Comments by "Winnetou17" (@Winnetou17) on "Intel's answer to Snapdragon PCs" video.

It wasn't Qualcomm which lit up the efficiency fire under Intel's bu^^. Because you can't answer to something like that in a couple of months. It was Apple, 4 years ago when they launched M1 which lit up the efficiency fire. That's how long architecting, designing and manufacturing a chip takes. And it can take even more than 4 years. I'm baffled about this "launch". Apparently actual reviews come only on 24th of September ? This seems to have been only a pre-launch hands on. Now wainting for another 3 weeks will be a bit painful. But I am excited about Lunar Lake, I really hope Intel didn't screw this one. I was mostly alone in battling people which were touting "hurr ArM iS sO mUcH mOrE eFfIcIeNt tHaN x86. durrr die x86" and I tried to explain that Apple's M efficiency comes from more sources, not ARM alone. Two of the major ones was memory directly on chip and better manufacturing node. Now Intel did both with Lunar Lake. Once the drivers for Linux are matured, we'll be able to see a proper ARM vs x86 comparison that's more apples to apples (pun unintended).
38
@BuggiEU Because that's normal ? I mean, normal it is for the reviews to come a bit before the launch. The launch means they are up for selling. Isn't it normal to have reviews so people know if it's worth buying ?
4
@BuggiEU What have phones have to do with this ? Look at Zen 5. And all CPU launches before that.
2
@systemBuilder Hey, it's true that the Intel chips were not efficient. But the heating part is totally Apple's fault. They had a colossal failure of a cooling system. Literally the worst in the market. I'm even questioning if they did it on purpose in order to justify the switch.
1
@felipematheus853 That's news to me, that the ARM GPU is absurdly efficient. Any sources on that ?
1
@felipematheus853 Pfffff, those comparisons are so off. 5600G is a desktop CPU. Second RX 580 is very old (in other words, meaningless in a comparison of today's state of anything) but also is ... a discreet GPU, which totally throws out of the window the efficiency comparison. Lastly it's ... not x86, lol. Still, if a chip can run CS:GO at 1 W that would be impressive. But I didn't find that. I did find that the M1 GPU runs at about 10 W, which is about as much as I expect. AMD's Z1 will also run at 10W on the lower end. Lunar Lake will be able to run the whole thing at 10W. Technically the others too, but LL should run a bit better when only fed 10W in total.
1