I recently purchased a new Studio but am still getting spinny wheels on some slower apps.
Ran Geekbench (with other apps off) and got scores:
3665 Single-Core Score
23527 Multi-Core Score
The comparison scores from Geekbench for the Mac Studio are
4017 and 26059.
That's about a 10 percent difference. Should I consider that 'acceptable' or should I be wondering if my Studio needs to be checked over?
Mac Studio, macOS 26.0
sentilinadho wrote:
I have a VPN but obviously closed it before running the bench test.
Closing it may not be enough. You may need to remove it completely. Unless you need it to access a secure server for work or school, it's possibly doing you more harm than good anyway.
sentilinadho wrote:
I have a VPN but obviously closed it before running the bench test.
Closing it may not be enough. You may need to remove it completely. Unless you need it to access a secure server for work or school, it's possibly doing you more harm than good anyway.
Consider downloading and running this little "discovery" utility, Etrecheck. It changes NOTHING. Etrecheck was developed by a senior contributor here, and uses mostly system calls and simple tests to collect often-needed information.
it contains little tests for speeds of devices, CPU utilization, memory usage, energy usage and a digest of recent problems, in one easy to use package. it does not even need to be Installed. Because less can be learned when your Mac is running great, best time to run is when your problems are actually occurring, if possible.
if you follow the directions faithfully, its report (pre-laundered of all personally-identifiable information) can be "Shared" to the System ClipBoard, then Pasted into an ‘Additional Text’ window in a reply on the forums.
Use Etrecheck Pro for free:
The amount of data you get can be daunting. If you POST your report, some Readers here are willing to look over those reports, and can provide valuable insights.
the start a reply on the forums, click the "additional text" icon, and PASTE
<< wondering if my Studio needs to be checked over? >>
Genius Bar:
if you take it to an Apple service provider first, they CAN run their diagnostics, which are a bit more comprehensive than the User diagnostics. In the very small amount of time they are run, “No faults detected” still does not mean “All is Well” but at least there are no detected GROSS problems. Unless they can find a fault, they will return your computer unchanged. Genius Bar technicians are NOT trained to read panic reports and computer logs.
When you present your computer at the Genius Bar with the expectation of a quote to repair it, you must generally meet these criteria:
• your Mac won't do anything
• your Mac fails diagnostics
• your Mac has a problem you can repeatably demonstrate on demand
one other oddball case:
• Apple online support has DIRECTED a Service Provider to replace specific components inside your Mac, such as the mainboard or display. Bring proof, such as case number.
Your appointment at an Apple owned store is free, in warranty our out. But If you don't have an obvious issue that can be replicated on demand, and it passes user diagnostics, you are likely wasting your TIME taking it to the Genius bar. They will likely just run their service diagnostics and return it to you, unmodified.
Any perceived slowdowns are likely due to software you added. That is why I led with, ¿Did you add any of the most resource-intensive junk-ware categories known to slow down an otherwise perfectly good Mac?
Since you seem to think you might have a problem, Readers could look over your Etrecheck report to see if there are other festering problems.
<< Much as Etrecheck, which you seem fond of, looks interesting, I'm not planning to download it, nor do I have info to see if it actually provides a standard benchmark for a Studio M4 against which I could compare our new Studio M4. >>
Etrecheck would be of very limited usefulness if all it did was compare to standards using benchmarking tools. Instead, it runs a number of different tests and uses some built-in system tools to get a broad picture of how various parts of your Mac are performing, and what software is dominating performance.
If you want to see if added software is limiting performance on your Mac (and nothing more) you could run the standard benchmark from Safe Mode.
Safe Mode does a number of different things. On older Intel Macs, hold shift at startup, but have your userid and password at the ready.
A parade of unusual things happens.
• Your Mac loads just enough of the kernel to do a disk check. Then it proceeds to do a disk check. This can take an extra about five minutes.
• your userid and password are required, even if you normally auto-login. So have them handy.
• Your Mac adds ONLY a minimal set of Apple-Only extensions, Not including graphics acceleration extensions. Screen updates will therefore be wonky and slow, but it ultimately should be correct.
• Your Mac assumes defaults for as many settings as possible, including screen resolution. This is the key for re-setting the screen, but there is a little more to it: Resolution is likely to be temporarily lower and settings ordinary. Use this as a starting point to customize settings to your liking.
If you make NO changes, regular mode will just revert to what you had set before, so DO make changes!
Any changes you make in Safe Mode will "stick" in regular mode after you restart.
• after a Restart in normal mode, your Mac will take slightly longer to start up [once] because it rebuilds some system caches.
"Works in Safe mode, fails in regular mode" implies "It's something you added".
How to use Safe Mode on your Mac
Start up your Mac in safe mode - Apple Support
By far the easiest way to cause poor performance, instability, overheating and crashing is to install ANY third-party speeder-uppers, Cleaners, Optimizers, or Virus scanners, Bit Torrent, or a VPN that you installed yourself.
¿Do you have anything like that installed?
sentilinadho wrote:
I understand about VPNs. I don't think a disconnected VPN would cause a 10% bench mark difference. If there is evidence to contrary I will change my opinion.
You don’t understand about VPNs. Just have a VPN profile installed, even it not turned, can cause serious performance hits, and even break connectivity entirely.
Here is some information to help you understand VPN: What you need to know - Apple Community
sentilinadho wrote:
I understand about VPNs. I don't think a disconnected VPN would cause a 10% bench mark difference. If there is evidence to contrary I will change my opinion.
As has been pointed out, merely closing it is not enough. You need to remove it.
Getting back to my question, "Is 10 per cent an acceptable bench mark variation from the mean or should get back to Apple?"
You can certainly contact Apple Support at any point. Or, provide feedback here:
sentilinadho wrote:
I recently purchased a new Studio but am still getting spinny wheels on some slower apps.
Ran Geekbench (with other apps off) and got scores:
3665 Single-Core Score
23527 Multi-Core Score
The comparison scores from Geekbench for the Mac Studio are
4017 and 26059.
That's about a 10 percent difference. Should I consider that 'acceptable' or should I be wondering if my Studio needs to be checked over?
I have not personally benchmark tested any Mac's in a very long time.
However, I do recall that a clean Mac with no third party App's and nothing running in the background, will produce a higher benchmark score than one that is in use.
I have a VPN but obviously closed it before running the bench test. What I'm wondering about is whether 10 per cent is an acceptable variation from the mean or whether I should get back to Apple. Years ago I had a MacPro that had to be returned to China and it came back seemingly ok. It was years later when I realised one of the components was still faulty. So if there's anything wrong with the Studio I'd want to return it promptly, not wait for years. Apple M4 Max, 4TB storage (less than half of which is used just now).
VPN:
A virtual private network, or VPN, is a private connection over the Internet from a device to a specific network. VPN technology is widely used in corporate environments. If you need to be "present" on an institutional network, a VPN is a great tool for accomplishing this. It is generally issued and controlled by the institution.
Almost all other uses are a SCAM. There is generally no need for you to have a private (and almost always MUCH slower) connection to a VPN vendor's Network, except to make it easier for them to harvest your data to sell. If you are behind a Router you control or Trust, there is NO security advantage whatsoever in using a VPN. Your connections are already encrypted in most cases.
If VPN vendors just stopped there, it would be bad. But many of these packages also insist on scanning all your files, non-stop, -- nominally looking for viruses, but who knows for sure what data they are harvesting. Their non-stop file reading punishes your computer's performance in the process.
Some also break into your other secure connections so they can be FIRST to examine your data, often leaving your Mac MORE vulnerable to attack.
https://gist.github.com/joepie91/5a9909939e6ce7d09e29
I understand about VPNs. I don't think a disconnected VPN would cause a 10% bench mark difference. If there is evidence to contrary I will change my opinion.
Getting back to my question, "Is 10 per cent an acceptable bench mark variation from the mean or should get back to Apple?"
Thank you for your time gentlemen.
To summarise, your opinion appears to be that the 10% drop from benchmark is due to the existence of a disconnected VPN (Proton, which fwiw has, I understand, just passed its 3rd independent audit confirming a no-logs policy). VPN is part of our I.T. policy for whatever reason on all devices and its proxies simplify various overseas business, so if it has caused a 10% drop from benchmark, so be it.) I mentioned a spinny wheel on some app, but to be more precise, Office 365 for Mac, and my other 'guess' is that it may possibly be connected to a) the size of the document (nearly 2000 pages, heavily formatted and with over 100 in a spreadsheet format) and b) MS repeatedly checking the licence on start up. Much as Etrocheck, which you seem fond of, looks interesting, I'm not planning to download it, nor do I have info to see if it actually provides a standard benchmark for a Studio M4 against which I could compare our new Studio M4.
I may have a chat about this when I next speak to our Business Adviser at the local Apple store. Thank you again for listing their diagnostics and test requirements, always useful to know, though I will more likely be having an informal chat as we do from time to time.
Kindest regards.
An important and frequently overlooked fact is that when you install VPN it configures itself to the specific hardware, OS version and network that the device uses. If any of those change it can break the VPN. So any time you change any of these you may need to delete the VPN app and profile, restart your device, then add them back (if you still want to use VPN) so it can configure itself to the new environment.
<< Office 365 for Mac, and my other 'guess' is that it may possibly be connected to a) the size of the document (nearly 2000 pages >>
Word for Mac has always had tools for producing documents that large as a "chapter-book", with each smaller chapter editable independently.
Word can calculate page numbers and index entries on the composite document, stored a series of chapters linked together. Each chapter mentions the local document that follows as the succeeding chapter. Simple but effective at making it so that you only need to edit one chapter, not the entire world.
EDIT: You are NOT stuck with page numbers of the form 12-57, unless you choose to do it that way.
↓Word for Mac has always had tools for producing documents that large as a
"chapter-book", with each smaller chapter editable independently. ↑
That's more or less what I do. I found with editing previous books that anything over 800 pages at once was giving my old iMac indigestion. The spreadsheet section of the current one seems the most resource-demanding, even run independently. Word probably isn't designed for spreadsheets where a single spreadsheet runs for several hundred pages of Word. There are probably better ways of formatting it: but it is what it is, and fortunately the Studio handles it even if there is a fraction of a second delay when editing! I'm still open to looking at possibly ways of doing the indexing, which involves several languages. They are rather large, as are the chapters, so I'm having an index for each chapter, subdivided automatically. The full work could probably just have an index of headings, together with the bibliography.
Mac Studio M4 low benchmark?
