Good point. I see no consumption measured, but Tiger Lake is set to 15W and 25W and tested separately against 15W 4800U. The 15W is competitive, the 25W clearly better than the AMD. Intel therefore needs more TDP space for this: Hardware info (lost link, will follow).
But in this review it seems a bit more nuanced. In any case, I recommend reading the first page of this article, almost essential for the context.
Here a direct link to Tiger Lake’s power consumption versus Ice Lake versus Renoir all at 15W:
So you see that in AVX-512 (which I don’t think is a fair comparison, really an Intel thing, AMD doesn’t do AVX-512), Tiger Lake shows a huge improvement compared to Ice Lake. Good news. The AMD takes more power here idd, I see that the AMD can absorb a higher power for longer. I think that’s good news too. Ultimately, however, AMD needs more joules to complete the Intel test.
The 2nd test is not an Intel test. In that Renoir even beats Tiger Lake.
I would have liked it if they had tested both at a higher TDP and had taken a 4800H (which can be up to 35TDP) so that you can test both at comparable high TDPs.
And this is all on 15W TDP (see the first page for an explanation of what that means).
In practice (Lenovo S540 13-ARE, can choose from 15-20-25W TDP in the Bios) I don’t notice anything at all.
I do believe that the Tiger Lake does better on battery, after all, it is precisely in terms of power consumption optimized compared to Ice Lake. So you can expect that too. AMD can do that trick with the next generation, as the 4800U is AMD’s first worthy low-cost laptop APU (I found rubbish before that!).
By the way, here’s a review that reasonably supports my point, the Tiger Lake 28W takes up more power than the 4700U 15W:
So it is actually comparing apples with pears. The Tiger Lake has to be set to 15W TDP to compare with this AMD.
[Reactie gewijzigd door Jazco2nd op 14 oktober 2020 00:19]