Robert Müller (Mueller.delete@this.mueller.com) on September 18, 2020 3:59 am wrote:
> Andrei F (andrei.delete@this.anandtech.com) on September 18, 2020 3:39 am wrote:
> > Michael S (already5chosen.delete@this.yahoo.com) on September 18, 2020 2:58 am wrote:
> > > anon (anon.delete@this.anon.com) on September 17, 2020 7:10 pm wrote:
> > > > AnandTech's (SPEC ST performance) review is here: anandtech.com/show/16084/intel-tiger-lake-review-deep-dive-core-11th-gen/8
> > > >
> > > > All of TigerLake's extra performance over IceLake is due to higher clocks. And it significantly
> > > > outperforms the Zen 2 based AMD Renoir chip (Ryzen 7 4800U). However not all is good: TigerLake
> > > > experiences a noticeable IPC regression compared to IceLake. The memory subsystem is unable
> > > > to keep up with the higher clocks, and the reworked cache is not enough.
> > > >
> > > > And it just barely beats Apple's A13 despite using significantly more power.
> > > > Makes you think the A14 based Mac coming later this year will blow past Intel's
> > > > best CPU microarchitecture in single threaded integer performance.
> > > >
> > > > https://images.anandtech.com/graphs/graph16084/111168.png
> > >
> > > The picture reminds me of 1951's Arthur C. Clarke classic short story "Superiority".
> > >

The ultimate cause of our failure was a simple one: despite all statements to the contrary, it was not
> > > due to lack of bravery on the part of our men, or to any fault of the Fleet's. We were defeated by one
> > > thing only - by the inferior science of our enemies. I repeat - by the inferior science of our
> > > enemies.

> > >
> > > I mean, minimally tweaked 5 y.o. uArch manufactured on maximally
> > > tweaked 6 y.o. process reigns on top of the chart.
> >
> > The difference that you can't see in that chart between the top 3 positions is that
> > it's 40W vs 21W vs 4W. The apparent superiority will come crumbling down soon enough.
>
> Andrei, do you have power draw estimates for the CPUs while running these workloads? I am still not
> 100% clear on the figures. If I understand your article correctly, Tiger Lake draws 50 watts at 4.8
> GHz. Does this mean that the 28W TDP profile actually means 50W power consumption? What confuses me
> even further is that a Coffee Lake (single core) running at 4.6 GHz only seems to consume around 30watts,
> I find it hard to believe that Tiger Lake would be a regression in energy efficiency.
>

Those are the figures, TGL single-core power is around 20-21W-ish at 4.8G.

The 50W figure is multi-core for a limited period of time before it throttles down to TDP levels at 15/28W sustained (with subsequently lower clocks). Comet Lake (10900K) is 40W-ish single-core at 5.3GHz. Essentially the new 10nm parts and new architecture is about double the efficiency of the 14nm Skylake parts.