Forum Discussion
MicrosoftNew blog post on performance improvements in Microsoft Edge 81
Hi All,
We posted a new blog article today that talks about some of the optimizations we have made in Microsoft Edge 81. Swing over to the blog site, https://blogs.windows.com/msedgedev/2020/02/13/performance-optimizations-edge-81/ and read more about it, then swing on back and let us know if you are seeing improvements, or have comments.
Thanks – Elliot
1st gen Surface Laptop i5 4GB RAM, I did it 7 times
Edge "stable" v80.0.361.50
1st run: 65.9
2nd run: 69.03
3rd run: 69.7
4th run: 69.99
5th run: 69.69
6th run: 69.5
7th run: 70.13
Dev v81.0.416.3
1st run: 71.7
2nd run: 79.9
3rd run: 80.4
4th run: 82.68
5th run: 82.43
6th run: 82.37
7th run: 82.7
I just got version 82. Really fast results on the Speedometer site. Normally I use Vivaldi and get around 110. Now 137 with Edge.
- The benchmark data looks promising. hope Edge will stay one step ahead from all other browsers, specially Chrome.
Elliot Kirk Its definitely faster than it used to be. On battery saving power modes, the speed was reduced drastically. I ran each one 3 times and restarted the browser every time I changed power modes.
Best Battery(power saving mode):
45.6
46.3
47.1Better Battery:
46.1
46.1
47.1
Better performance:
58.3
58.8
53.7
Best performance:
64.5
65.2
59.0Best performance plugged in
71.4
73.3
72.4- So it basically measures how good the browser can utilize CPU.
the better your CPU, the more score you get, otherwise why should the scores change that much with different power saving modes.
Elliot Kirk I have Canary 82.0.423.0 on a Surface Pro X, results are in the 36-38 range using the benchmarking software. I know that ARM is a very small sliver of the overall market, but when are we going to start seeing some love? Or is the benchmarking software off on ARM?
#ArmProblems
- Elliot Kirk
Hi tboggs, are you running the native Arm64 version of the x86 version. You should see a good bit of performance gain going from the x86 to the native client. Thanks - Elliot
I was also informed that we have enabled some of the optimizations on the ARM64 builds, so you should stick with it. you will see some gains, but probably not as much as on x64.
The current Edge Dev release is certainly the winner for me!
Edge Dev 81.0.416.6 (64 Bit) - Mean 137.1
Edge 80.0.361.54 (64 bit) - Mean 117.0
Firefox 73.0 (64 bit) - Mean 98.0
Elliot Kirk I got 58~ on beta and 45 on canary. Something aint right 😛
Jquery was very slow I noticed as well.- Oh, turning off dark reader gave me 133.3.
But I like making all things dark. Guess I'm stuck with a slow web 😕- Elliot Kirk
Pylot_Light, I understand the call to the Dark Side. I am glad that you can speed things up if you need to though! - Elliot
good improvementElliot Kirk
Just wondering if these performance improvements should be noticeable on macOS as well, I wasn't sure based on the post description talking about the Windows 10 version.
It looks like an opposite event and my results to speedometer 2.0 are nowhere consistent with the rest of the world. could be my Machine, My Internet, VPN? But everything is different on my Machine!
Version 80.0.361.62 (Official build) (64-bit) -108.xVersion 82.0.432.3 (Official build) dev (64-bit) -102.x
esarhaddon It could be all 3 of those things. Cheaper computers will be slower. VPNs create latency in the network, and faster networks will give better results than slower ones.
cjc2112 I would encourage you to ask what a person is using before you make comments about cheap computers. I had an Asus Z390e gamer with 32 gig Gskill DDR4-3600 and 3 M.2 NVME drive with 300 MBbs D/l internet. NOw I have an ASUS Z490e Gamer with the same other components as above and for some stupid reason, I don't think this is cheap junk!
Even with My VPN, I am running in excess of 180 Mbs D/L speeds and in excess of 300Mbs W/O and you are telling me that you have stated something relevant as to why I can't stream videos W/O buffering?
- Elliot Kirk
Hi esarhaddon, it is also important to turn off, or disable all extensions. These can make a huge difference to the speed of the browser. Just look at what Pylot_Light saw in their posts above. Thanks - Elliot
I have noticed that uBlock Origin has been cutting down on my speedometer 2.0 numbers, but only in edge. I haven't encountered the same differences in other browsers like Vivaldi or Brave. And this was using either version of uBlock Origin - the one on the Microsoft store or the one from the Chrome store.
