-
Notifications
You must be signed in to change notification settings - Fork 3.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
High CPU / power consumption with Groovy language server + others (MacOS Intel FPS) #16729
Comments
Since a couple of days ago, I'm having similar issues. Edit: adding system information |
I'm having similar issues too, I disabled language servers and still the consumption remained high. I rolled back to 0.148.0 which doesn't seem to have the issue. System info: |
@SchiavoAnto Yes you might be right, that it affects more languages than I thought in the beginning. I started to notice the same issues with Go. |
Same here with rust. With the latest releases, my Macbook went in full overdrive, even with the lid closed, three times so far, completely draining the battery in the process. I have to quit zed now before closing the lid. It is not the zed process creating the load though, always processes like kernel_task and Quick Look Satellite. But only if zed is running.
|
@mdelete Oh yeah the overheating with the lid closed happened to me too. I was using Zed and closed the laptop, like I always do, and found it very very hot and the battery was almost completely flat. |
Since Zed 0.149 my MacBook was getting hot, even with no LS running. I tried to leave Zed open with just a small txt file and all my personal settings removed and it's draining the battery in no time. Activity Monitor shows relatively high CPU + GPU usage. I went back to Zed 0.148 and this issue disappeared. So, there must be a serious regression between 0.148 and 0.149. |
There was quite a bit that happened between v0.148.x and v0.149.x (277 commits) as it was leading up the Zed.ai launch. Similarly a lot that has happened in the intervening time [between v0.149.x and v0.151.x] (200 commits). Are folks still seeing this issue with current releases (Zed Preview v0.151.x)? If yes, would anyone experiencing this be willing to do the git bisect from a working build to v0.149.x to help track down the offending commit? |
This comment was marked as off-topic.
This comment was marked as off-topic.
Just tried 0.151.1, the Mac ran hot within 30 minutes. I left Zed running with an empty window. No other apps were running except Activity Monitor. Activity Monitor is confusing: Zed 0.148: Zed 0.151 There's almost no difference! I don't even know how to measure the power drain. If I didn't trust you I'd say you're running a hidden process to train your next AI model 😉 😂. I'm happy to investigate as long as I don't need to build Zed (and install all the tools to do that). |
@notpeter I can build from source and give that a try but like @piechologist said I have no idea how to measure for the bad commit apart from feeling my Mac and listening to the fans running. I also don't see any differences between versions in Activity Monitor. Any other ideas would be appreciated! |
Hi, just to make basic triage:
|
Browsing through this issue and adding my machine:
I tried installing via homebrew (brew uses the intel binary) and by downloading the intel binary from zed.dev. |
I have the similar issue using js/ts
|
@notpeter it took longer than I thought but this was my experience with different commits on my Mac
|
I'd like to identify if there is a common pattern of hardware impacted by this. If you are experiencing this, please run the following and paste with a comment below and I'll collate the results and hide the comments.
|
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This reverts the FPS counter added in 1175391 (#16422) because in this issue someone bisected recent performance regressions down to this commit: - #16729 Another issue that's possibly related: - #17305 (comment) We're reverthing this in a PR to create a bundle that people can try out. Co-authored-by: Bennet <[email protected]>
Thanks @d-shah8. @bennetbo and I created a build (in https://github.com/zed-industries/zed/actions/runs/10735702994?pr=17485) for you all to try out:
In that build, the FPS counter code has been reverted (we couldn't spot an obvious reason for why it would decrease performance). Can you try out the Intel build and see if it fixes the problem? |
@mrnugget I am trying the Intel build and it looks like the issue is fixed |
@mrnugget yep issue seems fixed with that Intel build for me too. |
Ahhh... silence of the fans. Seems to fixed for me too (Intel build). |
Intel build looks good for me too. Thanks a lot! |
x86/Intel build looks very well. I wrote in a pull request comments at first, sorry. |
**UPDATE**: Response so far seems to be that this fixes the performance issues on Intel MacBooks. So we're going to go ahead and merge it. This reverts the FPS counter added in 1175391 (#16422) because in this issue someone bisected recent performance regressions down to this commit: - #16729 Another issue that's possibly related: - #17305 (comment) We're reverting this in a PR to create a bundle that people can try out. Assets: - Universal Binary: https://github.com/zed-industries/zed/actions/runs/10735702994/artifacts/1900460781 - x86/Intel: https://github.com/zed-industries/zed/actions/runs/10735702994/artifacts/1900461236 - Apple Silicon: https://github.com/zed-industries/zed/actions/runs/10735702994/artifacts/1900460978 Release Notes: - Removed the recently-added FPS counter since the changes it made to the Metal renderer on macOS could lead to performance regressions on Intel MacBooks. Co-authored-by: Bennet <[email protected]>
**UPDATE**: Response so far seems to be that this fixes the performance issues on Intel MacBooks. So we're going to go ahead and merge it. This reverts the FPS counter added in 1175391 (#16422) because in this issue someone bisected recent performance regressions down to this commit: - #16729 Another issue that's possibly related: - #17305 (comment) We're reverting this in a PR to create a bundle that people can try out. Assets: - Universal Binary: https://github.com/zed-industries/zed/actions/runs/10735702994/artifacts/1900460781 - x86/Intel: https://github.com/zed-industries/zed/actions/runs/10735702994/artifacts/1900461236 - Apple Silicon: https://github.com/zed-industries/zed/actions/runs/10735702994/artifacts/1900460978 Release Notes: - Removed the recently-added FPS counter since the changes it made to the Metal renderer on macOS could lead to performance regressions on Intel MacBooks. Co-authored-by: Bennet <[email protected]>
**UPDATE**: Response so far seems to be that this fixes the performance issues on Intel MacBooks. So we're going to go ahead and merge it. This reverts the FPS counter added in 1175391 (#16422) because in this issue someone bisected recent performance regressions down to this commit: - #16729 Another issue that's possibly related: - #17305 (comment) We're reverting this in a PR to create a bundle that people can try out. Assets: - Universal Binary: https://github.com/zed-industries/zed/actions/runs/10735702994/artifacts/1900460781 - x86/Intel: https://github.com/zed-industries/zed/actions/runs/10735702994/artifacts/1900461236 - Apple Silicon: https://github.com/zed-industries/zed/actions/runs/10735702994/artifacts/1900460978 Release Notes: - Removed the recently-added FPS counter since the changes it made to the Metal renderer on macOS could lead to performance regressions on Intel MacBooks. Co-authored-by: Bennet <[email protected]>
This fix is available in Zed Preview v0.152.1 and Zed Stable v0.151.2. Thanks for everyone who helped track this down and please accept our apologies for the two weeks of battery drain. @amaridev Apologies for hijacking your issue. As the original reporter you reported this on aarch64 while most of the participants in this discussion are seeing what I believe is an Intel x86_64 specific issue. If these builds don't solve your problem, please open a clean issue so we can track that separately. Thanks all! |
Check for existing issues
Describe the bug / provide steps to reproduce it
Hi,
I started to use zed recently and like it pretty much so far. I work a lot with Groovy and Jenkins Pipelines and face some issue with the Groovy plugin, I assume.
I already noticed that the syntax checking and auto format doesn't really work on Groovy files for me. Now I found that Zed sucked my battery empty in less than 2 hours. I am working on a Macbook Pro M1 Max.
I am not sure yet if it's a bug or some problem just on my machine. Hope to get some help figuring out where the issue is.
Environment
Zed: v0.149.5 (Zed)
OS: macOS 14.4.0
Memory: 32 GiB
Architecture: aarch64
If applicable, add mockups / screenshots to help explain present your vision of the feature
If applicable, attach your Zed.log file to this issue.
Zed.log
The text was updated successfully, but these errors were encountered: