(Update: laptop reviews) Not a bulldozing April fools joke - Intel new mobile processors announced

[mr moose]

17 minutes ago, TechyBen said:

I'd argue they are not. The position of AMD forced their hand in looking to reduce the cost of the chip, increase the production speed and thus improve the returns. Which resulted in the chiplet design.

That's what they were trying to do with 10nm.

Quote

Intel not having to worry, let the opportunity for them to get the chiplet design first (as a workable product, no idea on patents).

Ii don't understand what you are trying to say here.   Intel have played with chiplet and stacking before, their R+D is always trying to do new things.

https://www.anandtech.com/show/14211/intels-interconnected-future-chipslets-emib-foveros

 

EDIT: just in case you missed it, that article is a year old talking about existing chiplet products Intel have and stuff they have been working on for 2 years prior.   It's very likely Intel have been working on chiplet designs for many more years than that.

 

Quote

We would thus be in a very different position performance wise, perhaps with Intel having lower overall clocks, but higher core count. Not much of a difference, but the landscape and product bins would be different, if AMD had not had their hand forced.

 

This idea that Intel were caught off guard or were lazy, were resting on their laurels or any of those things is false.  They are a multinational company that has been working just as hard as any other to remain in front and retain a competitive edge.  The fact they failed does not mean they weren't working at it, a quick search of Intel's R+D  over the last 20 years should put all that to bed.  The whole time AMD had nothing on the table because BD was crap, Intel kept releasing new gen after new gen. Look at consumer demand and consumer software requirements over the last 10 years and look at Intel's offerings and you'll see why they kept selling new stuff even when the PC market took a massive shit dive. 

[RejZoR]

I'm pretty sure they were just being lazy and sleeping on laurels of past victories, downplaying AMD's capability to release anything worthwhile. It's been what, 3 years now since Ryzen launch and they are still shitting their pants, otherwise they wouldn't be re-releasing such sad refreshes.

 

I'm well aware they are working on it NOW, but that makes them very late to the game and for a multinational multibillion company that is always on top of things makes them look very unprepared... Coz if they were really prepared, they'd launch their new thing when AMD released first generation of Ryzen and crack them down before they could roll anything anywhere. Yet it's AMD doing that. They just relentlessly push out insane hardware. If first generation of Ryzen lacked a bit on highest clocks front, this 3rd generation not only matches Intel but sweeps the floor with it. Especially on the highest end with insane core counts and actual IPC gains which were mostly done by upping core clocks on Intel side. And still does.

[CephDigital]

3 hours ago, yaboistar said:

the only impressive thing here is 5.3GHz conditional boost on a mobile CPU. in every other respect, no thanks.

 

my 8750H is a good enough CPU and for the forseeable future if i need a new laptop i'll be getting one of the 4xxx series ryzens

Same here, I see no reason to upgrade my 8750H and RTX 2070 unless there's a massive performance boost to be had. A 4000 series Ryzen + RTX 2080 SUPER would be a nice upgrade though but unlikely. Let's just hope more laptop makers put these and future Ryzen CPUs in high end laptops. The lack of Thunderbolt 3 should be remedied with USB4 (even then, I only use my Thunderbolt port as a video out port)

 

Next laptop is looking to be 4000 or 5000 series Ryzen + RTX 3000

[5x5]

1 hour ago, TheUberMedic said:

Same here, I see no reason to upgrade my 8750H and RTX 2070 unless there's a massive performance boost to be had. A 4000 series Ryzen + RTX 2080 SUPER would be a nice upgrade though but unlikely. Let's just hope more laptop makers put these and future Ryzen CPUs in high end laptops. The lack of Thunderbolt 3 should be remedied with USB4 (even then, I only use my Thunderbolt port as a video out port)

 

Next laptop is looking to be 4000 or 5000 series Ryzen + RTX 3000

Thunderbolt is no longer ties to intel so Ryzen laptops with it are a possibility

[williamcll]

Update: the reviews are starting to come in.

[TechyBen]

10 hours ago, mr moose said:

That's what they were trying to do with 10nm.

Ii don't understand what you are trying to say here.   Intel have played with chiplet and stacking before, their R+D is always trying to do new things.

https://www.anandtech.com/show/14211/intels-interconnected-future-chipslets-emib-foveros

 

EDIT: just in case you missed it, that article is a year old talking about existing chiplet products Intel have and stuff they have been working on for 2 years prior.   It's very likely Intel have been working on chiplet designs for many more years than that.

 

 

This idea that Intel were caught off guard or were lazy, were resting on their laurels or any of those things is false.  They are a multinational company that has been working just as hard as any other to remain in front and retain a competitive edge.  The fact they failed does not mean they weren't working at it, a quick search of Intel's R+D  over the last 20 years should put all that to bed.  The whole time AMD had nothing on the table because BD was crap, Intel kept releasing new gen after new gen. Look at consumer demand and consumer software requirements over the last 10 years and look at Intel's offerings and you'll see why they kept selling new stuff even when the PC market took a massive shit dive. 

Yes. And AMD managed to leapfrog them. Was that by magic? Or by putting more effort into it? AMD put effort into dead and loss leading products which gave them the expertise to roll out Ryzen. Intel instead are still flogging a dead horse, hoping it won't die. They dabbled in other progress, but have, as far as I know, never actually released it. 10nm exception, they needed much more throughput on that!

[Bombastinator]

1 minute ago, TechyBen said:

Yes. And AMD managed to leapfrog them. Was that by magic? Or by putting more effort into it? AMD put effort into dead and loss leading products which gave them the expertise to roll out Ryzen. Intel instead are still flogging a dead horse, hoping it won't die. They dabbled in other progress, but have, as far as I know, never actually released it.

My suspicion is luck.  It was tsmc that came up with 7nm, not AMD.  You do a 14nm ryzen chip you got a ryzen1.  One wonders what intel might accomplish if it went tsmc too.  Their 14nm is a lot better than AMD 14nm

[TechyBen]

13 minutes ago, Bombastinator said:

My suspicion is luck.  It was tsmc that came up with 7nm, not AMD.  You do a 14nm ryzen chip you got a ryzen1.  One wonders what intel might accomplish if it went tsmc too.  Their 14nm is a lot better than AMD 14nm

? Is that luck or trade deals? What stops Intel asking/paying for a 7nm chip from TSMC?

 

As said, Intel have 10nm. There reliability and cost is through the roof though. So is that also down to luck with AMD, or working hard at the craft of 10-7nm? (Either in getting the contracts with the fabs, or with the design specs?).

 

Is it that AMD won a trade deal and has a monopoly on fabs? IIRC Samsung and Apple also get chips fabbed for phones at 10nm and below, right?

[Bombastinator]

1 minute ago, TechyBen said:

? Is that luck or trade deals? What stops Intel asking/paying for a 7nm chip from TSMC?

Intel has fabs.  I believe the quote was “real men have fabs”. I’m glad as heck they have fabs, honestly.  There’s very little manufacturing left in the US.  I’ve wondered a bit about what makes TSMC and AMD able to make 7nm chips when Intel can’t.  They’re missing something.  It might be in chip design, it might be in fab design, it might be someplace even weirder.  Iirc there was a blockage in chip design for a while because clean rooms couldn’t be made clean enough.  The solution was incredibly low tech.  They put a sticky pad on the floor outside the clean room that cleaned the bottoms of the shoes of workers.   It could be something wildly weird like that.  There’s something though.  Physics behaves the same way in Taipei as it does in California.  Maybe it’s the water?  Who knows? (Tsmc probably does winless they don’t)

[TechyBen]

3 minutes ago, Bombastinator said:

Intel has fabs.  I believe the quote was “real men have fabs”. I’m glad as heck they have fabs, honestly.  There’s very little manufacturing left in the US.  I’ve wondered a bit about what makes TSMC and AMD able to make 7nm chips when Intel can’t.  They’re missing something.  It might be in chip design, it might be in fab design, it might be someplace even weirder.  Iirc there was a blockage in chip design for a while because clean rooms couldn’t be made clean enough.  The solution was incredibly low tech.  They put a sticky pad on the floor outside the clean room that cleaned the bottoms of the shoes of workers.   It could be something wildly weird like that.  There’s something though.  Physics behaves the same way in Taipei as it does in California.  Maybe it’s the water?  Who knows? (Tsmc probably does winless they don’t)

But that's the thing. Did Intel put more money, people and time into progressing the design of chips than AMD and still lose? Possibly. But watching how businesses and meetings go on the little things (seeing the CEO of the bank I worked for, the shop I worked for etc in meetings/conferences) they had over the top confidence in "plan A" even if plan A was driving off a cliff. From Intel's monolithic approach, XE and the "more GHZ" options, I'd put them at a similar route, and it was a route that did not pay off.

 

Granted, AMD only really had the option of chiplets I guess. Perhaps they could have tried something else (Like Intel is doing on cheaper secondary chips using older tech while keeping the other fabs for better CPUs), but as they were losing the GHZ war and far behind on IPC, they had to do something, right?

 

Smaller companies, or even larger ones with the right structure, can adapt and change better. Those following the "company line" end up like Boeing, or Enron.

[mr moose]

1 hour ago, TechyBen said:

Yes. And AMD managed to leapfrog them. Was that by magic? Or by putting more effort into it? AMD put effort into dead and loss leading products which gave them the expertise to roll out Ryzen.

Intel AMD are in front for two reasons, 1, they built a decent product and 2 10nm failed for Intel.  If we are honest we don;t know and will never be able to say how critical the 10nm failure was in that success.

 

Quote

Intel instead are still flogging a dead horse, hoping it won't die.

They are still on 14 because 10nm failed, they are not choosing to flog 14nm. not by any stretch of the imagination.

Quote

 

They dabbled in other progress, but have, as far as I know, never actually released it. 10nm exception, they needed much more throughput on that!

 

Dabbled?  I wouldn't consider 4 years (probably more) of R+D and product development dabbling.  Why would you make Intel's biggest bock to a new architecture/process an "exception"?

 

The argument was quite clear that Intel were doing nothing, I have shown they were not doing nothing, in fact the they have been doing a lot more than just the two examples i have given, they have also been investing heavily in AI, stacking, HBM integration into CPU's.  The simple reality is that 10nm failed and that's why they are on so many reiterations of 14nm products, not because they were lazy.

[Bombastinator]

26 minutes ago, TechyBen said:

But that's the thing. Did Intel put more money, people and time into progressing the design of chips than AMD and still lose? Possibly. But watching how businesses and meetings go on the little things (seeing the CEO of the bank I worked for, the shop I worked for etc in meetings/conferences) they had over the top confidence in "plan A" even if plan A was driving off a cliff. From Intel's monolithic approach, XE and the "more GHZ" options, I'd put them at a similar route, and it was a route that did not pay off.

 

Granted, AMD only really had the option of chiplets I guess. Perhaps they could have tried something else (Like Intel is doing on cheaper secondary chips using older tech while keeping the other fabs for better CPUs), but as they were losing the GHZ war and far behind on IPC, they had to do something, right?

 

Smaller companies, or even larger ones with the right structure, can adapt and change better. Those following the "company line" end up like Boeing, or Enron.

That would qualify as “something”.  I don’t think intel is going for ghz because they want to.  They’ve been banging their heads on nanometer reduction for ten years and got nowhere.  AMD programming somehow runs 10% faster than intel on intel hardware.  Maybe 7nm isn’t what is making AMD chips fast after all.  Maybe it’s something else.  I have no idea.  Maybe something was passed over.  Maybe AMD is getting faster chips by playing looser with its code than it should.  Intel chips are still better than AMD chips at the enthusiast level, they just cost more.  An intel 6/12 whales on an AMD 6/12.  The thing is they cost a hundred dollars more.

[justpoet]

On 4/1/2020 at 3:12 PM, williamcll said:

Update: reviews are now coming in:

None of the 3 you've posted are reviews at all.  Two are advertisements by the laptop company, and the other is "I have a preproduction model as payment to advertise this, can't talk about performance, and here are some specs".

[xAcid9]

Quote

It does mean that in order to hit 5.3 GHz, the Core i9 is by default allowed to take 135 W across two cores, or 67.5 W per core. Even at 60W per core, you're looking at 50A of current per core... in a laptop.

https://www.anandtech.com/show/15687/intel-details-10th-gen-comet-lake-h-for-45-w-notebooks-up-to-53-ghz

 

[TechyBen]

13 hours ago, mr moose said:

Intel AMD are in front for two reasons, 1, they built a decent product and 2 10nm failed for Intel.  If we are honest we don;t know and will never be able to say how critical the 10nm failure was in that success.

 

They are still on 14 because 10nm failed, they are not choosing to flog 14nm. not by any stretch of the imagination.

 

Dabbled?  I wouldn't consider 4 years (probably more) of R+D and product development dabbling.  Why would you make Intel's biggest bock to a new architecture/process an "exception"?

 

The argument was quite clear that Intel were doing nothing, I have shown they were not doing nothing, in fact the they have been doing a lot more than just the two examples i have given, they have also been investing heavily in AI, stacking, HBM integration into CPU's.  The simple reality is that 10nm failed and that's why they are on so many reiterations of 14nm products, not because they were lazy.

I agree, I could see stacking letting them keep that top performance power.

[5x5]

@williamcll soooooo, yeahhhhh