I can't see where Intel Power Gadget shows the M7 as 1.2? The solid line representing base frequency looks like it is on 1.3 for the M7 and 1.2 for the M5 and 1.1 for the M3 in all charts (referring to the Notebookcheck review linked in post#121).
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Any benchmarks for the core m7?
- Thread starter SR71
- Start date
- Sort by reaction score
I can't see where Intel Power Gadget shows the M7 as 1.2? The solid line representing base frequency looks like it is on 1.3 for the M7 and 1.2 for the M5 and 1.1 for the M3 in all charts (referring to the Notebookcheck review linked in post#121).
If you install the Intel Power Gadget and let it write to a log file, you'll notice that the base frequency is always at 1.2 GHz regardless that it's an m7 @ 1.3 Ghz. That's because Apple decided to slightly overclock the m7.
Intel's Gadget is based on the OEM clock speeds.
OK, not sure why that invalidates the Notebookcheck review?
EDIT, that table also shows the M3 base as 0.9 and the M5 as 1.1, so presumably same is true for them?
I can't verify the m3 and m5 speeds myself, but I assume so.
Regarding the review: it states the m7 is slower because it throttles quicker, but as long as the readings are not 100% correct it's difficult to tell if that statement is valid.
Regarding the review: it states the m7 is slower because it throttles quicker, but as long as the readings are not 100% correct it's difficult to tell if that statement is valid.
Last edited:
My comment about that review eliminating M7 envy was tongue-in-cheek, hence the smiley. Maybe time will prove it to be misleading, but I suspect even if the absolute values change (which is likely to apply to M3 and M5) there will be an element of truth in the conclusions.
Before reading that review I had expected the M7 would be better in all situations, and by more than 0.1Ghz because the turbo boost difference is 0.4 Ghz which applies a lot of the time. I had even been considering trading up. I am relieved to find that the M7 benefit is not so clear cut and there may even be some scenarios where the M5 wins out. Enough doubt to remove all thought of trading up.
Here is a quote from Anandtech:
However, it gets more confusing: Anandtech has put together a table listing the Intel HD Graphics 515 max. dynamic frequency speed at 950 MHz. Intel's own ARK lists the max. processor graphics at 1000 for the m7. The Intel Power Gadget reports GPU speed at 950 MHz during gaming. It's a mess!
I'll run more benchmarks this really bothers me now!
However, it gets more confusing: Anandtech has put together a table listing the Intel HD Graphics 515 max. dynamic frequency speed at 950 MHz. Intel's own ARK lists the max. processor graphics at 1000 for the m7. The Intel Power Gadget reports GPU speed at 950 MHz during gaming. It's a mess!
I'll run more benchmarks this really bothers me now!
Last edited:
I created a slide based on my benchmark runs to visualize how throttling works. Used a log file from Intel Power Gadget while running Cinebench R15.
I should also mention that this slide was entirely created on my tiny rMB 12".
Yes, you can absolutely work on this lightweight, yet powerful, machine.
Last edited:
Interesting and impressive! I had a quick go at doing the same on my M5 but can't find the IPG log....I have it turned on in IPG prefs but can't see it in the default /var/ or on desktop if I change location to desktop.
You might be interested in my post #11 in this thread where I monitored HIARCS Deep Chess in analysis mode (uses 350% CPU) with IPG to see how throttling worked.
EDIT: IPG log working now. Wish I had it when doing my HIARCS throttling tests.
Last edited:
There is something odd about the Notebookcheck review multicore Cinebench result for the M5 MacBook.
The Cinebench CPU value shown in the table of results for the M5 is 260. I have tested mine several times, including from cold, and have never got higher that 243. The Techradar review of the M5 gives a Cinebench score of 237.
The value shown for the M7 in the table is 251, identical to your result in your chart above (post#132). This is a credible amount more than my M5's 243 or Techradar's 237.
It would be a bit of a headline shock news if the M7 was really slower than the M5 as this review says, and Geekbench certainly isn't slower.
The 22 run/1 hour chart in the review would look very different if the M5 line was lowered so that the initial M5 result was 243 instead of 260.
So while I am prepared to believe the M7 may have a tendency to throttle off its peak more rapidly, I think all the multicore tests in the review undersell the M7........either that or I have a duff M5, but my M5 results are consistent Techradar and with the published Geekbench scores and the IPG behaviour shows it running at spec GHz.
Last edited:
lol i got the m3 3 days ago and now i sold it, i do 300 dpi image for tshirt in photoshop and i stupidly thought the core m3 was enough. it wasn't. now im thinking about the core m5. no core m7 being sold in my country.
i ready sacrifice performance for ultra portability, but the core m3 feels really slow in 300dpi photoshop editing. is the core m5 gonna be much better?
btw i cant return the product i bought in my country, so i had to sell it in local online marketplace for $150 less than the price i bought it only after 3 days of usage, it sucks.
i ready sacrifice performance for ultra portability, but the core m3 feels really slow in 300dpi photoshop editing. is the core m5 gonna be much better?
btw i cant return the product i bought in my country, so i had to sell it in local online marketplace for $150 less than the price i bought it only after 3 days of usage, it sucks.
"300dpi" does not tell us anything about the size of the image you are printing, only the setting you are using for printing. 300dpi is the normal one for printing photos. Assuming a T shirt image is not likely to be bigger than 18 x12ins, then a 300 dpi image would be 3597x5395 pixels, i.e. 55.5 mb. I have an M5, and I have just been playing around with an image this size, sharpening, gaussian blur etc, and everything is instant. The well known Retouch Artists speed test for Photoshop with 20 history/4cache is 40 seconds, which is faster than my original Mac Pro. I have previously done some work with very large files (over 1Gb) with my M5 and it handled them OK. I would not have thought the M3 would be so much slower
When you say printing is slow with your M3, what exactly was the problem? How were you sending the image to the printer? What were you using before you got the M3?
Last edited:
yes it's not bigger than 18x12 inch. what felt slow on the m3 was using 2 finger to move within the image. like when you hold space in photoshop to move around the image. it felt jerky and it's not enjoyable to work with.
i dont print my work, i only design the tshirt. printing is the other division job.
btw, i tried working the image in my friend's 2015 11" air (1.6ghz) and i can comfortably working with that, but the lack of retina display is really annoyance since i work on the computer at least 6 hours a day so retina display is a must as it's easier on my eyes.
heard the m5 is faster or at least has the same speed as the 1.6ghz 205 air am i right? but the air has less pixel to drive.
Not quite sure what the two finger moving within the image is, but I tried moving the 18x12x300 image around the screen with the space bar held down in Photoshop. It is a bit jerky if you move it fast, but I wouldn't have thought twice about it if it wasn't for this thread. However I also did the identical thing on my late 2013 maxed out quad-core 15" rMBP which has a dedicated graphics card....and it was just the same. So if we are talking about the same issue, and if a 11" MBA does not do it, then it may be a retina thing.
Things that don't matter for casual use become important when you work on a machine for six hours a day so I guess you will need to try out on any machine you are thinking of buying.
it's jerky too in my 2015 air, but the core m3 was worse, not only jerky but the movement was really lagging and that lag made it unusable. i can deal with jerkiness in moving image but not with lagging.
i guess im waiting to try the new base mbp 13" in store first before getting the core m5. as much as i love the power in the new mbp, i really love the lightness and the shape of the macbook, it's so sexy lol.