On this note, I am relatively sure Ryzen 7000 is the desktop alternative to Ryzen 6000! But yes, it is a very valid point that if they were to spend a lot of time developing a new mainboard and architecture for AMD (I do not know what this entails), that Ryzen 6000 by that time would be outdated.
Ryzen 6000 is Zen3+, Ryzen 7000 is Zen4. It remains to be seen if the mobile-specific sucessor to Zen3+ is vanilla Zen4 or if itās something like a Zen4+.
ā¦though considering that Zen4 is looking to launch later this year, the timing would suggest that the sucessor to Zen3+ will indeed be just plain-old Zen4, and therefore would likely still be Ryzen 7000 on mobile in the same way that Ryzen 5000 on mobile is bog-standard Zen3 (and Zen3 also launched second half of the year).
Just a bystander here, even though I do side with your opinions, it was clarified in a Intel Technology episode, that you can disable all the P cores except one, which is needed to boot, and for important functions. I too prefer quieter laptops, that have a battery that can actually last.
Hmm! Yeah I just got the feeling from the CES presentation, although Iām sure AMD will release a Zen 4 mobile APU soon after! Hopefully Zen 4 retains Zen 3+'s power efficiency though, considering itās a new architecture it may not retain it. 
Edit: Oy! @NM64 Youāre making me anxious! Youāre still typing!!! Jk
Itās my understanding that a big part of why Intel has almost seemingly regressed in efficiency is because their 10nm was such a borked up node. By comparison, we can refer to Appleās success with TSMCās 5nm node to see that efficiency should theoretically still be fine.
ā¦emphasis on theoretically, because thereās always the chance of a Bulldozer MKII (though that itself was arguably already Netburst MKII, so I guess Netburst MKIII? But even then Rocket Lake was arguably on its way to being an early Netburst MKIII ala Willamette before Alder Lake aborted that arc)
Iām just OCD about fan noise, even on my desktops. 
On my primary desktop PC, I run a wacky dual-monitor setup with an LCD and a CRT monitor, and my desktop is so quiet that, unless the CPU is really cranking (not all 100% workloads generate the same amount of heat - I recently discovered that CPU-based waifu2x is one example with greater-than-usual heat-load) or Iāve intentionally set it to 4.6GHz 1.31v instead of the more sensible 4.5GHz 1.24v, the minor hum and/ro coil whine from the CRT monitor is actually louder than the PC fans. Heck one of my SSDs and some of my RAM has coil whine that, when hit heavily, is louder than my PCās fans!
Even on an older HP DM1 laptop with an AMD E-350 I would commonly straight up unplug the fan and underclock the CPU to around 2/3 of its stock clocks, and coil whine by far becomes the loudest thing with CPU temps never really going above 80c when combined with an undervolt (interestingly the iGPU can push up above that when specifically doing hardware-accelerated video decoding).
If/when underclock/undervolt functionality becomes possible on the Framework, I canāt help but wonder if one could similarly ācheatā by just straight-up unplugging the fan.
I started typing one post to your initial post and then, before posting it, I copied the text and combined it into one big combo-post.
ā¦Iām also a bit of a chronic editor if you couldnāt tell.
1 Like
Mmm! Although I personally just set the fan locked to 30% usage, to ensure thereās at least a bit of airflow! Completely silent too. Iām not to sure about CPUs actually though! You can lock the CPU usage in Windows directly, Iām not to sure about undervolting though, is it already availible with the 11th gen CPUs? Would be great. I too am a fan of limiting my hardware for the sweet silence.
BTW regarding just unplugging the fan, some laptops will straight-up refuse to go past the POST screen with the fan unplugged - a 15 year old Gateway laptop I got from a cousin that had a faulty fan was really annoying in this regard since it wouldnāt let me boot unless I replaced the fan!
The aforementioned HP DM1 laptop instead, when the fan is unplugged, just throws up a warning on the POST screen that requires you to press the enter key to continue (if you donāt press the enter key itāll instead just shut down after 15 seconds).
Itās my impression that Intel has locked down that feature due to a (supposed) security issue with it.
Of course, open firmware could theoretically work around thatā¦
1 Like
I like the sound of that 
. Of course Intel locked down the feature! 
What did I expect, being able to control the CPU you own? Hahaha. On a more serious note though, yeah with Framework releasing the firmware, thereās limitless possibilities! Just a bit of a patience, and Iām sure thereāll be a dedicated modding community!
Although I am getting awfully off topic here, so just to throw in my cent, I would love AMD CPUs, although I would purchase a Framework once I have the money either way, as I want to support the company.
Iām not quite sure why the motherboard matters for ecc now that the imcs are on the cpu die? surely if the imc does it and the ram does it, then its a simple on/off in the bios if its not enabled by default?
Iāll throw in another additional reason to prefer an AMD CPU.
You know the idea where you can run a Framework laptop motherboard all on itās own? Iāve basically kind of done that already with a previous HTPC that was an 8440p laptop where I just removed the screen, put some desktop case fans on top in-place of the keyboard, and called it a day.
So the fact that you can do this with the Framework in a less derpy manner is something Iām definitely interested in.
But wait, isnāt this supposed to be related to AMD? Why yes it is - I recently had the ājoyā of discovering that Intel on Linux does not support aspect ratio-preserved resolution scaling on external displays without the use of imperfect 3rd party software (look at the username on that forum thread):
https://forums.linuxmint.com/viewtopic.php?f=59&t=366662&p=2134263
Worse still is that there was even a freedesktop.org issue listing for this very issue that was closed over 2 years ago now under the impression that it was āno longer an issueā (protip: itās still an issue even on a Xubuntu 22.04 daily build from 2022-02-05; at least it is on both Haswell and 965GMA):
So how does that relate to AMD? Simple - on AMD graphics, you can just use the xrandr argument --set "scaling mode" "Full aspect" and everything will ājust workā and the setting is even remembered anytime the resolution is changed in any manner, including software switching to exclusive fullscreen at a set resolution, until you turn off your PC (so the easy solution is to simply have the xrandr command automatically ran whenever you boot).
The irony is that both Intel and AMD are using the mesa graphics driver stack rather than anything proprietary (though apparently the setting also exists in Nvidiaās proprietary graphics blob).
Perhaps this is exactly the case, but the BIOS is still handled by the motherboard.
I do recall a thing where Gigabyteās AM4 boards only support ECC on boards with 6 layer PCBs or more as stated directly by a Gigabyte rep on Hardforum, so perhaps itās something to do with making sure thereās quality-enough signal integrity to handle the additional bit that ECC RAM uses?
1 Like
sure bios is on the motherboard, but it can be updated, and is often modified by third parties to add features or unlock options hidden by the manufacturer.
I find that to be more common on desktop enthusiast motherboards (see: running 8th gen Intel on 100-series boards), but maybe Iām just bitter that the previously-mentioned Gateway laptop had an ATI 3200-based chipset that definitely has the hardware for AHCI (Iāve another PC with the same chipset and AHCI) but it was not present in the BIOS and, if I knew about that back when I was originally given the laptop years and years ago, the bios-modding community may have had more interest in fulfilling my request as that was a much more common request back then than it is now.
(for those that donāt know, AHCI is required for TRIM which itself is required for basically any SSD made in the last decade to maintain proper performance; only some really old ones like OCZ Onyx can have their early equivalent of TRIM manually activated without AHCI and even on WinXP via a āwiper.exeā program which was notable as TRIM support was only introduced with Win7)
Again, yet another example of why Iām really starting to value the idea of open firmware.
2 Likes
If FW decides to support Ryzen 7000 for their current chassis, they would likely use a 28 W āPhoenix-Uā. There are some rumours that Phoenix-H will use Zen 4, which probably means Phoenix-U will also use Zen 4.
Then thereās the mysterious āSabrinaā that showed up in the coreboot repo, which might be an alias for Phoenix-U or perhaps the actual codenameā¦
1 Like
Just praying Framework doesnāt have a exclusivity deal with Intel 
. Very possible.
2 Likes
This is exactly what I am worried about. Specifically with some of the major investors involved with this project may steer them away from Open Source friendly hardware and be more Windows centric.
2 Likes
Exactly who are you referring to? The only investor Iām aware of is Linus and I assume a bunch of hedge funds/VCās that likely donāt care so long as they get returns
1 Like
Have you also considered the fact that AMD support with Thunderbolt is less mainstream than Intel. That said, I would too love to see an AMD option having had all my devices switch to AMD as I find them to have nicer business morals than Intel.
3 Likes
USB4 basically is Thunderbolt, and Ryzen 6000 is the first AMD product that integrated USB4.
(this is why there are concerns at all around Pluton in this thread, because USB4 is kind of a baseline requirement for a Framework device, but the first AMD processor with USB4 is also the first AMD processor with Pluton which makes it very much a double-edged sword for a Framework device)
1 Like
Welcome to the Forums!
I think it is a mistake to believe that any company has intrinsically better business morals than others. Public companies have one responsibility that they are legally bound to uphold, to extract maximum shareholder value.
Corporations are amoral instruments, there is no such thing as bad corporations or good corporations. There are immoral people in charge of corporations but that is not the same thing.
Intel is not immoral and AMD is not moral or vice versa.
Please donāt ever assume that a company is on your side. There are valid reasons to want AMD over Intel. Business practices is not one of them.
1 Like
Yeah didnāt mean it like that, not sure how to phrase it 
. Its just that with AMDās drivers, for example, they are all open source, and thatās what this laptop is about in many ways! I feel as though they are more accessible, also AMDās profit margin is lower so I am probably biased as I love good value for money
3 Likes