Patch will slow down Intel CPUs up to 30%
- KVRAF
- 24455 posts since 7 Jan, 2009 from Croatia
Fix will be non-optional, and actually partial fix is already in 10.13.2 and more work on it will be in 10.13.3. W10 also got their fix this Wed.
Performance hit is negligible for most things - on W10 what got affected the most are SSD 4K read times. Seems there's no impact to ASIO latencies etc.
Performance hit is negligible for most things - on W10 what got affected the most are SSD 4K read times. Seems there's no impact to ASIO latencies etc.
-
- Banned
- 2238 posts since 19 Dec, 2014
EvilDragon wrote:Performance hit is negligible for most things - on W10 what got affected the most are SSD 4K read times. Seems there's no impact to ASIO latencies etc.
is that speculative, or are you seeing actual benchmarks (concerning a reasonably encompassing DAW range of activities) ?
the only two anecdotal performance reports I've seen thus far, from DAW people, have not been good.
-
- Banned
- 2238 posts since 19 Dec, 2014
ya i agree. my DAW/Design machine exists only for that purpose anyway. keeping it off the net, other than to update, would be easy. even keeping it off the net entirely, if needs be.aMUSEd wrote:They should make the fix optional - I am not crippling my Mac just for something that might remotely happen someday - this is just risk averse scaremongering. I'd rather just keep my Mac off the net.
- KVRAF
- 37510 posts since 14 Sep, 2002 from In teh net
On Mac it's optional to the extent that I don't have to install updates anymore. They are not forced like on Windows, but it does mean I can't also install updates that might be useful, I can't ignore just this as they are all bundled together and that is wrong.EvilDragon wrote:Fix will be non-optional, and actually partial fix is already in 10.13.2 and more work on it will be in 10.13.3. W10 also got their fix this Wed.
- KVRAF
- 24455 posts since 7 Jan, 2009 from Croatia
Daags wrote:EvilDragon wrote:Performance hit is negligible for most things - on W10 what got affected the most are SSD 4K read times. Seems there's no impact to ASIO latencies etc.
is that speculative, or are you seeing actual benchmarks (concerning a reasonably encompassing DAW range of activities) ?
the only two anecdotal performance reports I've seen thus far, from DAW people, have not been good.
People on GS reporting their PT rigs working just fine and as they did before the fix (and you know how PT can be fussy). People on VI-Control (so, orchestral composers mostly reporting no changes, or within margin of error 1-2%). There are now also quite some post-fix benches around, just google.
Anyone with a Haswell CPU onwards should really have marginal impact on performance (because of the PCID instruction that was introduced then, which helps with reducing the performance hit).
-
- Banned
- 2238 posts since 19 Dec, 2014
interesting, thanks, I will follow up on those.EvilDragon wrote:Daags wrote:EvilDragon wrote:Performance hit is negligible for most things - on W10 what got affected the most are SSD 4K read times. Seems there's no impact to ASIO latencies etc.
is that speculative, or are you seeing actual benchmarks (concerning a reasonably encompassing DAW range of activities) ?
the only two anecdotal performance reports I've seen thus far, from DAW people, have not been good.
People on GS reporting their PT rigs working just fine and as they did before the fix (and you know how PT can be fussy). People on VI-Control (so, orchestral composers mostly reporting no changes, or within margin of error 1-2%). There are now also quite some post-fix benches around, just google.
Anyone with a Haswell CPU onwards should really have marginal impact on performance (because of the PCID instruction that was introduced then, which helps with reducing the performance hit).
-
fluffy_little_something fluffy_little_something https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=281847
- Banned
- 12880 posts since 5 Jun, 2012
In W7 there was a tool for measuring the performance potential of one^s system configuration, it would yield some value, like 7.2.
Can someone tell me where that feature is in W10? I can't find it. I would like to measure my configuration now and then again after the patch has been applied (so far that has not happened, it seems, judging from the Windows Update list).
Can someone tell me where that feature is in W10? I can't find it. I would like to measure my configuration now and then again after the patch has been applied (so far that has not happened, it seems, judging from the Windows Update list).
- KVRAF
- 24455 posts since 7 Jan, 2009 from Croatia
https://winaero.com/blog/windows-experi ... indows-10/
This performance index won't show you much after the update. You need to do more adequate benchmarks (like CrystalMark for testing SSDs, or DAWbench for testing if anything impacted your DAW).
This performance index won't show you much after the update. You need to do more adequate benchmarks (like CrystalMark for testing SSDs, or DAWbench for testing if anything impacted your DAW).
-
- Banned
- 2238 posts since 19 Dec, 2014
I think you're talking about the Windows Experience Index (WEI), which was in the system section of the control panel on W7. It was removed since Windows 8 ...fluffy_little_something wrote:In W7 there was a tool for measuring the performance potential of one^s system configuration, it would yield some value, like 7.2.
Can someone tell me where that feature is in W10? I can't find it. I would like to measure my configuration now and then again after the patch has been applied (so far that has not happened, it seems, judging from the Windows Update list).
quick google for a W10 equivalent surfaced this:
https://answers.microsoft.com/en-us/win ... 378?auth=1
honestly, I'd put much more trust in a stress test (specific to whatever process you are interested in, i.e audio or 3d) over some windows diagnostic.
http://www.dawbench.com/benchmarks.htm
-
- Banned
- 2238 posts since 19 Dec, 2014
snap!EvilDragon wrote:https://winaero.com/blog/windows-experi ... indows-10/
This performance index won't show you much after the update. You need to do more adequate benchmarks (like CrystalMark for testing SSDs, or DAWbench for testing if anything impacted your DAW).
-
fluffy_little_something fluffy_little_something https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=281847
- Banned
- 12880 posts since 5 Jun, 2012
Yes, I just read that the Windows index is based on graphics performance to a large extent, which, however, is deliberately almost non-existing on my computer 
- KVRAF
- 24455 posts since 7 Jan, 2009 from Croatia
So after some encouraging results from people at VI-C and GS, I pulled the patch. I did a DAWbench VI in Reaper before the patch and after, and I'm getting much the same numbers: 820 voices at 128 samples buffer with my RME UFX+, and with the factory library on Samsung 850 EVO. i7-6700K at 4.5 GHz here.
So, it seems that ASIO performance wasn't affected one bit (at least as far as Reaper is concerned). Looks like DAW users fall into "average workload" crowd.
So, it seems that ASIO performance wasn't affected one bit (at least as far as Reaper is concerned). Looks like DAW users fall into "average workload" crowd.
-
- KVRAF
- 35689 posts since 11 Apr, 2010 from Germany
As i mentioned somewhere else, i'm not surprised about that at all... nothing but hot air. They needed a big thing for the new year.EvilDragon wrote:So after some encouraging results from people at VI-C and GS, I pulled the patch. I did a DAWbench VI in Reaper before the patch and after, and I'm getting much the same numbers: 820 voices at 128 samples buffer with my RME UFX+, and with the factory library on Samsung 850 EVO. i7-6700K at 4.5 GHz here.
So, it seems that ASIO performance wasn't affected one bit (at least as far as Reaper is concerned).
- KVRAF
- 24455 posts since 7 Jan, 2009 from Croatia
I'm sure some OTHER workloads WILL suffer a much worse fate, targetted benchmarks confirm that.
-
- KVRian
- 890 posts since 9 May, 2005
Anyone else having flashbacks to Y2k? 