M4 Rumors (requests for).

[dada_dave]

I haven’t encountered this person. Now I’m half way between checking his posts out and avoiding entirely!

Edit: first post saw something about KDE Plasma…I”m out.

Edit2: second post I read is some drivel about “your computer isn’t yours”. Are we sure this isn’t another mi7chy alt lol.

No way worse I think - like this person's posts are filled with Apple and government spooks tracking everything you say and do. They push this on every topic no matter how ancillary. As annoying as they are, mi7chy just trolls with off-the-wall and ignorant comparisons to make it seem like Apple hardware is substandard compared to PC, especially AMD CPUs and Nvidia GPUs. But this individual, as far as I can tell, their main thing is trying to convince everyone that Apple and the government are out to get them and they seem to really, truly believe it. That's scarier - basically one step from believing that vaccines have microchips to control your mind. In fact, if someone told me Minghold had an alt on Reddit spreading exactly that, I wouldn't be surprised. That and his user name/Ming the Merciless profile picture is ... well it's a choice to use something that steeped in racism as their profile picture and name. @mr_roboto already got suspended for interacting with them once.

 

[mr_roboto]

That and his user name/Ming the Merciless profile picture is ... well it's a choice to use something that steeped in racism as their profile picture and name. @mr_roboto already got suspended for interacting with them once.

Yeah, he's really a piece of work. I forget what I said to get suspended, probably just mildly insulted him for his insanely paranoid interpretation of, well, everything.

 

[Cmaier]

Yeah, he's really a piece of work. I forget what I said to get suspended, probably just mildly insulted him for his insanely paranoid interpretation of, well, everything.

when i went to see who this guy was, i saw a post you made where you called him paranoid, and thought “how did that not get him suspended?”

 

[mr_roboto]

when i went to see who this guy was, i saw a post you made where you called him paranoid, and thought “how did that not get him suspended?”

Well, apparently I ate a ban. Had a nice reply typed up dissecting how incredibly wrongheaded he was from a more technical perspective too.

 

[Jimmyjames]

Well, apparently I ate a ban. Had a nice reply typed up dissecting how incredibly wrongheaded he was from a more technical perspective too.

Entirely in keeping with the quality of moderation over there.

 

[NotEntirelyConfused]

It really is astonishing how bad the mods are there. Or perhaps I should say, how poorly their apparent goals align with their actual interests.

Regardless of your political leanings, I think it would be noncontroversial to say that introducing rabid partisanship into unrelated discussions by having a very large and extreme political sig is not conducive to good forum hygiene. I reported such a couple days ago and they responded within an hour saying "political sigs are allowed". Really really dumb.

 

[dada_dave]

Meant to put this here in the M4 thread (sorry):

Published analysis of die shots comparing Qualcomm Elite, Strix Point, and Apple M4:

D

Post in thread 'Nuvia: don’t hold your breath'

Sep 30, 2024

Comparing the die shots of the Snapdragon Elite with the M4 and the AMD Strix Point:

Qualcomm Snapdragon X die shot reveals massive CPU cores with huge caches

GPU is quite modest though.

www.tomshardware.com

Check out this beautiful die shot of AMD's new Ryzen AI 300 'Strix Point' APU

AMD's new Ryzen AI 300 series 'Strix Point' APU die shot revealed: TSMC 4nm process node, Zen 5 CPU, RDNA 3.5 GPU, XDNA2 NPU in great detail.

www.tweaktown.com

So one thing I didn't realize is that I thought the AMD chip was on regular N4, which I believe is what the Qualcomm Snapdragon SOC is fabbed on, but in fact the AMD chip is fabbed on the slightly newer N4P which has 6% more performance than N4 but...

• dada_dave

If anyone has annotated Apple M2 Pro die shots that they'd like to make a comparison with, that'd be great!

==============

I have to admit: while Apple adopted the 6 E-core design for the base M4, if the M4 Pro doesn't have its own bespoke SOC design and is a chop of the Max, then, depending on how Apple structures the upcoming M4 Max/Pro SOC, it'll be a shame to see Apple lose a product at this performance/efficiency point. The M3 Pro is rather unique. Also, its 6+6 design really highlights how improved the E-cores (and P-cores) were moving from the M2 to the M3, especially in this workload.

Thinking about this statement I made in my last M3 analysis post, actually now that I think about it, there doesn't seem to be a reason why even if the M4 Pro is a die chop of the M4 Max that the additional die area added to the Max's SOC couldn't simply have an additional P-core cluster. Just making numbers up but say the M4 Pro had one six-P-core cluster and one six-E-core cluster, the bottom part of the die could simply add another 6-P-core cluster. Sure, Apple didn't do that for the M1/M2 but there is no reason why they couldn't and the M3 and M4 clearly show a departure for the M1/M2 design. It could also be two four-P-core clusters and one six-E-core clusters and an additional four-P-core cluster, but you get the idea.

 

[NotEntirelyConfused]

I assume most have now seen the leaked M4 MacBook. The posted GB6 numbers are unimpressive - which is as expected, since running them shortly after first power-on (which is what they did, AFAIK) should pollute the results with lots of background tasks running (like mdworkers).

Does anyone have a reasonable rule of thumb for how results are likely to change after things settle down? Like, "SC will go up by 3-4%, MC by 7-9%"?

 

[leman]

I assume most have now seen the leaked M4 MacBook. The posted GB6 numbers are unimpressive - which is as expected, since running them shortly after first power-on (which is what they did, AFAIK) should pollute the results with lots of background tasks running (like mdworkers).

What makes you say they are unimpressive? This is a better single-core score than anything else currently available on the market — by a substantial margin. It completely outclasses Lunar Lake, mobile Zen5, and the Snapdragon X. And it matches or outperforms high-end desktop Zen 5.

If this score represents the average GB6 result for the base M4 Mac, I'd say Apple has a big winner. If instead this is an early score and the final average GB6 score is 3-5% faster, well, that would be just crazy.

 

[NotEntirelyConfused]

What makes you say they are unimpressive? This is a better single-core score than anything else currently available on the market — by a substantial margin. It completely outclasses Lunar Lake, mobile Zen5, and the Snapdragon X. And it matches or outperforms high-end desktop Zen 5.

If this score represents the average GB6 result for the base M4 Mac, I'd say Apple has a big winner. If instead this is an early score and the final average GB6 score is 3-5% faster, well, that would be just crazy.

Sorry, I meant "unimpressive for M4 in a cooled Macbook Pro chassis" and specifically in comparison to the un-cooled M4 iPad. Obviously they demolish everything else that exists.

 

[NotEntirelyConfused]

Sorry, I meant "unimpressive for M4 in a cooled Macbook Pro chassis" and specifically in comparison to the un-cooled M4 iPad. Obviously they demolish everything else that exists.

I'll also note that it's running at the same 4.4GHz that's the nominal speed of the M4 iPad. I was a little surprised - I thought they'd push it at least a little harder - but upon reflection, given the rarity of sightings of iPads actually running at that speed for any length of time, I shouldn't have been. It makes sense that Apple would give the M4 its head on the iPad, at least along the reasonable part of the power/performance curve, letting thermals control actual speed. So (assuming the leak is legit) they've probably done the same thing here, and the fan will let it stay at or closer to 4.4GHz for longer than the iPad, but it will likely still throttle back some.

 

[quarkysg]

Sorry, I meant "unimpressive for M4 in a cooled Macbook Pro chassis" and specifically in comparison to the un-cooled M4 iPad. Obviously they demolish everything else that exists.

Well, if you look at it that the M4 could run to it's max in a passively cooled iPad enclosure, it is really impressive. Of course it (the iPad) wouldn't be able sustain it for long.

 

[leman]

Sorry, I meant "unimpressive for M4 in a cooled Macbook Pro chassis" and specifically in comparison to the un-cooled M4 iPad. Obviously they demolish everything else that exists.

Ah, I see. Well, the score of 3864 is roughly 5% better the averages reported for M4 iPad, so I think it makes sense. We see less difference between the passively and actively cooled M2 models for example.

I'll also note that it's running at the same 4.4GHz that's the nominal speed of the M4 iPad.

On my iPad Pro the actual reported peak frequency is closer to 4.05-4.1 Ghz. It also starts to draw quite some power at higher frequency.

 

[Jimmyjames]

Difficult to draw conclusions from one result taken soon after unboxing. I think it’s a very nice score and I can easily see the Max pushing 4000 sc without Geekerwan and NO2! Perhaps 24000 or 25000 mc.

 

[Cmaier]

Seems like a very nice score. Presumably this is the base model?

 

[thenewperson]

Seems like a very nice score. Presumably this is the base model?

That's the assumption (or hope, given the 16/512 config)

 

[NotEntirelyConfused]

Ah, I see. Well, the score of 3864 is roughly 5% better the averages reported for M4 iPad, so I think it makes sense. We see less difference between the passively and actively cooled M2 models for example.

Sure, but we know that this score came right after unboxing. That should mean that there's a relatively large amount of background processing going on, which can directly impact MC scores, and (through reduced thermal headroom) indirectly lower SC scores too.

So I'm hoping we see final typical scores a bit better than that. The point of my initial post was to ask if anyone has a good rule of thumb about how scores settle out after the Mac's done with that burst of initial activity.

 

[Yoused]

How much indexing happens on a new machine out of the box? Unless you are doing a MA setup with a lot of transferred files, what is there to index?

 

[NotEntirelyConfused]

How much indexing happens on a new machine out of the box? Unless you are doing a MA setup with a lot of transferred files, what is there to index?

That's a very good question. Which is why I asked mine.