development. I should have said "narrows the gap." Apple obviously still has a significant head start and this one hire won't put Google on equal footing, but it gets Google closer than they were before.
Do you seriously believe Google is "serious" about this venture, what with their 1million pixel sales over 8 months????
I think they're serious about it, but not just for smartphones. You're right that if was only about a Pixel processor it might not be terribly serious.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Hmmm... Let's think... What use would Google have for processors and/or other chips? Servers? ...that's a biggie since Google buys more server chips than almost anyone on the planet including companies that actually sell servers.
Perhaps Smarthome devices? Automotive? VR/AR headsets and gear? Phones? Chromebooks? An upcoming line of devices that hasn't been publicly mentioned yet running a new OS?
FWIW Google has already designed their own chips, in service now. This looks like Google plans to expand those efforts.
Thanks and Tmay posted something similar as well. In essence, Google trying to become an "Apple"??? I would say "good luck" with that.
LOL... Hardly, tho who wouldn't like to be a little Apple.
Spreading their wings a bit to rely a little less on services/advertising is just as smart as Apple spreading their wings a bit more to depend a tad less heavily on hardware and increase services revenue.
Agreed on all points. But the evidence so far does not give any confidence in Google's hardware adventures. Only time will tell, I guess.
In post 32 I tred to clarify something I wrote earlier, but chose poor wording again. I said this poaching "closes the gap" between Google and Apple with regard to chip development. I should have said "narrows the gap." Apple obviously still has a significant head start and this one hire won't put Google on equal footing, but it gets Google closer than they were before.
Do you seriously believe Google is "serious" about this venture, what with their 1million pixel sales over 8 months????
I think they're serious about it, but not just for smartphones. You're right that if was only about a Pixel processor it might not be terribly serious.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Also, has Google announced it's intention to develop custom SoCs? As others have rightly pointed out, just one high profile hiring does not make you a company capable of making custom SoCs. Isn't this (custom SoC development part, not the hiring part) still a rumor?
Correction: Apple HAD a ten year lead on custom silicon. The man responsible just took all that knowledge and experience to Google. This brings Google a lot closer to being current, and slows Apple's development efforts.
If Manu is that important to Apple's chip development, I wonder why he would change companies, especially with Apple's new campus and the great stuff in the pipeline? Is he bored? Was it purely a monetary decision (Google offering him more?)?
In post 32 I tred to clarify something I wrote earlier, but chose poor wording again. I said this poaching "closes the gap" between Google and Apple with regard to chip development. I should have said "narrows the gap." Apple obviously still has a significant head start and this one hire won't put Google on equal footing, but it gets Google closer than they were before.
Do you seriously believe Google is "serious" about this venture, what with their 1million pixel sales over 8 months????
I think they're serious about it, but not just for smartphones. You're right that if was only about a Pixel processor it might not be terribly serious.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Also, has Google announced it's intention to develop custom SoCs? As others have rightly pointed out, just one high profile hiring does not make you a company capable of making custom SoCs. Isn't this (custom SoC development part, not the hiring part) still a rumor?
Chromebooks.
Chromebooks??? Seriously??? The ASP of Chromebooks is around $200, if I am not wrong. For that price range, you do NOT need a custom SoC. Even low-quality Mediatek SoCs would do, leave alone high quality SoCs from Qualcomm/Samsung/Huawei !!! And we are talking about one level above that, as a potential requirement so as to justify custom SoC development. Is Google thinking of Chromebooks being competitors to iPad Pros or MacBooks? That too 5 years down the line, in PREMIUM segment?
Correction: Apple HAD a ten year lead on custom silicon. The man responsible just took all that knowledge and experience to Google. This brings Google a lot closer to being current, and slows Apple's development efforts.
If Manu is that important to Apple's chip development, I wonder why he would change companies, especially with Apple's new campus and the great stuff in the pipeline? Is he bored? Was it purely a monetary decision (Google offering him more?)?
Apple's head of Siri Natural Language Understanding just left for another company too (Facebook). Money, new challenges, change of scenery... people change jobs for any number of reasons.
Correction: Apple HAD a ten year lead on custom silicon. The man responsible just took all that knowledge and experience to Google. This brings Google a lot closer to being current, and slows Apple's development efforts.
If Manu is that important to Apple's chip development, I wonder why he would change companies, especially with Apple's new campus and the great stuff in the pipeline? Is he bored? Was it purely a monetary decision (Google offering him more?)?
I find it hard to believe that Apple would be unable to match Google's offer if the guy was worth it, so I don't think money is the reason he moved.
Correction: Apple HAD a ten year lead on custom silicon. The man responsible just took all that knowledge and experience to Google. This brings Google a lot closer to being current, and slows Apple's development efforts.
If Manu is that important to Apple's chip development, I wonder why he would change companies, especially with Apple's new campus and the great stuff in the pipeline? Is he bored? Was it purely a monetary decision (Google offering him more?)?
I find it hard to believe that Apple would be unable to match Google's offer if the guy was worth it, so I don't think money is the reason he moved.
This largely depends on corporate culture. Many companies strive to have a stable salary scale and stick to it with perks and performance bonuses added to flesh out the deals in an attempt to avoid being held to ransom by their own brightest employees. The policy that no one irreplaceable is what keeps negotiations in the comfort zone.
I have a friend who is currently earning almost 2,500€ a day on a basic salary.
He is happy because he has job stability, family, roots etc and values his 'easy' life.
As a computer scientist he had a far more stimulating job in government but the salary wasn't its strong point. He then became independent and was doing contract work for years, setting his own prices and living all over the world. It sounds like fun but it is tough. Very tough on every level but he was able to set his own price. With no ties, it made sense.
Today, he says he still misses the government work in spite of the poor salary, working underground and living in a town where everybody else also worked for the same institution and nobody being able to talk about what they were doing.
However, he is now 'coasting' and basically having his cake and eating it but when he was younger he was driven by other, purely work related factors
So, this move could be for any amount of reasons but I also think it was not related to money. More likely someone gave him a challenge and a certain amount of freedom to meet it and he decided to accept it.
Yes, the opposite is also possible, perhaps he was feeling the heat at Apple and finding it too much and Google offered a less pressurised environment. I don't think thay.is the case here though.
Well here's a processor space that I hadn't realized Google was as serious about as they appear to be: Quantum computing. Google announced their own Quantum chip design is now available for researchers, putting them ahead of many others working in that field according to industry articles. https://research.googleblog.com/2018/03/a-preview-of-bristlecone-googles-new.html
Well here's a processor space that I hadn't realized Google was as serious about as they appear to be: Quantum computing. Google announced their own Quantum chip design is now available for researchers, putting them ahead of many others working in that field according to industry articles. https://research.googleblog.com/2018/03/a-preview-of-bristlecone-googles-new.html
They're all seeking dominance in AI, so presumably they think such developments will give them a leg up (...a neuron up?)
Comments
https://electrek.co/2016/01/28/tesla-jim-keller-apple-processor-architect-2/
I agree with Gatorguy that this wasn't about the pixel per se; just a necessary move to own an important piece of the hardware stack for any projects under development. I don't think the competition is necessary, but I would argue that 2 of 3 of the world's consumer operating systems are fully committed to ARM architecture development, something not lost on the x86/x64 world or the booming fab business.
Apart from smartphones, what else do you think Google is planning to use the custom SoCs for? Tablets? Anything else? It would take at least 2 to 3 years to first roll out your own SOC which would be inefficient at best and then to improve it and learn to make it efficient, it would take another 2 years. Is Google ready to play the long game, to what benefit? I doubt it.
Also, has Google announced it's intention to develop custom SoCs? As others have rightly pointed out, just one high profile hiring does not make you a company capable of making custom SoCs. Isn't this (custom SoC development part, not the hiring part) still a rumor?
We converse frequently here of the performance of ARM, and how, in some cases it outperforms Intel designs for specific applications. We speak of an ARM Mac like it's a sure thing, but it is as well easy to argue that Google will build a Chromebook followup at some point that will compete directly with a segment of Mac Books and Pros and Surface devices; Google already does that to some extent. More to the point, as Apple and others have demonstrated, their is a need for purpose built satellite processors for accessories like AirPods and Alexa devices, or VR headsets.
Google is just being prudent,
Servers?
...that's a biggie since Google buys more server chips than almost anyone on the planet including companies that actually sell servers.
Perhaps Smarthome devices?
Automotive?
VR/AR headsets and gear?
Phones?
Chromebooks?
An upcoming line of devices that hasn't been publicly mentioned yet running a new OS?
FWIW Google has already designed their own chips, in service now. This looks like they plan to expand those efforts and believe Mr Gulati's expertise is needed.
Now, I get what you are saying. But there are so many other question marks. Apple has always been a hardware & software company, having expertise in both for decades!!! Google has been a software company (a damn good one at that), with MANY half hearted attempts at hardware, all ending up failures. Even the "Pixels" were touted as Google taking hardware seriously, guess where it ended up? Worser than the earlier Nexus phones which had better success due to right pricing. The natural question then arises - Do they take hardware seriously? Do they really have any expertise of whatsoever to be successful at making their own hardware? Wouldn't they be better of just making software and leaving the hardware development to their partners? If they really want to succeed at hardware, what would be the relationship with their hardware partners, whom Google needs for their 'ad' business?
Thanks and Tmay posted something similar as well. In essence, Google trying to become an "Apple"??? I would say "good luck" with that.
Spreading their wings a bit to rely a little less on services/advertising is just as smart as Apple spreading their wings a bit more to depend a tad less heavily on hardware and increase services revenue.
Chromebooks??? Seriously??? The ASP of Chromebooks is around $200, if I am not wrong. For that price range, you do NOT need a custom SoC. Even low-quality Mediatek SoCs would do, leave alone high quality SoCs from Qualcomm/Samsung/Huawei !!! And we are talking about one level above that, as a potential requirement so as to justify custom SoC development. Is Google thinking of Chromebooks being competitors to iPad Pros or MacBooks? That too 5 years down the line, in PREMIUM segment?
I have a friend who is currently earning almost 2,500€ a day on a basic salary.
He is happy because he has job stability, family, roots etc and values his 'easy' life.
As a computer scientist he had a far more stimulating job in government but the salary wasn't its strong point. He then became independent and was doing contract work for years, setting his own prices and living all over the world. It sounds like fun but it is tough. Very tough on every level but he was able to set his own price. With no ties, it made sense.
Today, he says he still misses the government work in spite of the poor salary, working underground and living in a town where everybody else also worked for the same institution and nobody being able to talk about what they were doing.
However, he is now 'coasting' and basically having his cake and eating it but when he was younger he was driven by other, purely work related factors
So, this move could be for any amount of reasons but I also think it was not related to money. More likely someone gave him a challenge and a certain amount of freedom to meet it and he decided to accept it.
Yes, the opposite is also possible, perhaps he was feeling the heat at Apple and finding it too much and Google offered a less pressurised environment. I don't think thay.is the case here though.
https://research.googleblog.com/2018/03/a-preview-of-bristlecone-googles-new.html