He doesn't know that the PL1 and PL2 defining the peak power consumption can be changed by the vendor.That s the power measured at the main with 4W idle power, so that makes about 9W in ST and not 14W as claimed by this tweetos.
80 watts in cb 2024 I think let me find that tweetWonder how much this thing would pull at 3.8G all core?
Some atomwaffen stuff right here.
Don’t trust marketing slides when it’s release dayNo need.
View attachment 101405
hell yeah it's a hand nuke.Here it is. This the 78 SKU
This has more perfs and a 4 more LITTLEs on the side.It was M3 Max
It's around 10W SOC 1t at measely 3.4.
At 4.2 this thing gonna smelt titanium.
This is why the next generation should reduce the frequency, not increase it!
It's still fine.Cinebench is their best score overall since they manage to match Phoenix@35W, but Kakan will use 25W for the same throughput, that s 40% better perf/watt at same throughput than both those chips.
They just need to work on their physdes and binning capabilities.This is why the next generation should reduce the frequency, not increase it!
Yeah the issue here is phydes, and possibly (?) the pmic BS Charlie claimed — where they front smartphone pmics and pcbs and marred power at higher current.It's still fine.
now, 85W peak nT power is not fine, that's luggable territory.
They just need to work on their physdes and binning capabilities.
It's not rocket science.
Physical design.ze what?
Genuinely didn't expect GWIII to deliver 55W R23 handnukes on pedestrian SKUs.Yeah the issue here is phydes, and possibly (?) the pmic BS Charlie claimed — where they front smartphone pmics and pcbs and marred power at higher current.
Changed employersWonder what happened.
WhyMan now I'm feeling really funny about SD 8 gen 4.
Freedom at last.Changed employers
See that 3.4G power being what it is?
Apple hweng is slave galley so the change was almost certainly positive.and the sudden change in environment probably had an impact
RKL was a weird one-off backport by an unrelated team, this is prim and proper setpiece for a gen1 Nuvia core and it's...After looking at this I genuinely believe that v1 will be like Rocket lake and v2 will be much better.
Yep. I’m ankle-deep in on “a few things were rushed or went wrong” even tho it’s still a good part on paper just lateMan now I'm feeling really funny about SD 8 gen 4.
Physical design.
AMD's favourite.
Genuinely didn't expect GWIII to deliver 55W R23 handnukes on pedestrian SKUs.
Wonder what happened.
Works ok for Cortex-X/Zen/Atom et al, server and mobile design targets for non-SIMD have a lot of overlap.The server origins of the core might not be helping...
Not comparable to rocket but I do think we’ll know if they have it in them with 8G4 & then 2025.You don’t go from executing so well from A7 to A14 and the sudden change in environment probably had an impact. After looking at this I genuinely believe that v1 will be like Rocket lake and v2 will be much better.