Apple quietly adds 10Gb networking option to the M1 macMini

[hishnash]

Summary

Finally the M1 MacMini gets a 10Gb networking option to the macMini, this was something that we all expected to be on the way after some details leaked in dec with 10Gb parts.  The option is a $100 upgrade.

 

Quotes

Quote

There was, however, a sign that 10 Gigabit Ethernet was coming. MacRumors discovered a mention of M1-based ‌Mac mini‌ logic boards with 10 Gigabit Ethernet in an internal parts list provided to Apple Authorized Service Providers.

 

My thoughts

Shame this did not ship day one for the M1 macMini but late is better than never.

 

Sources

https://www.macrumors.com/2021/04/20/mac-mini-10-gigabit-ethernet-add-on/
https://www.apple.com/shop/buy-mac/mac-mini/apple-m1-chip-with-8-core-cpu-and-8-core-gpu-256gb#

[bcredeur97]

Linus will be happy lol 

[SeraphicWings]

*cues Linus happy noise*

[leadeater]

Nice, odd it wasn't an option from the start but better late than never

[captain_to_fire]

3 hours ago, leadeater said:

Nice, odd it wasn't an option from the start but better late than never

Then there's me who thought that the reason why the M1 Mac mini doesn't have 10 gig ethernet is due to not enough PCIe lanes

[LAwLz]

That's nice.

I wonder what excuse people who said "the M1 is only more power efficient than x86 because it has less IO" will come up with now.

[leadeater]

20 minutes ago, LAwLz said:

That's nice.

I wonder what excuse people who said "the M1 is only more power efficient than x86 because it has less IO" will come up with now.

Ah yes still salty about that I see, continue to ignore the evidence that proved it as true. Now that we have both made off topic comments we can end this here.

 

Spoiler

FYI Infinity Fabric is I/O 

 

[avg123]

imac vs separate display with mac mini?

[leadeater]

3 minutes ago, avg123 said:

imac vs separate display with mac mini?

I'd take the iMac (the new M1 models), you'll be getting a better display unless you went with a higher end option that would make the mini + screen combination cost more than the iMac.

[Video Beagle]

48 minutes ago, avg123 said:

imac vs separate display with mac mini?

I can think of at least 5 youtube channels planning that very video right now

[Kisai]

1 hour ago, captain_to_fire said:

Then there's me who thought that the reason why the M1 Mac mini doesn't have 10 gig ethernet is due to not enough PCIe lanes

It was likely one of three things:

1) Apple didn't design a chip with it (if it's in the SoC)

2) Apple got caught in the chip shortage (10Gb is normally server NIC's)

3) Apple assumed people would use USB-C Thunderbolt or a TB docking station if they needed 2.5G or 10G ethernet

 

Like it costs $300 for a docking station or $150 for a 10GB TB NIC, so $100 addon is kind of a steal.

 

 

[Sauron]

Pretty reasonable price to be honest

[AlexGoesHigh]

Now if 10Gb NICs in general could get cheaper...

[WolframaticAlpha]

2 hours ago, leadeater said:

I'd take the iMac (the new M1 models), you'll be getting a better display unless you went with a higher end option that would make the mini + screen combination cost more than the iMac.

lel. I hate the new imac. So colourful. They should've accounted for people who hate color(My pc doesn't have a speck of color, all white and black)

 

 

Edit: JUst discovered a silver model. Would've loved a space gray tho

 

[WolframaticAlpha]

1 hour ago, Video Beagle said:

I can think of at least 5 youtube channels planning that very video right now

Imo, they should have put a 6 big cores and 2 small cores. Also a few more gpu cores. But hey, if it works, then don't fix it.

[LAwLz]

2 hours ago, leadeater said:

Ah yes still salty about that I see, continue to ignore the evidence that proved it as true. Now that we have both made off topic comments we can end this here.

Why would I be salty about being right?

Also, there was never any evidence posted. You posted some information like power consumption numbers for the total package and cores, but then you made your own guesses and assumptions based on those two numbers for how much power the IO used, while I argued that you couldn't just do that because there are a lot of shared resources between the cores apart from IO.

 

Also, my comment wasn't off-topic. The addition of the 10Gbps proves that the M1 can still retain its very low power consumption (the same) while also providing high speed IO, which has been something people have (sadly) been repeating on this forum quite a lot.

 

2 hours ago, leadeater said:

 

  Reveal hidden contents

FYI Infinity Fabric is I/O 

 

Moving the goalpost? This is the first time you have mentioned that while in our other conversations we mentioned PCIe lanes, USB ports, NICs etc when we talked about IO.

In fact, here is a quote from you:  

On 11/16/2020 at 11:28 AM, leadeater said:

It's not really a case of not being able to add more PCIe lanes it's a design decision for the target application. I/O is not very power efficient and also does not like dense and very small transistor/traces, even on small nodes I/O portions are purposely not actually at these in the design. The Apple M1 would not be as power efficient as it is if you start adding thins like more PCIe lanes and integrated 10Gb networking, this is one of the misunderstandings around the Apple M1 SoC and ARM when people try and compare to Intel or AMD CPUs/SoCs.

Emphasis added by me.

[leadeater]

40 minutes ago, LAwLz said:

Emphasis added by me.

Yep, but whole bunch of ignoring context and ignorance added in too. Because saying not as power efficient must somehow means a huge power increase 

 

May I ask that you repeat the reason why AMD kept Ryzen Mobile 5000, which uses Zen 3, at PCIe 3.0 and did not incorporate PCIe 4.0?

 

Not as power efficient, then you take a giant leap off the deep end about what that means. Not my problem.

 

40 minutes ago, LAwLz said:

Also, there was never any evidence posted.

Oh yes there was, a lot of it. You just refuse to believe it because "you're right". It's not my problem you can't literally look at Ryzen Mobile parts and figure out that I/O is the direct cause of the bad idle power of the desktop parts. End of discussion on this one, I've already decided it's a lost cause to bother, shouldn't have bothered with this post but w/e, I can't see this going anywhere productive.

 

P.S. No that goalpost was literally in my discussion about the IOD on the desktop platform, exhibiting that you indeed read nothing I wrote back then further showing how futile this is. Edit: You're the one that interjected back then too to try and incorrect me and you're the one now going back and selective quoting sections of posts for a discussion that spanned many, I can for sure tell you I did talk about the IF links between the CCD and IOD during that entire discussion back then, and I noticed back then you ignored it too. So I have no idea what to tell you.

 

Edit 2:

Just for you I put some more effort in to getting you the proof.

On 11/17/2020 at 8:48 PM, leadeater said:

I never said it will balloon, it's a well known fact that I/O is not power efficient and also does not respond well to node shrinks. M1 I/O design is totally different, it doesn't need to drive Infinity Fabric links and it doesn't need to drive higher power DRAM that is much more distant to the memory controller, both these being a significant portion of the IOD power usage.

So yea 100% did talk about it to you back then.

[LAwLz]

34 minutes ago, leadeater said:

Because saying not as power efficient must somehow means a huge power increase 

That's what you were implying, yes.

If you ask all the people who thumbed you up back then "how much power do you think adding more IO to the M1 will add?" then I think most people would answer "a lot" after reading your posts.

 

34 minutes ago, leadeater said:

May I ask that you repeat the reason why AMD kept Ryzen Mobile 5000, which uses Zen 3, at PCIe 3.0 and did not incorporate PCIe 4.0?

Could be for any number of reasons. Saying that it must be because of power consumption reasons is certainly a big assumption that is not based on any evidence.

"Ryzen 5000 mobile does not have PCIe 4.0. It must be because PCIe 4.0 uses a lot of power so therefore ". The first sentence is a fact. The second sentence is an assumption. 

[Vanderburg]

I'm so glad I bought a Mini less than a month ago. I would have happily paid the extra $100.

[LAwLz]

5 minutes ago, Vanderburg said:

I'm so glad I bought a Mini less than a month ago. I would have happily paid the extra $100.

Not sure how it looks in Texas, but in Sweden Apple lets you return a product for no additional charge up to 14 days after the purchase.

Some other stores offers longer return periods as well.

[Vanderburg]

18 minutes ago, LAwLz said:

Not sure how it looks in Texas, but in Sweden Apple lets you return a product for no additional charge up to 14 days after the purchase.

Some other stores offers longer return periods as well.

Bought it from apple.com, delivered March 24th, sadly.

[Video Beagle]

10 hours ago, Vanderburg said:

Bought it from apple.com, delivered March 24th, sadly.

I'd say you could try going thru legit means "I just got this, but would happily pay the extra blah blah"

or not so legit "This doesn't bring me joy and I'm returning it!" then "This new one may bring me joy!"

[leadeater]

30 minutes ago, Video Beagle said:

I'd say you could try going thru legit means "I just got this, but would happily pay the extra blah blah"

or not so legit "This doesn't bring me joy and I'm returning it!" then "This new one may bring me joy!"

From what I know/heard US Apple support is very good and accommodating, I would expect them to help out.

[Vanderburg]

1 hour ago, leadeater said:

From what I know/heard US Apple support is very good and accommodating, I would expect them to help out.

I might give that a go. The worst they could say is no.