one drawback on the M1 system:
when the computer sets to sleep the ethernet port is closed (and as far as I can see Thunderbolt ports too) so the network connection between FMClient and FMS is cut - annoying.
I talked to Christian this afternoon and he told me I could circumvent this with some nifty MBS function. But this has yet to be tested...
Interesting results Markus.
The number of cores isn’t usually a factor with these type of tests. Usually a processor with higher Ghz, but less cores would normally be expected to be quicker than a slower processor with more cores.
As you say, the M1 continues to impress and I can’t wait to try the native FileMaker when it is released and, as @Torsten mentioned, native FileMaker Server will be interesting, where the cores will really come into play.
Thanks everyone for contributing to the testing. I’ve some heavy grunt work scheduled for next week, so will add time tests to this and run alongside the MacBook Air.
I agree although on the Mac mini showing Big Sur on the display was the best experience I have had over the last years: Philips 43'' 4K Monitor was recognized on startup, no fall-back to lower resolution on the big display like on older Macs in frequent years...
FMP 19.1.3, 2017 MacBook Pro 16 GB, 2.9 GHz 4-core i7: 4:29 (local)
FMS 19.1.2, 2012 Mac Mini Server 16 GB, 2.3 GHz 4-core i7: 13:39 (PSOS, wait for completion On)
FMS 19.1.2, 2012 Mac Mini Server 16 GB, 2.3 GHz 4-core i7: 12:47 (PSOS, wait for completion Off)
FMS 19.1.2, 2012 Mac Mini Server 16 GB, 2.3 GHz 4-core i7: 14:16 (scheduled script)
The 2012 Mac Mini Server has a slower i/o subsystem, which seems to be the limiting factor here.
All test have been executed only once.
All machines on MacOS 10.15.7
that seems to be quite slow, how much capacity does the SSD have?
It has 1 TB and 810 GB available.
| Device Name: | Samsung SSD 860 EVO 1TB |
|---|---|
| Media Name: | AppleAPFSMedia |
| Medium Type: | SSD |
| Protocol: | SATA |
| Internal: | Yes |
| Partition Map Type: | Unknown |
| SMART Status: | Verified |
MacBook Pro 2017 with internal SATA SSD? - astonishing
Sorry, I did not read carefully. The MacBook has a 500 GB SSD. 250 GB available.
You can try ProcessActivity.beginActivity function and see if your solution can prevent the sleep while you are using it.
What is interesting:
my MacBook Pro 2016 2,9 GHz Intel Core i7 16GB Ram macOS 10.12.6 FMPA 17 03:04min
compared to your setup with Catalina and FMP 19 using plus 50%
Where is the reason behind this? MacOS version, FM version, a mixture of both?
comparing the results (fm-test) - Apple seems to be on the right path concerning cpu...
Frankly, I have no idea and did not test against FMP 17.
just to add one more figure
FMS 19.1.2, Intel NUC NUC8i3BEH2 Core i3-8109U, 32 GB, 240GB SATA SSD, centOS 7.9: 08:14 scheduled script
but I ordered an NVMe SSD (black friday...) and maybe this can get a bit faster next week
Thank you so much for your joureney and text file.
My Late 2013 2,6 GHz Intel Core i7 15" Macbook Pro takes 3'29"
This makes your M1, under rosetta, 1,77x faster than my Mac, with native code we could expect a x2 speed gain, that's very promising
It would be very interesting to test this on your M1 with Filemaker GO
there is a report (in German): FileMaker Forum-Beitrag: iOS SDK App auf dem Mac mit M1 Processor – erste Testergebnisse :: FileMaker Magazin
yep, but what would be interesting is to launch the same file's test on same M1 machine, so we get a comparison point
Great idea @Vincent_L. I;m on it, but so far it looks to be very, very slow or something has gone wrong. I’ll also test it on my 2019 11” iPad Pro as well.
Update, the first test result so far is the first M1 system error after leaving it running for quite a long time. I’ll try again and check there is nothing in the script that is not FMG compliant.
Just resetting all the display preferences now 
OK, FM Go got to 308,646 records, but it was running for a good 20 minutes before it blew out. So not particularly optimistic.
That's 2 system errors resulting in a restart of the MacBook Pro, this time only 4080 records have been created. The test is still running on the iPad Pro, but it seems FileMaker Go's performance doesn't take advantage of the potential processing power available.
For the record, I've attached the 2nd crash report.Crash Report.txt (11.3 KB)