Google brings on former Apple chip designer

[captain_to_fire]

Source: LinkedIn via Android Authority

 

Quote

 

Manu Gulati announced the move on his LinkedIn profile, though it likely won’t be the last such announcement — The Verge observed that Google has several open job postings that deal with chip development. When reached for comment, mum was the word for Google and Apple, though Gulati’s move is one that is a big one.

 

After all, Gulati was a key part of Apple’s efforts to build custom chipsets for its iPhones and iPads. The company started using its custom chips in 2010 for the original iPad and iPhone 4, both of which were powered by the A4. Since then, Apple’s custom chips have made their way inside of the Apple Watch, iPod Touch, Apple TV, and even the diminutive AirPods.

 

By comparison, whether it was a Nexus or Pixel device, Google has relied on Qualcomm for its chip needs. It makes sense, seeing how Qualcomm has been the go-to chip company for Android manufacturers for quite some time. If Google wants its Pixel smartphones to compete with the iPhone, however, it will have to also compete at the chipset level.

 

From Android Authority (Why are Apple’s chips faster than Qualcomm’s? – Gary explains):

Quote

What is different about Apple’s CPU cores?

 

There are several key things to recognize about Apple’s CPU cores.

 

First, Apple had a head-start over just about everyone when it comes to 64-bit ARM based CPUs. Although ARM itself announced the Cortex-A57 back in October 2012, the proposed timeline was that ARM’s partners would ship the first processors during 2014. But Apple had a 64-bit ARM CPU in devices during 2013. The company has since managed to capitalize on that early lead and has produced a new CPU core design every year.

 

Second, Apple’s SoC efforts are tightly coupled to its handset releases. Designing a high performance mobile CPU is hard. It is hard for Apple; for ARM; for Qualcomm; for everyone. Because it is hard, it takes a long time. The Cortex-A57 was announced in October 2012, but it didn’t appear in a smartphone until April 2014. That is a long lead time. That lead time is changing.

 

Third, Apple’s CPUs are big and in this game, big means expensive. According to a 2016 report by the Linley Group, the Hurricane cores in the Apple A10  are “about twice the size of other high-end mobile CPUs”. Even the smaller Zephyr cores are much larger than their low-power counterparts, “nearly twice as large as Cortex-A53.” The key here is that Apple sells smartphones, not chips. As a result, it can afford to make the SoCs more expensive and recoup the money in other places, including the final retail price.

 

ARM and Qualcomm, however, are in the chip selling business. ARM does the CPU core design for Qualcomm (and others like MediaTek) and Qualcomm designs the chips, which it in turn sells to handset makers like Samsung, Sony, LG, etc. ARM needs to make a profit. Qualcomm needs to make a profit. All the OEMs need to make profits. The practical result is that Qualcomm can’t afford to make overly expensive processors or OEMs will start looking elsewhere.

 

Fourth, Apple’s CPUs have big caches. Silicon costs money and for some chip makers their profit margin can be found in just 0.5 mm² of silicon saved. Like the third point above, Apple is able make bigger chips (in terms of silicon costs) and that includes large caches.

 

Fifth, and finally, Apple’s plan of making processors with wide pipelines at (initially) lower clock speeds has come to fruition. In very broad terms, SoC makers can either make a CPU core with a narrow pipe, but run that pipe at high clock frequencies; or use a wider pipe, but at lower clock speeds. Like a real world water pipe, you can either pump water at high pressure through a narrower pipe or at lower pressure through a wider pipe. In both cases you can theoretically achieve the same throughput. ARM falls squarely in the narrow pipeline camp, while Apple is in the wider pipeline camp. The Cortex-A75 can be clocked at a max frequency of 3 GHz on 10 nm, where as the Apple A10 is clocked at 2.34 GHz. We are still waiting to find out the clock speed of the A11.

 

This doesn't surprise me. Google is aware that one of the reasons why Android is fragmented is because of Qualcomm. Qualcomm only releases drivers up to two years and after that, they're done and move to the new version of Snapdragon and on top of that, Google is aware that at the moment, Qualcomm's Snapdragon SoC is very good but not on par with Apple's own A series chips. I remember back in 2013 when the die hard Android fans are dismissing Apple because they still use dual core SoCs (typical Apple hater) but Apple surprised everyone with the iPhone 5s's A7 chip which is the first 64-bit ARM SoC for consumers whereas Qualcomm simply doubled their core count with the Snapdragon 600 (previously dual core Snapdragon S4 Pro).

 

Geekbench resuslts: Samsung Galaxy S4 (32-bit Qualcomm Snapdragon 600 quad core) vs iPhone 5s (64-bit A7 chip dual core) - 2013

Spoiler

Geekbench Results: Google Pixel 2XL (Qualcomm Snapdragon 835) vs iPhone X (A11 Bionic: two high performance cores + four power efficient cores) - 2017

Spoiler

 

*I know there are valid reasons to dismiss Geekbench results but it's probably when comparing an x86 CPU vs an ARM SoC only.

 

Google has been taken many pages from Apple's playbook and having their own in-house designed SoCs is one of them. Android 8.1 update enabled the Pixel Visual Core SoC in the Pixel 2/2XL which allows for better local machine learning while taking pictures and video and it shows as the Pixel 2/2Xl can do bokeh without relying on a secondary sensor like the iPhone X/8+ and the Samsung Galaxy Note 8 aside from the fact that the Pixel 2XL scored higher by a few points than the iPhone X in DxO Mark.

Since Google is building out their own SoCs and probably ditch Qualcomm, it raises a few questions:

1. Would this mean that the Pixel line will no longer be a reference hardware to Android OEMs and will become a direct competitor not only to Apple but also to the likes of Android OEMs like Samsung, etc?
2. Will Google allow licensing of their own in-house SoC to other OEMs? (Unlikely just like Apple but who knows?)
3. How will Qualcomm respond to Google using their own SoCs? Will they finally release drivers for more than two years?
4. Will Qualcomm finally build their own smartphone?

So basically, Google is kinda like Intel by hiring an engineer from their competitor. I for one, thinks this is a good move from Google. As Gary from Android Authority has said, Qualcomm back in 2013 behaved lazily like Intel. 

 

Edited December 26, 2017 by hey_yo_
Edited the core count

[ScratchCat]

Obligatory but can it run crysis?

 

Finally we may see some innovation in terms of IPC for the android camp, the last few launches have been uninteresting at best. The supposed A53 replacement A55 has not been seen yet, offers only a 10-20 improvement for 10-20 more power resulting in a power efficiency increase which is negligible in most cases. 

 

These cores (if they are in fact cores) will probably not be seen until 2020, enough time to give Apple a health lead on whatever will be released as a result of this.

[Trixanity]

Correction: A11 has two high performance cores and four (not six) efficiency cores.

 

Google might make an application processor but let's not get ahead of ourselves. It takes years, tons of money and tons of engineers to get anywhere especially if starting from scratch.

 

Remember Apple didn't just pick up an engineer here and there whilst designing their chip in-house. They bought entire companies with lots of talent to jump start their endeavor.

 

Google needs more than a couple of guys to do this.

 

Google might just be making more specialized hardware like they do now with the Visual Core and their tensor cores.

[captain_to_fire]

38 minutes ago, ScratchCat said:

Obligatory but can it run crysis?

“Can Windows 10S run Crysis?” ?

 

38 minutes ago, ScratchCat said:

Finally we may see some innovation in terms of IPC for the android camp, the last few launches have been uninteresting at best. The supposed A53 replacement A55 has not been seen yet, offers only a 10-20 improvement for 10-20 more power resulting in a power efficiency increase which is negligible in most cases. 

 

These cores (if they are in fact cores) will probably not be seen until 2020, enough time to give Apple a health lead on whatever will be released as a result of this.

Basically, Qualcomm is kinda like Intel...

[Taf the Ghost]

Confirmation of Google getting into the SOC market is interesting. Though this could still be embedded devices and not mobile.

[captain_to_fire]

7 minutes ago, Taf the Ghost said:

Confirmation of Google getting into the SOC market is interesting. Though this could still be embedded devices and not mobile.

There’s nothing wrong with hoping that it’ll reach smartphones as the main SoC and not just their Iot devices 

[captain_to_fire]

49 minutes ago, Trixanity said:

Correction: A11 has two high performance cores and four (not six) efficiency cores.

Corrected it now. 

50 minutes ago, Trixanity said:

Google needs more than a couple of guys to do this.

Who knows? They already have a lot of staff including engineers and for something as big as Google, it shouldn’t be a problem. 

[Doobeedoo]

Ow this will be interesting. I wonder how much will it come close to A chips in future. 

[Matu20]

Should this mean lower priced Google phones at higher perf? Sign me up, love me some stock Android.

[mrchow19910319]

8 hours ago, hey_yo_ said:

 

Kudos to OP for such a informative post. 

[Not_Jony_Ive]

As long as Google doesn't hire me, android devices will never be as magical as iPhones. 

 

#Dealwithit.

 

 

[Commodus]

You know where Google getting into custom SoCs would particularly help?  Smartwatches.

 

Part of Apple's dominance of smartwatches stems from, well, actually giving a damn about smartwatch hardware.  Qualcomm's Snapdragon Wear chips feel like throwaways, and you can see that in the performance: the general rule is that Android Wear watches have mediocre battery life and still don't feel terribly fast.  Contrast that with the Apple Watch, where custom chips have seen giant speed boosts in two years and the improvements are obvious.  My original Apple Watch ended the day with 30-40 percent battery despite shorter exercises; my Series 3 can end the day with 66-70 percent even after a 60-minute workout.

 

If Google ever makes Pixel smartwatches, custom SoCs could go a long way toward making Android Wear worthwhile instead of the punchline it tends to be today.

[dizmo]

11 hours ago, ScratchCat said:

 

Finally we may see some innovation in terms of IPC for the android camp, the last few launches have been uninteresting at best. The supposed A53 replacement A55 has not been seen yet, offers only a 10-20 improvement for 10-20 more power resulting in a power efficiency increase which is negligible in most cases. 

 

These cores (if they are in fact cores) will probably not be seen until 2020, enough time to give Apple a health lead on whatever will be released as a result of this.

I think you mean Qualcomm? Android is an operating system and has nothing to do with IPC improvements.. 

 

I think Samsung's chips have been doing pretty well. 

5 hours ago, Matu20 said:

Should this mean lower priced Google phones at higher perf? Sign me up, love me some stock Android.

What makes you think it'll be lower cost? They'll still ant to make profit, and if people are used to paying a price you're not going to lower it, especially when you're competing with Apple. 

[Matu20]

18 minutes ago, dizmo said:

I think you mean Qualcomm? Android is an operating system and has nothing to do with IPC improvements.. 

 

I think Samsung's chips have been doing pretty well. 

What makes you think it'll be lower cost? They'll still ant to make profit, and if people are used to paying a price you're not going to lower it, especially when you're competing with Apple. 

Because in-house is usually cheaper alternative when mass produced.

[dizmo]

39 minutes ago, Matu20 said:

Because in-house is usually cheaper alternative when mass produced.

You won't see lower prices. You'll see higher profit margins. 

[Matu20]

1 hour ago, dizmo said:

You won't see lower prices. You'll see higher profit margins. 

Or that, yeah

[mrchow19910319]

all the android die hard "apple go fuck yourself" fanboies should avoid using google's hardware from now on. cos it is designed by a former apple employee. 

 

[mr moose]

2 hours ago, mrchow19910319 said:

all the android die hard "apple go fuck yourself" fanboies should avoid using google's hardware from now on. cos it is designed by a former apple employee. 

 

They should have been avoiding apple then because a lot of their staff used to work for HP.  Including years back the Woz himself. 

[mrchow19910319]

29 minutes ago, mr moose said:

They should have been avoiding apple then because a lot of their staff used to work for HP.  Including years back the Woz himself. 

yeah. sometimes i post sarcastic comments and people actually wrote an essay debate with me. 

lmao. 

"IM EXTREMELY BIASED TOWARDS MY OWN BELIEFS AND DEFINITELY NOT A HAPPY PERSON"

"ALL OF THE PEOPLE ON INTERNET NEED TO ALIGN THEIR BELIEFS WITH ME"

 

next time people go all the way out argue with me I will set up a separate paypal account and tell them "if you want to feel the sense of pride and accomplishment of winning an argument with me, please donate 9.99 opps sorry i mean 99.99 (this is for topic about apple) to this account. then I'll definitely more than happy to argue with you"

[mr moose]

29 minutes ago, mrchow19910319 said:

yeah. sometimes i post sarcastic comments and people actually wrote an essay debate with me. 

lmao. 

"IM EXTREMELY BIASED TOWARDS MY OWN BELIEFS AND DEFINITELY NOT A HAPPY PERSON"

"ALL OF THE PEOPLE ON INTERNET NEED TO ALIGN THEIR BELIEFS WITH ME"

 

next time people go all the way out argue with me I will set up a separate paypal account and tell them "if you want to feel the sense of pride and accomplishment of winning an argument with me, please donate 9.99 opps sorry i mean 99.99 (this is for topic about apple) to this account. then I'll definitely more than happy to argue with you"

1. I knew you were being sarcastic, that's why I added the laughing face to the end of my sentence.

2. You should consider ending your posts with /s  because poe's law is a thing.

 

3. I debate for free,  becasue I don't like people discovering that they paid money to find out they were wrong. 

[mrchow19910319]

47 minutes ago, mr moose said:

1. I knew you were being sarcastic, that's why I added the laughing face to the end of my sentence.

2. You should consider ending your posts with /s  because poe's law is a thing.

 

3. I debate for free,  becasue I don't like people discovering that they paid money to find out they were wrong. 

haha i knew you were't one of them of course. but those who did... i will add a /s though!