Chunk loading is slow, causing lag spikes, void walls, and lighting errors.

Same on my survival world on ipad pro 2018

Same on my survival world on ipad pro 2018

chunk generating is slow. loading is not

chunk generating is slow. loading is not

same here on win10. takes around half a minute to join, chat is slow and chunks load overly slowly, even on fresh world

same here on win10. takes around half a minute to join, chat is slow and chunks load overly slowly, even on fresh world

i felt like i should probably post some more info:

freshly generated worlds are somewhat laggy, though fine after some time (didnt check for long last time)

i have a 300mb world, and no matter where i put it, server, multiplayer, singleplayer, the chunks have issues loading, though on singleplayer and multiplayer invite, it's fine.

i usually host the world on a bedrock dedicated server, and the same issue is over there.

i felt like i should probably post some more info:

freshly generated worlds are somewhat laggy, though fine after some time (didnt check for long last time)

i have a 300mb world, and no matter where i put it, server, multiplayer, singleplayer, the chunks have issues loading, though on singleplayer and multiplayer invite, it's fine.

i usually host the world on a bedrock dedicated server, and the same issue is over there.

Alright, so i did some testing with around 5 people flying around a world at once.

Around 400mb in, and chunk loading started to slow down to a point that it was similar to shown in the video, 60mb later, at 460mb, we started to run into invisible walls in the void.

After reaching around 490mb, we started to run into invisible walls every few seconds, with barely any time to reorient ourselves before hitting another one. At this point we all headed out.

I still have the world, and have left a link to it below. Be warned that some idiot friend of mine burnt a bunch of villages along the way and may have left a sign or two, so be warned.

I feel like when a world has buildings, it's most likely a lot more complex, so therefore will start lagging at a smaller size. This one was just pure land generation, nothing more.

I'll test this world every update and try to report the results here until it's fixed.

Thanks to Kotikissa, Durable, doh007 and KoleckOLP for spending an hour flying around a world with me. 🙂

Download world

Alright, so i did some testing with around 5 people flying around a world at once.

Around 400mb in, and chunk loading started to slow down to a point that it was similar to shown in the video, 60mb later, at 460mb, we started to run into invisible walls in the void.

After reaching around 490mb, we started to run into invisible walls every few seconds, with barely any time to reorient ourselves before hitting another one. At this point we all headed out.

I still have the world, and have left a link to it below. Be warned that some idiot friend of mine burnt a bunch of villages along the way and may have left a sign or two, so be warned.

I feel like when a world has buildings, it's most likely a lot more complex, so therefore will start lagging at a smaller size. This one was just pure land generation, nothing more.

I'll test this world every update and try to report the results here until it's fixed.

Thanks to Kotikissa, Durable, doh007 and KoleckOLP for spending an hour flying around a world with me. 🙂

Download world

also i have an simple texture bug with no lag ithe chunks need to be refreshed

also i have an simple texture bug with no lag ithe chunks need to be refreshed

@Jax logan In my experience both generated chunks and new chunks load at the same speed in my testing

This was tested at 12 chunks render distance, and 8 simulation distance on Windows 10 so far.

@Jax logan In my experience both generated chunks and new chunks load at the same speed in my testing

This was tested at 12 chunks render distance, and 8 simulation distance on Windows 10 so far.

Some loading lag can be reduced by lowering your render distance or the world's simulation distance. When commenting, please mention what you have these set on.

Some loading lag can be reduced by lowering your render distance or the world's simulation distance. When commenting, please mention what you have these set on.

This ticket replaces MCPE-19308, which reported similar slow chunk loading in release 1.0 to 1.2. Those problems were mostly corrected by tuning efforts, and I espect that will be what we need in this case as well.

 I have experienced this slowness on Windows 10 while sprint flying in creative mode. It's especially slow over the ocean. (Ocean chunks seem to take longer to generate than terrain chunks since the Update Aquatic). It is possible for me, at full speed, to fly over a void space before the generated chunk gets rendered.

This ticket replaces MCPE-19308, which reported similar slow chunk loading in release 1.0 to 1.2. Those problems were mostly corrected by tuning efforts, and I espect that will be what we need in this case as well.

 I have experienced this slowness on Windows 10 while sprint flying in creative mode. It's especially slow over the ocean. (Ocean chunks seem to take longer to generate than terrain chunks since the Update Aquatic). It is possible for me, at full speed, to fly over a void space before the generated chunk gets rendered.

For me, changing the render distance and the simulation distance seem to have little to no effect in terms of chunk loading and chunk generating (for me they both seem to be around the same speed), on windows 10, whilst flying in max speed, chunks keep up for a while, until they start to slow down, and you are nearly flying over the void, and you sometimes hit a wall, as soon as you add another player, you are both constantly hitting chunk walls whilst flying into new chunks and sometimes already generated chunks.

On Nintendo Switch, chunk walls are hit a bit faster, but other then that it seems to have similar results to the windows 10 version. In multiplayer the same issues occur.

These were both tested with worlds being relatively small at around 10MB. On larger worlds, the impact is severe in multiplayer (and also in singleplayer chunks load slower than a fresher world), being that neither devices can load any chunks up to a certain point for minutes.

Here is the link to the 120MB survival world if you want to do more testing: Achievement Get Copy

For me, changing the render distance and the simulation distance seem to have little to no effect in terms of chunk loading and chunk generating (for me they both seem to be around the same speed), on windows 10, whilst flying in max speed, chunks keep up for a while, until they start to slow down, and you are nearly flying over the void, and you sometimes hit a wall, as soon as you add another player, you are both constantly hitting chunk walls whilst flying into new chunks and sometimes already generated chunks.

On Nintendo Switch, chunk walls are hit a bit faster, but other then that it seems to have similar results to the windows 10 version. In multiplayer the same issues occur.

These were both tested with worlds being relatively small at around 10MB. On larger worlds, the impact is severe in multiplayer (and also in singleplayer chunks load slower than a fresher world), being that neither devices can load any chunks up to a certain point for minutes.

Here is the link to the 120MB survival world if you want to do more testing: Achievement Get Copy

@Hasan i experience this too, at the lowest chunk setting, on both the server and client, the same lag still applies. on small worlds its not noticeable at all like i said above, but in bigger worlds it takes minutes, yeah

@Hasan i experience this too, at the lowest chunk setting, on both the server and client, the same lag still applies. on small worlds its not noticeable at all like i said above, but in bigger worlds it takes minutes, yeah

I also have game slower in general.

slow chat and inventory? takes me around 2 seconds to open inv or send message

slow chat and inventory? takes me around 2 seconds to open inv or send message

https://youtu.be/1GhLd91aowI video of the lag i'm experiencing

https://youtu.be/1GhLd91aowI video of the lag i'm experiencing

 Am getting a lot of lag too, can't even fly in a straight line with rockets in single player on min settings without hitting void wall... And I have a decent gaming rig so count that out....

 Am getting a lot of lag too, can't even fly in a straight line with rockets in single player on min settings without hitting void wall... And I have a decent gaming rig so count that out....

Results seem the same on v1.16.1, will do more testing later.

v1.16.1 is not out on other platforms yet.

Results seem the same on v1.16.1, will do more testing later.

v1.16.1 is not out on other platforms yet.

will test my world now.

will test my world now.

finally transferred files to the server (really slow server lol) and tested. i was really excited leading up to this buuut didn't fix anything sadly.

finally transferred files to the server (really slow server lol) and tested. i was really excited leading up to this buuut didn't fix anything sadly.

Problems with performance aren't like normal bugs. After all, it's not that the code is doing something wrong, it's just doing something slower than you want it to. What you have to do to fix a performance problem is tune the code. It's kind of like tuning an engine, except there are probably dozens to hundreds of places where you can tweak it and none of them are marked; you have to find them. Then you have to think out of the box to come up with a replacement part that does exactly the same thing, just faster. And even after you get it working, you may have to do lots more small tweaks, each helping just a little, before they add up to enough improvement to be considered a fix.

My point is that performance problems are even harder to fix than regular bugs. If we're very lucky, one of the devs may have a moment of brilliant thinking that lets them make a major improvement in not too much time. Realistically, though, chances are that this will only be fixed incrementally, as a series of small improvements over a period of time. And some of those improvements may be too small for many people to even notice them. So settle in and be prepared for the long game. This probably isn't something that's going to be fixed in one hotfix release.

Problems with performance aren't like normal bugs. After all, it's not that the code is doing something wrong, it's just doing something slower than you want it to. What you have to do to fix a performance problem is tune the code. It's kind of like tuning an engine, except there are probably dozens to hundreds of places where you can tweak it and none of them are marked; you have to find them. Then you have to think out of the box to come up with a replacement part that does exactly the same thing, just faster. And even after you get it working, you may have to do lots more small tweaks, each helping just a little, before they add up to enough improvement to be considered a fix.

My point is that performance problems are even harder to fix than regular bugs. If we're very lucky, one of the devs may have a moment of brilliant thinking that lets them make a major improvement in not too much time. Realistically, though, chances are that this will only be fixed incrementally, as a series of small improvements over a period of time. And some of those improvements may be too small for many people to even notice them. So settle in and be prepared for the long game. This probably isn't something that's going to be fixed in one hotfix release.

@Auldirck, I understand that these are not necessarily "bugs" and they are engine performance problems, part of this issue (multiplayer chunk loading) was "caused" by the v1.16 update, which suggests that this issue may still be a "bug".

Maybe I could purchase I realm to temporarily avoid this issue to continue playing my survival world.

@Auldirck, I understand that these are not necessarily "bugs" and they are engine performance problems, part of this issue (multiplayer chunk loading) was "caused" by the v1.16 update, which suggests that this issue may still be a "bug".

Maybe I could purchase I realm to temporarily avoid this issue to continue playing my survival world.

@Hasan is this not an issue on realms then?

you mention "This doesn't seem to affect servers and realms in any way, so this may be a problem with regular multiplayer mode.", but the issue is definitely there on my BDS, but not sure about realms.

servers are definitely affected

@Hasan is this not an issue on realms then?

you mention "This doesn't seem to affect servers and realms in any way, so this may be a problem with regular multiplayer mode.", but the issue is definitely there on my BDS, but not sure about realms.

servers are definitely affected

@Hubz, ah sorry, I meant featured servers, I'll change the description now. Also, realms aren't effected to my knowledge, but more testing will need to be done (I can't do any right now, because my realm expired).

Also @Hubz did this affect you before v1.16?

@Hubz, ah sorry, I meant featured servers, I'll change the description now. Also, realms aren't effected to my knowledge, but more testing will need to be done (I can't do any right now, because my realm expired).

Also @Hubz did this affect you before v1.16?

@unknown: The fact that it started with a new release doesn't make it a bug. If that was the case, every new feature would fit the definition of a bug. But my purpose wasn't to correct terminology, it was to supply a rationale for my subsequent argument that bug watchers should not be expecting the prevailing one-bug/one-fix paradigm they're used to seeing in bug reports. I was encouraging them to adopt more realistic expectations about how this fix will proceed.

Also, I'm not sure it's useful to try to distinguish between single-player and multiplayer behaviors in this case. If there's a performance problem in single-player and local multiplayer modes, it's to be expected that networked multiplayer mode, which adds network delays to the problem, will only make it worse, so a difference in degree isn't necessarily noteworthy. On the other hand, if networked multiplayer mode produces additional or different symptoms or an extreme difference of degree, and other comments affirm that experience, it starts to look more like a difference of kind, and in that case we should probably split it off with its own report. I'm not clear on whether you think the networked multiplayer experience constitutes a difference of kind, so feel free to enlighten me on what's unique about it.

One more thing: The featured servers are not maintained by Mojang and don't necessarily use Mojang's server code, so we and the developers are generally going to disregard references to a problem reported here to be affecting them. If a problem like this affects them, the server operators have their own, separate communication channel for reporting it and collaborating with the developers on a fix. I'm not saying you shouldn't have mentioned featured servers, just warning you that fixing that aspect of the problem isn't going to be one of the goals associated with this report.

@unknown: The fact that it started with a new release doesn't make it a bug. If that was the case, every new feature would fit the definition of a bug. But my purpose wasn't to correct terminology, it was to supply a rationale for my subsequent argument that bug watchers should not be expecting the prevailing one-bug/one-fix paradigm they're used to seeing in bug reports. I was encouraging them to adopt more realistic expectations about how this fix will proceed.

Also, I'm not sure it's useful to try to distinguish between single-player and multiplayer behaviors in this case. If there's a performance problem in single-player and local multiplayer modes, it's to be expected that networked multiplayer mode, which adds network delays to the problem, will only make it worse, so a difference in degree isn't necessarily noteworthy. On the other hand, if networked multiplayer mode produces additional or different symptoms or an extreme difference of degree, and other comments affirm that experience, it starts to look more like a difference of kind, and in that case we should probably split it off with its own report. I'm not clear on whether you think the networked multiplayer experience constitutes a difference of kind, so feel free to enlighten me on what's unique about it.

One more thing: The featured servers are not maintained by Mojang and don't necessarily use Mojang's server code, so we and the developers are generally going to disregard references to a problem reported here to be affecting them. If a problem like this affects them, the server operators have their own, separate communication channel for reporting it and collaborating with the developers on a fix. I'm not saying you shouldn't have mentioned featured servers, just warning you that fixing that aspect of the problem isn't going to be one of the goals associated with this report.

@Auldirck, the multiplayer was mentioned because when it is another player joins a (large) world, chunk loading becomes incredibly slow (or ceases to stop) making even walking hard.

@Auldirck, the multiplayer was mentioned because when it is another player joins a (large) world, chunk loading becomes incredibly slow (or ceases to stop) making even walking hard.

Ah, ok. Assuming the new player is at a distance from the first, that would imply that the chunk loader thread is trying to load twice as many chunks as before, so it's just a difference in degree. Thanks for explaining.

Ah, ok. Assuming the new player is at a distance from the first, that would imply that the chunk loader thread is trying to load twice as many chunks as before, so it's just a difference in degree. Thanks for explaining.

After more testing, I am certain this affects large worlds more (tested on switch and windows 10, 120MB survival world yet again, v1.16).

Admittedly, copying the world, then deleting the original and playing the copy did slightly speed up chunk loading (this clip is the copied world).

In smaller worlds, chunk rendering gradually slows down when flying for a while, but here on the large world it slows instantly.

After more testing, I am certain this affects large worlds more (tested on switch and windows 10, 120MB survival world yet again, v1.16).

Admittedly, copying the world, then deleting the original and playing the copy did slightly speed up chunk loading (this clip is the copied world).

In smaller worlds, chunk rendering gradually slows down when flying for a while, but here on the large world it slows instantly.

> Also @Hubz did this affect you before v1.16?

Nope, sure didn't. it was fine before the 1.16 update, now its broken.

> Also @Hubz did this affect you before v1.16?

Nope, sure didn't. it was fine before the 1.16 update, now its broken.

On nintendo switch in local multiplayer (on the same console), especially with 3 or 4 players, chunks loading/generation is extremely slow in the overworld. Even walking is hard and also mobs or minecart are stuck in position when trying to move into unloaded chunks. My world size is approx. 300 MB, but these chunk problems are present also in already generated chunks with player relatively close (50-100 blocks) from each other. I also tried to lower the simulation distance to minimum and disable all video settings, but it does almost nothing. In the nether it is more playable, but even there the chunks load very slowly in multiplayer. This problem was present also in 1.14.X.

On nintendo switch in local multiplayer (on the same console), especially with 3 or 4 players, chunks loading/generation is extremely slow in the overworld. Even walking is hard and also mobs or minecart are stuck in position when trying to move into unloaded chunks. My world size is approx. 300 MB, but these chunk problems are present also in already generated chunks with player relatively close (50-100 blocks) from each other. I also tried to lower the simulation distance to minimum and disable all video settings, but it does almost nothing. In the nether it is more playable, but even there the chunks load very slowly in multiplayer. This problem was present also in 1.14.X.

I'll test local multiplayer soon, although I only have three (pairs of) controllers.

It seems that local multiplayer exihbits the same problems as (online) multiplayer.

I'll test local multiplayer soon, although I only have three (pairs of) controllers.

It seems that local multiplayer exihbits the same problems as (online) multiplayer.

Edit: I feel like (on switch atleast) moving the game to the internal storage (not micro-sd card), slightly sped up chunk loading, but we have no way of testing that.

What's weird though, is on switch save data can only be on internal storage.

Edit: I feel like (on switch atleast) moving the game to the internal storage (not micro-sd card), slightly sped up chunk loading, but we have no way of testing that.

What's weird though, is on switch save data can only be on internal storage.

it makes sense that the faster the storage device the faster the loading, for example when i move my world from my (from my knowledge) hdd vps to my local ssd it speeds up quite a bit, though still not good enough to be playable sadly

it makes sense that the faster the storage device the faster the loading, for example when i move my world from my (from my knowledge) hdd vps to my local ssd it speeds up quite a bit, though still not good enough to be playable sadly

For me as a mobile player and a builder, its soo hard to do large scale builds when i get lagspikes every few seconds, this affects realms the most, for singleplayer worlds it doesnt seem to cause too much lag.

For me as a mobile player and a builder, its soo hard to do large scale builds when i get lagspikes every few seconds, this affects realms the most, for singleplayer worlds it doesnt seem to cause too much lag.

I've noticed similar issues on four different devices which each have different storage. Two Android and two PC devices. One has UFS 2.1 storage, the other has emmc, the laptop has tcl ssd, my pc has traditional rotating hdd drive. So it's not just the storage and chunk loading which causes the performance issues... but the game changes overall.

I've noticed similar issues on four different devices which each have different storage. Two Android and two PC devices. One has UFS 2.1 storage, the other has emmc, the laptop has tcl ssd, my pc has traditional rotating hdd drive. So it's not just the storage and chunk loading which causes the performance issues... but the game changes overall.

According to my testing, split screen seems to have the same effects as online multiplayer (fine in smaller worlds, bad in larger worlds), will need more testing to account for the fact that render distance is paired back in split screen multiplayer.

According to my testing, split screen seems to have the same effects as online multiplayer (fine in smaller worlds, bad in larger worlds), will need more testing to account for the fact that render distance is paired back in split screen multiplayer.

this does not happen to me. my computer has 8gb ram(7.4 usable 933mhz), r7 integrated graphics, HDD, 4 core,no hyperthread, 2.7ghz. 1gb+world, no bugs

this does not happen to me. my computer has 8gb ram(7.4 usable 933mhz), r7 integrated graphics, HDD, 4 core,no hyperthread, 2.7ghz. 1gb+world, no bugs

A test of what happens with 3 players in a 110MB survival world (14 chunk render distance, 8 chunks simulation).

With one player in the world it was fine, with two players the game started to slow but was manageable, and with three players there was extreme lag everywhere (the loading bar in the centre of the screen was me trying to open my inventory).

Edit: Yes this was the actual framerate in realtime not just the video.

A test of what happens with 3 players in a 110MB survival world (14 chunk render distance, 8 chunks simulation).

With one player in the world it was fine, with two players the game started to slow but was manageable, and with three players there was extreme lag everywhere (the loading bar in the centre of the screen was me trying to open my inventory).

Edit: Yes this was the actual framerate in realtime not just the video.

is this still an issue?

edit: on my world, it seems much, much better, chunks are loading somewhat fine, though its still a bit laggy. i'm not sure if there was an update at one point recently that fixed it or not, but it works for me now!

is this still an issue?

edit: on my world, it seems much, much better, chunks are loading somewhat fine, though its still a bit laggy. i'm not sure if there was an update at one point recently that fixed it or not, but it works for me now!

A work around I have found is to set the simulation distance to 4, which makes the problem less severe, although chunk loading can still be a bit slow.

Not tested 1.16.10 although yet.

A work around I have found is to set the simulation distance to 4, which makes the problem less severe, although chunk loading can still be a bit slow.

Not tested 1.16.10 although yet.

> set the simulation distance to 4

i've had it set to 4 since before this happened for performance reasons, haha - think you could check on 1.16.10 when you get the time? it seems it may possibly be somewhat fixed, atleast sped up alot

> set the simulation distance to 4

i've had it set to 4 since before this happened for performance reasons, haha - think you could check on 1.16.10 when you get the time? it seems it may possibly be somewhat fixed, atleast sped up alot

Yeah when setting the simulation distance to 4, chunks can actually load but you can still outrun their loading speeds (issues still occur),

This is more prevalent when the game is autosaving (along with block lag).

The amount of players in the game (at four simulation distance) doesn't seem to make a difference in terms of world speed (comparing single player to multiple players).

EDIT: I might try and make a creative copy then "/kill @e" to observer any possible changes

Yeah when setting the simulation distance to 4, chunks can actually load but you can still outrun their loading speeds (issues still occur),

This is more prevalent when the game is autosaving (along with block lag).

The amount of players in the game (at four simulation distance) doesn't seem to make a difference in terms of world speed (comparing single player to multiple players).

EDIT: I might try and make a creative copy then "/kill @e" to observer any possible changes

I found out the issue on my world! The tickspeed was set to 2000.... Setting it to 3 fixed everything.

I don't think my issues were related to this in any way, luckily, but sadly

I found out the issue on my world! The tickspeed was set to 2000.... Setting it to 3 fixed everything.

I don't think my issues were related to this in any way, luckily, but sadly

Edit: In version 1.16.20 and 1.16.21 the problem seems to have moved; players can now move around normally with chunks generating for them (on the host switch it can still be somewhat slow, but on other devices in the world chunks load fairly fast).

The problem occurs when autosaving starts, a variety of things can occur with the severity depending on your world size and simulation distance:

Block lag, chunks not loading (causing empty voids), lighting not loading, entity and mob jittering/lag.

 
Table for my large world

The duration of the autosaving also contributes to the severity of the lag, if the autosaving is short (less than 10 seconds) barely any lag happens, but if the autosaving is lengthy (1+ minutes) lots of lag occurs.

Edit: In version 1.16.20 and 1.16.21 the problem seems to have moved; players can now move around normally with chunks generating for them (on the host switch it can still be somewhat slow, but on other devices in the world chunks load fairly fast).

The problem occurs when autosaving starts, a variety of things can occur with the severity depending on your world size and simulation distance:

Block lag, chunks not loading (causing empty voids), lighting not loading, entity and mob jittering/lag.

 
Table for my large world

The duration of the autosaving also contributes to the severity of the lag, if the autosaving is short (less than 10 seconds) barely any lag happens, but if the autosaving is lengthy (1+ minutes) lots of lag occurs.

In the stone circle in the Bad_Chunks jpg my ram usage on my device jumps from 5 GB to 16+ GB. If I stay outside the circle I have no issues.

It is very easily repeatable

More details in MCPE-98243

In the stone circle in the Bad_Chunks jpg my ram usage on my device jumps from 5 GB to 16+ GB. If I stay outside the circle I have no issues.

It is very easily repeatable

More details in MCPE-98243

@unknown: I'm pretty sure yours is a different problem. I have reopened MCPE-98243 and added comments to it you might find helpful.

Unfortunately, by "repeatable" you only mean you can reproduce the failure, not the data corruption that causes the failure. The data corruption could have been caused by a bug, but other things could cause it too (power excursions, hardware errors) so there's no guarantee it's even potentially fixable, and in any case the corruption occurred too long ago for there to be any usable evidence anymore.

@unknown: I'm pretty sure yours is a different problem. I have reopened MCPE-98243 and added comments to it you might find helpful.

Unfortunately, by "repeatable" you only mean you can reproduce the failure, not the data corruption that causes the failure. The data corruption could have been caused by a bug, but other things could cause it too (power excursions, hardware errors) so there's no guarantee it's even potentially fixable, and in any case the corruption occurred too long ago for there to be any usable evidence anymore.

This chunk-generation also affects most of the Android devices like mine. (OPPO F3, Android 6) - also affects 1.16.100.58 and above versions.

This chunk-generation also affects most of the Android devices like mine. (OPPO F3, Android 6) - also affects 1.16.100.58 and above versions.

Affects 1.16.200.53 Beta

Affects 1.16.200.53 Beta

This definitely isn’t as bad as before, this generally now only seems to happens if the player pushes the game to extreme measures with extreme elytra speeds or if the world has a million dropped items otherwise its pretty normal.

This definitely isn’t as bad as before, this generally now only seems to happens if the player pushes the game to extreme measures with extreme elytra speeds or if the world has a million dropped items otherwise its pretty normal.

If I'm not mistaken, 1.16.100 introduced RenderDragon, a complete replacement for the rendering system. It's not surprising that replacing such a large and pervasive part of the game with completely new code would result in performance issues, as code generally can't be thoroughly tuned for performance until it's incorporated into the full release.

In response to feedback (including, perhaps, this report), RenderDragon was backed off in a later Beta version. I believe it has since been reinstated in one of the 1.16.200 Beta releases. So there may have been a period when it was causing performance issues, but those have been resolved now and it seems better than before. On the other hand, there's no guarantee that the newest updates to RenderDragon are as good as it can possibly get. It's hard to say whether this issue is still active, because performance problems are never "fixed" so much as "fixed well enough", and what constitutes "enough" can vary depending on the device.

If you feel like your device is not performing well enough, it would be a good idea to get specific about what type and model of device it is, what platform it runs, and platform details such as how much RAM is available and whether you're trying to run other apps simultaneously.

If I'm not mistaken, 1.16.100 introduced RenderDragon, a complete replacement for the rendering system. It's not surprising that replacing such a large and pervasive part of the game with completely new code would result in performance issues, as code generally can't be thoroughly tuned for performance until it's incorporated into the full release.

In response to feedback (including, perhaps, this report), RenderDragon was backed off in a later Beta version. I believe it has since been reinstated in one of the 1.16.200 Beta releases. So there may have been a period when it was causing performance issues, but those have been resolved now and it seems better than before. On the other hand, there's no guarantee that the newest updates to RenderDragon are as good as it can possibly get. It's hard to say whether this issue is still active, because performance problems are never "fixed" so much as "fixed well enough", and what constitutes "enough" can vary depending on the device.

If you feel like your device is not performing well enough, it would be a good idea to get specific about what type and model of device it is, what platform it runs, and platform details such as how much RAM is available and whether you're trying to run other apps simultaneously.

@auldrick I don't think ram is the issue here if you have enough. I am on WIndows 10, 32gb of ram, 1050ti, and an i7 8750H, and I am experiencing lower FPS. The ram usage is a tad higher than with the original minecraft renderer, hence why the lower end devices are struggling. But for me, a PC user with plenty of ram, I am getting around 300-400 fps in the beta as opposed to 500-600 in 1.16.40. Comparatively, there is little difference but the numbers speak for themselves. The effects of this dip can certainly be felt on console and mobile. The render dragon is introduced to improve performance, not hinder it. 

@auldrick I don't think ram is the issue here if you have enough. I am on WIndows 10, 32gb of ram, 1050ti, and an i7 8750H, and I am experiencing lower FPS. The ram usage is a tad higher than with the original minecraft renderer, hence why the lower end devices are struggling. But for me, a PC user with plenty of ram, I am getting around 300-400 fps in the beta as opposed to 500-600 in 1.16.40. Comparatively, there is little difference but the numbers speak for themselves. The effects of this dip can certainly be felt on console and mobile. The render dragon is introduced to improve performance, not hinder it. 

This chunk loading also affects my phone, using OPPO F3 with Android 6 hookup, I ended up traversing in the local world with chunk loading (in other words, slow chunk loading will lag a little bit), so this must be fixed in the next few days.

This chunk loading also affects my phone, using OPPO F3 with Android 6 hookup, I ended up traversing in the local world with chunk loading (in other words, slow chunk loading will lag a little bit), so this must be fixed in the next few days.

@unknown: I guess you mean you don't think RAM is an issue for you. My suggestion wasn't addressed to anyone in particular, and knowing how much RAM somebody has is important in gauging what kind of performance can be expected on their device.

Keep in mind that, as I've said before, it doesn't make sense for the devs to do a lot of work tuning a beta release, for several reasons. They do want reports about poor performance in beta releases, so if a major problem is introduced they can address it immediately, but you shouldn't normally expect to see a lot of performance improvement while it's still in beta. Of course, the only way they can know whether a performance problem is major is by comparing the performance you're getting to what can be reasonably expected on your device, and for that they need the specific details I was talking about.

And yes, RenderDragon is expected to eventually improve performance, but being new it probably still needs/can benefit from more tuning whereas the old engine has been tuned to exhaustion, so we won't necessarily see a huge improvement with RD in its first live release.

Edit for clarification: When I say "it doesn't make sense to tune while still in beta releases", I don't mean they wait until after the regular release before they start concentrating on improving performance. Rather, they do that as they get closer to the end of the beta phase, because the feature work is mostly done by then and won't interfere with performance measurement.

@unknown: I guess you mean you don't think RAM is an issue for you. My suggestion wasn't addressed to anyone in particular, and knowing how much RAM somebody has is important in gauging what kind of performance can be expected on their device.

Keep in mind that, as I've said before, it doesn't make sense for the devs to do a lot of work tuning a beta release, for several reasons. They do want reports about poor performance in beta releases, so if a major problem is introduced they can address it immediately, but you shouldn't normally expect to see a lot of performance improvement while it's still in beta. Of course, the only way they can know whether a performance problem is major is by comparing the performance you're getting to what can be reasonably expected on your device, and for that they need the specific details I was talking about.

And yes, RenderDragon is expected to eventually improve performance, but being new it probably still needs/can benefit from more tuning whereas the old engine has been tuned to exhaustion, so we won't necessarily see a huge improvement with RD in its first live release.

Edit for clarification: When I say "it doesn't make sense to tune while still in beta releases", I don't mean they wait until after the regular release before they start concentrating on improving performance. Rather, they do that as they get closer to the end of the beta phase, because the feature work is mostly done by then and won't interfere with performance measurement.

Affects 1.16.200.55 Beta

Affects 1.16.200.55 Beta

One thing to add is that the game lags out significantly when flying around with an elytra equipped (at least in 1.16.200.55). When the player equips an elytra, the game lags substantially and continues to lag substantially until the player unequips the elytra. I have attached a video showing this in action since I thought this issue related to this bug report.

One thing to add is that the game lags out significantly when flying around with an elytra equipped (at least in 1.16.200.55). When the player equips an elytra, the game lags substantially and continues to lag substantially until the player unequips the elytra. I have attached a video showing this in action since I thought this issue related to this bug report.

@Connor McNally: the issue with FPS dropping when wearing elytra is being tracked separately at MCPE-105632. I will move your video over to that report. You may wish to add a vote.

@Connor McNally: the issue with FPS dropping when wearing elytra is being tracked separately at MCPE-105632. I will move your video over to that report. You may wish to add a vote.

In 1.16.200.55 Beta, i'm getting alot of lag. I did disable fancy graphics, fancy bubbles, fancy leafs, lower down my sim distance, lowered my render distance, disabled beautiful skies, to reduce lag, but me doing all of that improved a little bit only. I know it is a Beta and RenderDragon is not fully optimized until the full version comes out, but still the lag is noticeable. 

In 1.16.200.55 Beta, i'm getting alot of lag. I did disable fancy graphics, fancy bubbles, fancy leafs, lower down my sim distance, lowered my render distance, disabled beautiful skies, to reduce lag, but me doing all of that improved a little bit only. I know it is a Beta and RenderDragon is not fully optimized until the full version comes out, but still the lag is noticeable. 

Mostly fixed in 1.16.100, but is a bit more difficult to repro so affects 1.16.100.

Mostly fixed in 1.16.100, but is a bit more difficult to repro so affects 1.16.100.

I'd like to add that the slow chunk loading creates intense block lag when insta-mining. Only noticed in 1.16.100 full release Nintendo Switch edition (all lowest settings possible). The lag wasn't there yesterday, and makes the game unplayable. Not sure why y'all released a version like this

I'd like to add that the slow chunk loading creates intense block lag when insta-mining. Only noticed in 1.16.100 full release Nintendo Switch edition (all lowest settings possible). The lag wasn't there yesterday, and makes the game unplayable. Not sure why y'all released a version like this

Noticeable on PS4 since updating to 1.16.100. 

Noticeable on PS4 since updating to 1.16.100. 

Windows 10, v1.16.100: Chunks are loading very very slowly for me, especially when I use elytra. Also occurs somewhat frequently when just riding in a boat as well. Not sure if it's the game or my computer but I have a feeling it's the game since chunk loading was, although not great, better before 1.16.100

After 1.16.100 update, I can't explore with elytra properly. Before the update all the chunks loaded fast enough to fly. Now, I hit invisible walls every 200-300 meters. It is impossible to play now!

After 1.16.100 update, I can't explore with elytra properly. Before the update all the chunks loaded fast enough to fly. Now, I hit invisible walls every 200-300 meters. It is impossible to play now!

Me personally I notice a major improvement on iOS, here is some documentation:
iPhone SE 2nd gen:
Before 1.16.100 - 24 chunk render distance: 15-30 FPS
After 1.16.100 - 24 chunk render distance: 30-60 FPS
iPhone 7:
Before 1.16.100 - 24 chunk render distance: CRASH
After 1.16.100 - 24 chunk render distance: 20-40 FPS
I think as far as I can tell about performance being improved is true, though I think due to MCPE-105892 that causes it to seem worse due to entities and related stuff lagging out. Now I'm not saying this issue is fixed, I personally can still reproduce but chunk loading is definitely better, and looking into coding it's been refined and shortened to help performance. Now do keep in mind I tested this with an iPhone SE2, iPhone 7, Xbox one and a lower end Win 10 PC for these results. But the main cuase of this issue seems to be the game being unable to allocate the devices RAM properly and CPU usage, I'm looking into the coding and hope to find more. Thanks for reading!

Me personally I notice a major improvement on iOS, here is some documentation:
iPhone SE 2nd gen:
Before 1.16.100 - 24 chunk render distance: 15-30 FPS
After 1.16.100 - 24 chunk render distance: 30-60 FPS
iPhone 7:
Before 1.16.100 - 24 chunk render distance: CRASH
After 1.16.100 - 24 chunk render distance: 20-40 FPS
I think as far as I can tell about performance being improved is true, though I think due to MCPE-105892 that causes it to seem worse due to entities and related stuff lagging out. Now I'm not saying this issue is fixed, I personally can still reproduce but chunk loading is definitely better, and looking into coding it's been refined and shortened to help performance. Now do keep in mind I tested this with an iPhone SE2, iPhone 7, Xbox one and a lower end Win 10 PC for these results. But the main cuase of this issue seems to be the game being unable to allocate the devices RAM properly and CPU usage, I'm looking into the coding and hope to find more. Thanks for reading!

Can confirm that the game is absolutely unplayable in any capacity in a multiplayer setting, issue is still occurring even in single player. I can confirm this is happening via windows 10, Android Devices (LG Stylo 6), iPhone XR, and Nintendo Switch. Entities move like a slideshow, when you hit them, the slowly fly back into the air after a few seconds. Loading chunks is next to impossible. Often, invisible blocks will appear or the void will consume the entire chunk in question. 

Can confirm that the game is absolutely unplayable in any capacity in a multiplayer setting, issue is still occurring even in single player. I can confirm this is happening via windows 10, Android Devices (LG Stylo 6), iPhone XR, and Nintendo Switch. Entities move like a slideshow, when you hit them, the slowly fly back into the air after a few seconds. Loading chunks is next to impossible. Often, invisible blocks will appear or the void will consume the entire chunk in question. 

it is very hard to fly with elytra and fireworks because you are moving so fast and the chunks aren't loading, so you freeze. In Java, flying with elytra and fireworks is much easier.

I'm still able to experience chunks loading slowly when flying with an elytra. I lowered down all my settings to improve performance and also turned off wifi as well but, only a bit performance increase, which could be noticeable but also the same time cannot. I know it's on beta, and beta's are meant for testing. Hoping this will completed be fixed in a full release. You guys are doing great job fixing bugs. 

I'm still able to experience chunks loading slowly when flying with an elytra. I lowered down all my settings to improve performance and also turned off wifi as well but, only a bit performance increase, which could be noticeable but also the same time cannot. I know it's on beta, and beta's are meant for testing. Hoping this will completed be fixed in a full release. You guys are doing great job fixing bugs. 

Chunk rendering appears to have degraded in Realms, namely chunks load but don't visually render until a few seconds have passed (other than tile/block entities). This also applies to subchunks, especially in the Nether, leading to increased risk of death due to the elytra deactivation bug when hitting a wall.
I just noted the new fix version after typing this so I will wait for it.

Chunk rendering appears to have degraded in Realms, namely chunks load but don't visually render until a few seconds have passed (other than tile/block entities). This also applies to subchunks, especially in the Nether, leading to increased risk of death due to the elytra deactivation bug when hitting a wall.
I just noted the new fix version after typing this so I will wait for it.

@unknown: In future, when updating a description, please try to make all your corrections before submitting once, rather than making one tiny correction at a time. Also, it's not necessary to fix every tiny grammatical and spelling mistake, only the ones that will likely cause a misunderstanding.

Our problem is that each time you save your changes, a complete copy of the loooong before and after is appended to the history. There are dozens of copies in there now, and each fills a whole page on my screen, which makes it difficult to scroll past all your tiny, mostly unimportant changes when I'm looking to see who updated one of the other fields.

@unknown: In future, when updating a description, please try to make all your corrections before submitting once, rather than making one tiny correction at a time. Also, it's not necessary to fix every tiny grammatical and spelling mistake, only the ones that will likely cause a misunderstanding.

Our problem is that each time you save your changes, a complete copy of the loooong before and after is appended to the history. There are dozens of copies in there now, and each fills a whole page on my screen, which makes it difficult to scroll past all your tiny, mostly unimportant changes when I'm looking to see who updated one of the other fields.

This issue, together with MCPE-105892 for 1.16.100 and later, has been fixed in the 1.16.200.57 Beta release. Provided no further problem is found with the fix, it will be included in the next full release. Until then, we will leave this ticket open.

No date for the full release can be given, but they're looking at releasing it in weeks, not months, if all goes as expected.

This issue, together with MCPE-105892 for 1.16.100 and later, has been fixed in the 1.16.200.57 Beta release. Provided no further problem is found with the fix, it will be included in the next full release. Until then, we will leave this ticket open.

No date for the full release can be given, but they're looking at releasing it in weeks, not months, if all goes as expected.

same issue but my world have 312MB

same issue but my world have 312MB

In beta 1.16.210.50, chunk loading is OK/better for me. I have not yet experienced too much lag like 1.16.200 beta's. Flying with elytra in superfast, only entities appear in loading. Chunks won't visually appear. It does after 2 to 7 seconds.

In beta 1.16.210.50, chunk loading is OK/better for me. I have not yet experienced too much lag like 1.16.200 beta's. Flying with elytra in superfast, only entities appear in loading. Chunks won't visually appear. It does after 2 to 7 seconds.

changelog says this has been improved in 1.16.200 when flying w/ elytra

I'd say it has been improved a bit from 1.16.100 but it's still too slow in my opinion as reaching fast speeds with an elytra still results in flying into unloaded chunks a lot of the time, and since the unloaded chunks load as soon as you enter them, there is a high possibility of hitting blocks that you wouldn't have been able to see as they were unloaded before

Chunk loading is still not great. I'm still hitting void walls flying around, even with the vastly reduced max render distance.

Chunk loading is still not great. I'm still hitting void walls flying around, even with the vastly reduced max render distance.

Its still horrible nothing changed since 1.16.100 chunks dont load, you cant fly with an elytra properly, Block lagging its just sad im afraid of 1.17 when the chunks are 64 blocks deeper, more Blocks loading means more lag and it will take even longer for chunks to load

,

Its still horrible nothing changed since 1.16.100 chunks dont load, you cant fly with an elytra properly, Block lagging its just sad im afraid of 1.17 when the chunks are 64 blocks deeper, more Blocks loading means more lag and it will take even longer for chunks to load

,

I still experience this bug in multiplayer, but rarely singleplayer.

I still experience this bug in multiplayer, but rarely singleplayer.

I still experience this issue on 1.17.40.06. Ever since 1.16.100 this issue still hasn't been fixed for me.

I still experience this issue on 1.17.40.06. Ever since 1.16.100 this issue still hasn't been fixed for me.

We are currently tracking delayed chunk loading at MCPE-120971.

We are currently tracking delayed chunk loading at MCPE-120971.

I still notice lag chunks loading, also under the suggested render distance (12 and i set to 10), at around 6-8 looks pretty ok, but is the half SUGGESTED render distance. I wanna to specify that I'm using an iPad with a M2 chip... pretty fast, no? Still lag... with or without resource packs. I'm on the 1.21.2

I still notice lag chunks loading, also under the suggested render distance (12 and i set to 10), at around 6-8 looks pretty ok, but is the half SUGGESTED render distance. I wanna to specify that I'm using an iPad with a M2 chip... pretty fast, no? Still lag... with or without resource packs. I'm on the 1.21.2