> There were also 1,5 ghz ones that came with the 32mb radeon.

I've never heard or read anything like that.

> Note that the 1,5 ghz minis ar really 1.42 ghz but overclocked.

I don't think so. The 1.42 GHz G4 Mac mini has a 7447B, which has been available up to 1.7 GHz (besides, the 7447A likewise):

https://www.nxp.com/docs/en/reliability-and-quality-information/MC7447AQUAL.pdf

So the 1.5 GHz G4 Mac mini can very well have a genuine (i.e. not overclocked) 1.5 GHz 7447B. To know it for sure one would have to open his 1.5 GHz G4 Mac mini and read what's printed on the CPU ...or find someone who did:

"Got adventurous this evening with a speed bumped Mini g4. I removed the heat sink and confirmed that the chip was in fact rated for 1.5GHZ and it was."
http://www.applefritter.com/node/10820/52363#comment-52363

And even if the 1.5 GHz was overclocked from 1.42 GHz (which it is not, apparently): There's a difference if Apple did it or somebody else. In the former case the Mac mini would have 64 MiB VRAM, in the latter one it has 32 MiB VRAM.


Edit: changed link to PDF

[ Edited by Andreas_Wolf 25.04.2022 - 22:51 ]

>> There were also 1,5 ghz ones that came with the 32mb radeon.
>I've never heard or read anything like that.

AFAIK those 1.5 GHz 32MB are only overclocked 1.42 GHz versions. Some companies sold refurnished versions which they also overclocked.

Geit

> AFAIK those 1.5 GHz 32MB are only overclocked 1.42 GHz versions. Some companies
> sold refurnished versions which they also overclocked.

Alright, thanks. So not overclocked by Apple, just like I said. Whether it's overclocked by such company (different from Apple) or by end user doesn't make much difference on the second hand market, in my opinion.

Quote:

I don't understand when some make statements like the above, without any evidence to back up such a claim, when others have already made statements with references to a source that confirms the information. Nothing personal intended toward ThEcRoW, as he has made many useful and constructive postings in the past, and I am sure he will do more in the future. If you have some reference, documentation, or other source that can back up your claim that Apple did distribute some 1.5GHz MacMini's with only 32mb VRAM, please point us to that source, because from what I have researched on Apple's own support specification pages, the 1.5GHz G4 was ONLY available with 64mb VRAM and the amount of VRAM is not changeable by the consumer (and probably not an upgrade that even Apple, or a independent shop tech could do).

Acill has confirmed that his 32mb VRAM, 1.25GHz MacMini runs MOS2.4 very fast and with no problems regarding running out of VRAM. So, as he and some others have written, buying a lower spec MacMini will still get you one of the fastest MorphOS computers on the planet. Only the 1.42GHz and 1.5GHz models will be a little faster.

I have not checked the price difference between the 1.25GHz models and the faster models, but if it is significant dollar amount, the 1.25GHz model is likely the better deal.

Edit: @Andreas, if the G4 in the MacMini was available up to a 1.7GHz speed, I wonder if it would be possible to safely overclock the 1.5 chip to 1.7 or higher?

[ Edited by amigadave on 2009/10/29 16:03 ]

Im not Andreas but on a.org piru wrote a thread of how to overclock the macmini and up to 1,5ghz seems a bit unstable.

I was referring to a info i read some time ago that those minis were called "silent updates" by apple, and there is on another forums people who has the mini at 1,5ghz but with 32 vram. Mine has 32vram and its perfect, the more memory the better, but 32 mb vram seems enough for morphos, at least the things i'm testing and doing.
I'm not claiming what i said was true over the another posters opinion, just stated what i had heard on forums and on the net, so no offense taken, dave.
Morphos is great, and as someone pointed out, even the less clocked mini will give you a fast morphos experience.
Sorry if i have not explained well, but english is not my primary language.
Cheers morphers!!!

> a info i read some time ago that those minis were called "silent updates" by apple

The 1.5 GHz G4 Mac mini "silent upgrade" *always* has 64 MiB VRAM.

> there is on another forums people who has the mini at 1,5ghz but with 32 vram.

That's no "silent upgrade" 1.5 GHz G4 Mac mini then, but overclocked either by end user or by reseller company.

> if the G4 in the MacMini was available up to a 1.7GHz speed, I wonder if it would be
> possible to safely overclock the 1.5 chip to 1.7 or higher?

Check my link to Applefritter where someone stated he successfully overclocked his 1.5 GHz rated 7447B to 1.83 GHz. But that's always to be taken with a grain of salt because the 1.5 GHz Mac mini's 7447B is probably rated at 1.5 GHz for a technical reason. The existence of higher rated 7447Bs doesn't mean all 7447Bs are able to run at that clock frequency savely. And alway have in mind that overclocking provokes a shortened life span.

Why one would want to overclock a Mac Mini (specially if it already runs The Lightning OS) is well beyond me...

My Mac mini 1.5 GHz (silent update) showed 64 MB VRAM under MacOS, but under MorphOS the GfXCard tool shows 32 MB VRAM only.

Is this a bug of the tool or there are some settings to be changed?

Quote:

there' s a true performance benefit in overclocking a 1,5ghz mini to 1,8ghz (max you can reach). It's simply faster. Now it's not about the OS but the apps, some ressource hungry apps (Mame, some demos, some games, encoding tools etc.) would benefit from it.

Now it has some downsides, especially about the noise, a 1,8ghz overclocked macmini is pretty noisy, the fan being blowing all the time. And some macmini work very well at 1,8, some get unstable.

Qwert44: as far as remember you have to type reset reset-nvram and then
reset-all in Open Firmware.

> Is this a bug of the tool or there are some settings to be changed?

That's a bug in the actual GFX driver that can occur with 64 MiB VRAM Mac minis and will be fixed in the next MorphOS release according to bigfoot. The GraphicBoards tool is always showing the VRAM amount MorphOS can access. So it's not just a wrong number in the tool, unfortunately. I have the same problem when autobooting MorphOS. When starting MorphOS from the graphical boot menu it shows full 64 MiB VRAM.

Thanks Andreas &al, -- the GUI bootmenu solves the issue, and the system seems to be even snappier.

Needless to say, if you could cool the processor in a Mac Mini (I would probably leave the case disassembled and try to add a better CPU cooler) you'd probably have a better chance of a stable overclock.
After overclock numerous X86 systems, there's one common thread I've noticed. You may be able to get a system to boot at a specific overclock, but the real test is how stable and cool your system is under load (I don't doubt the 1.83 overclock, I doubt its stability under load).
Better cooling definitely helps any overclock and I doubt the (with its compact form factor) that the Mac Mini's CPU cooling is all that great.

[ Edited by Jim on 2010/10/4 0:23 ]

> I don't doubt the 1.83 overclock, I doubt its stability under load

Still the same sceptic? ;-) Statement from pampers as recent as September:

"My Mini is overclocked from 1.5 to 1.83 - no problems at all."

But it seems he's about to sell it in favour of a PowerMac.


Btw, I think you replied to the wrong message ;-)

He makes no statement about its operation under load Andreas. I've had PCs that overclocked for 20 or 30 minutes under heavy load, then crashed. Since most operations under MorphOS are not that CPU intensive, I could see a Mac Mini running under MorphOS at 1.83 exhitbiting fairly good stability and only crashing every once in awhile.

Although, if the Mac Minis do use a 7447B, then operating at this speed is only a little above what the best of those could do. Since some third party upgrades used to run at 2.0GVhz, I have no reason to doubt that it can be done. It's just pushing it a bit.

I frequently reply to the wrong message. Just following the entire content of the thread.

> He makes no statement about its operation under load

True. But I suspect that he tested it under load to make the general statements he does make.

> most operations under MorphOS are not that CPU intensive

That depends on your usage pattern really. It's not the OS itself consuming much CPU time obviously but some applications and games that can really get it running hot. Mind you that there're people using MorphOS as their everyday system, only switching to another OS in case of real need.

> I could see a Mac Mini running under MorphOS at 1.83 exhitbiting fairly
> good stability and only crashing every once in awhile.

pampers has been running his Mac mini at 1.83 GHz from March through September at least, that's about half a year. Seems that during that time he had not one single crash that he'd contribute to the overclocking.

> if the Mac Minis do use a 7447B, then operating at this speed is
> only a little above what the best of those could do.

Rather: ...only a little above what Freescale rated them to do. And as mentioned by me in this very thread, 7447A is rated for the same maximum of 1.7 GHz. But in this context that's moot anyway because Mac mini G4 uses 7447B in fact.

> some third party upgrades used to run at 2.0GVhz

...which is 18% overclocking from 1.7 GHz vs. 22% from 1.5 to 1.83 GHz. Not that much of a difference, especially considering the possibility that Freescale underrated the 1.5 GHz parts (i.e. may possibly have as well been rated as 1.7 GHz).

> I have no reason to doubt that it can be done.

Alright. Sounds different now than just moments ago ;-)

> I frequently reply to the wrong message.
> Just following the entire content of the thread.

My message you replied to doesn't have anything to do with overclocking.

True, but if we held strictly to a thread's original intent or even its overall drift, we wouldn't end up (as we often frequently do) exploring some really informative tangents.

Your posts are as exact as a scalpel. Their precise, factual, and informative. But even you occasionally return to a topic previously mentioned. You've faulted me on this before, but my reply while after your statement wasn't directed exclusively at you.

And you know, I am beginning to realize that your thinking is far more linear and precise than my own.

> if we held strictly to a thread's original intent or even its overall drift

You misunderstand what I say. I'm not talking about holding to the thread topic but about holding (at least in a slight way) to the content of the message you chose to reply to by clicking its 'Reply' button. That's a difference.

> even you occasionally return to a topic previously mentioned.

Yes, and I always reply to an older message where my new message fits, both contentswise and in a formal way (by clicking the proper 'Reply' button).

> You've faulted me on this before

Yes, there. And it's only now that I realize you didn't follow my twofold advice to use threaded view mode to see what I'm about.

> but my reply while after your statement wasn't directed exclusively at you.

It was in *no* way directed at my message you replied to, and yet you specifically chose to reply to it by clicking its 'Reply' button.

> I am beginning to realize that your thinking is far
> more linear and precise than my own.

Just switch to threaded view mode and try to make any sense out of this discussion (or any other discussion where certain participants click random 'Reply' buttons or use the buggy one).

Cool I've never explored that option, although I've seen you mention it more than once. I'll use it.

> Cool I've never explored that option [...]. I'll use it.

Threaded view is nice for better coherency, yes. But, as indicated, here on MorphZone there're currently some problems with this which prevent me from using it exclusively:

#1: Some people use 'Post Reply' button at the bottom of the page, which is faulty, i.e. generates a reply to the most recent message of the thread instead of the thread's opening message.
#2: Some people use one reply message to answer more than one messages from different authors.
#3: Threaded view mode takes some time to be displayed, especially with long threads.

> Rather: ...only a little above what Freescale rated them to do. And as mentioned by me in this very thread, 7447A is rated for the same maximum of 1.7 GHz. But in this context that's moot anyway because Mac mini G4 uses 7447B in fact.

I'm a little bit confused...Morph System Monitor shows 7447A/1500.
(just starting to move on in the PowerPC fog...)

> I'm a little bit confused...Morph System Monitor shows 7447A/1500.

Extensive explanation over there:

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6993&forum=3&post_id=77226#77226

Jim:

There was no problem at all with my overclocked MacMini under MorphOS and under OSX, with light and heavy load. You can ask Ruud how is the Mini getting on - he bought that for me.

I also will be overclocking my PowerMac. I've got spare processor board and water cooling, so hopefully at some stage in few weeks time I will keep you updated :)