Quote from: Vaidyanathan on February 27, 2020, 16:17:21
Guess we won't be able to answer anything about efficiency until we get our hands on an actual review sample. There's still a lot of time for that, though. I doubt leaks leading up to the launch will even remotely talk about the chip's efficiency leave alone the iGPU.

dont take it personally, i even forgot you are media guy. it is what it is, efficiency should be everywhere and as the main question, meanwhile im keep finding cpu/gpu reviews without it at all.

QuoteWhile at the moment it appears as though 8 cores would be a huge task on current 10nm yields, I am sure they will improve on that coz Tiger Lake-H is expected to arrive around end of Q1 next year. That surely cannot be a measly 4-core. But yeah, it will take time to trickle that into a 15W envelope.

really strange to see separate mobile chips for 15 and 45 watts.
(and on paper zen3 would destroy tiger lake-h)

Quote from: william blake on February 27, 2020, 07:34:22

Quote from: Vaidyanathan on February 27, 2020, 05:30:16
Yeah even if you play around with the cTDP, there's only so much you can extract from the iGPU. This still uses the UHD 630 so there's not much incentive on the GPU side anyways. It has to be paired with a capable Turing (or Ampere next year) part.

have you heard about new intel gpus(discrete and integrated) energy efficiency? Xe or some s***..im not. and for me this is outrageous, there is no more frequency of cores questions, these are secondary. we live in a world with almost infinite(for the user) scalability now, and the main question should be  always about efficiency. this determines how much productivity from socket of from tdp or from another limit you can obtain.
i dont understand media/tech guys, seems like they are all stuck the past. who cares about tiger lake igpu if there is no big improvement in efficiency. maybe better replace it with the cheap ancient pascal and half of the tiger's chip is a waste of silicon?
so far i'm very sceptical about the direction intel has chosen. they have no tdp room for 8 cores+big igpu, all mobile situation for them looks fucked up until 7nm. its just a present for them that amd is so small and unable to produce 3 times more.

Guess we won't be able to answer anything about efficiency until we get our hands on an actual review sample. There's still a lot of time for that, though. I doubt leaks leading up to the launch will even remotely talk about the chip's efficiency leave alone the iGPU.
While at the moment it appears as though 8 cores would be a huge task on current 10nm yields, I am sure they will improve on that coz Tiger Lake-H is expected to arrive around end of Q1 next year. That surely cannot be a measly 4-core. But yeah, it will take time to trickle that into a 15W envelope.

Quote from: Vaidyanathan on February 27, 2020, 05:30:16
Yeah even if you play around with the cTDP, there's only so much you can extract from the iGPU. This still uses the UHD 630 so there's not much incentive on the GPU side anyways. It has to be paired with a capable Turing (or Ampere next year) part.

have you heard about new intel gpus(discrete and integrated) energy efficiency? Xe or some s***..im not. and for me this is outrageous, there is no more frequency of cores questions, these are secondary. we live in a world with almost infinite(for the user) scalability now, and the main question should be  always about efficiency. this determines how much productivity from socket of from tdp or from another limit you can obtain.
i dont understand media/tech guys, seems like they are all stuck the past. who cares about tiger lake igpu if there is no big improvement in efficiency. maybe better replace it with the cheap ancient pascal and half of the tiger's chip is a waste of silicon?
so far i'm very sceptical about the direction intel has chosen. they have no tdp room for 8 cores+big igpu, all mobile situation for them looks fucked up until 7nm. its just a present for them that amd is so small and unable to produce 3 times more.

Quote from: william blake on February 26, 2020, 23:09:07
same gpu, comet is basically a rebrand, slightly higher clocked.
but gpu performance does not matter without increasing in energy efficiency somewhere, on cpu or gpu side, because 15w is 15w(or 45=45)
so, every time a manufacturer says "our mobile something is faster and bigger now" we should ask him, is this a new arch or node and you made it more energy efficient?
this is why every high-end mobile igpu is overrated as hell. low end is kinda underrated tho. lets say vega3 in ryzen 3 or athlon. this vega's are very good with 2-core cpu and 300usd market.

Yeah even if you play around with the cTDP, there's only so much you can extract from the iGPU. This still uses the UHD 630 so there's not much incentive on the GPU side anyways. It has to be paired with a capable Turing (or Ampere next year) part.

Quote from: xpclient on February 26, 2020, 22:45:58
Does this Comet Lake come with any GPU performance improvement or it is the same rehashed UHD Graphics 630? Of course I understand it will be paired with discrete GPU but still. If the AMD 4000 Series laptops have Thunderbolt 3 and NVIDIA dGPU, then I am sold on AMD.

same gpu, comet is basically a rebrand, slightly higher clocked.
but gpu performance does not matter without increasing in energy efficiency somewhere, on cpu or gpu side, because 15w is 15w(or 45=45)
so, every time a manufacturer says "our mobile something is faster and bigger now" we should ask him, is this a new arch or node and you made it more energy efficient?
this is why every high-end mobile igpu is overrated as hell. low end is kinda underrated tho. lets say vega3 in ryzen 3 or athlon. this vega's are very good with 2-core cpu and 300usd market.

Does this Comet Lake come with any GPU performance improvement or it is the same rehashed UHD Graphics 630? Of course I understand it will be paired with discrete GPU but still. If the AMD 4000 Series laptops have Thunderbolt 3 and NVIDIA dGPU, then I am sold on AMD.

Quote from: DF on February 26, 2020, 19:50:07
It's only disappointing that mobile Ampere will not be ready for this launch. 

mobile ampere is a huge step forward in energy efficiency and productivity. new tech process, new architecture.
mobile comet lake is the same thing as prior mobile lake. same tech, same arch.
hard to find any correlation between them.

It's only disappointing that mobile Ampere will not be ready for this launch.  2xxx super's won't really move the needle enough.  And thus for all designs which call for a separate GPU, I suspect demand for the new chips will be much lower except for those going for the "red" team.  AMD is highly likely to capitalize on the low improvement of the super series with launches of its own.  We'll see if I'm right.

Quote from: william blake on February 26, 2020, 18:31:55
i bet 10880h on par with 3800x in consumption also :)

There's a 10980HK expected as well.

i bet 10880h on par with 3800x in consumption also :)

Quote from: Imglidinhere on February 26, 2020, 16:22:16
If it's possible to find a laptop that can handle such a processor at max turbo, so far none of the options that house the 9880H can actually keep the processor from throttling, then that's great. But most options are enthusiast-grade and beyond what extremist payers are willing to pay imo.

I like the idea of this processor but as so many others have said, it doesn't matter how well you refine a 14nm processor with the same architecture, it can only get so efficient. If you run the older i7-8750H as per it's specs, not using the boost, as a 45w CPU, it sits at around 2.8GHz across all twelve threads! That's all!

Now you know why most laptops usually set the 8750H and 9750H at 55w, which pushes each up to around 3.5/3.6GHz, which is far better across the board. The issue is that most laptop heatsinks can handle MAYBE 140w for a combined CPU and GPU load, which is not much. Bob Of All Trades on YouTube makes a seriously good argument, which is where my statements are also coming from. He makes a good point.

Ultimately it boils down to whether or not this newer chip is worth its time and/or Ryzen mobile lineup coming out will bring the necessary competition to 'heat things up'. AMD is outselling Intel ten to one in desktop CPU sales around the planet, so here's to hoping we see better options cropping up in laptops and at least equal sales numbers when things get rolling.

Yeah you have a point. I expect this CPU to perform at its peak (or at least near to its peak) only in some Clevo models with gargantuan chassis designs. I don't think the boost will sustain long in say an Aero 15x-type design. But then, Intel will only brag about the chips themselves and not necessarily worry about how OEMs actually implement it.

If it's possible to find a laptop that can handle such a processor at max turbo, so far none of the options that house the 9880H can actually keep the processor from throttling, then that's great. But most options are enthusiast-grade and beyond what extremist payers are willing to pay imo.

I like the idea of this processor but as so many others have said, it doesn't matter how well you refine a 14nm processor with the same architecture, it can only get so efficient. If you run the older i7-8750H as per it's specs, not using the boost, as a 45w CPU, it sits at around 2.8GHz across all twelve threads! That's all!

Now you know why most laptops usually set the 8750H and 9750H at 55w, which pushes each up to around 3.5/3.6GHz, which is far better across the board. The issue is that most laptop heatsinks can handle MAYBE 140w for a combined CPU and GPU load, which is not much. Bob Of All Trades on YouTube makes a seriously good argument, which is where my statements are also coming from. He makes a good point.

Ultimately it boils down to whether or not this newer chip is worth its time and/or Ryzen mobile lineup coming out will bring the necessary competition to 'heat things up'. AMD is outselling Intel ten to one in desktop CPU sales around the planet, so here's to hoping we see better options cropping up in laptops and at least equal sales numbers when things get rolling.