I'm skeptical but haven't found a blatant smoking gun yet; I don't know how task manager behaves with that many logical processors. However, I'm pretty sure the claimed hardware config is impossible, so it's at the very least a VM, and probably sped up.
-
-
-
This claims Windows Server 2019 does up to 1024 logical processors (threads), which wouldn't be enough. Though it's just an HP spec. I can't find a limit on MS's site. https://assets.ext.hpe.com/is/content/hpedam/documents/a00056000-6999/a00056655/a00056655enw.pdf …
1 reply 0 retweets 11 likes -
One Windows Kernel: „pre-release Windows DataCenter class machine with 896 cores supporting 1792 logical processors and 2TB of RAM!“https://techcommunity.microsoft.com/t5/windows-kernel-internals/one-windows-kernel/ba-p/267142 …
1 reply 0 retweets 7 likes -
Replying to @aschilling @winocm
So those are the exact same specs... same with the cache (and the L3 cache number is definitely wrong). The tiny size of the CPU usage in the cells, and the implication that this is a HiDPI screen bothers me a lot. Also, the weird gaps around cols 3-5.
1 reply 0 retweets 12 likes -
The gaps imply some kind of alignment issue with columns that are not an exact divider of the window width, but the screenshots I'm finding of task manager don't do that. Could be a poor attempt at editing it. But really, HiDPI on a server? Maybe it's RDP? Still, kinda fishy.
1 reply 0 retweets 14 likes -
I think the real task manager is supposed to have a blue border around the CPU display too, and the CPU section on the left bar is supposed to be selected. So I dunno, unless this is a significantly different version of Windows / task manager, this doesn't really look real.
3 replies 0 retweets 13 likes -
Yeah, the fonts look wrong, "Base Speed" is capitalized "Base speed" in the real one, and there are other minor issues that are unexplainable. I think someone recreated the entire Task Manager UI to fake this (instead of using a screenshot as a base).
1 reply 5 retweets 48 likes -
I'm calling it fake.
2 replies 1 retweet 44 likes -
I'm thinking also it would be very difficult to prevent "screen" tearing from occurring in this situation. you'd have to be synchronized with the measurement interval very closely. I also kind of doubt the averaging window is small enough to allow for this
1 reply 0 retweets 4 likes
Yeah, @winocm's version is a lot messier due to sampling issues like this. I just focused on UI problems since those are more obvious smoking guns. I've been wanting to try improving the code that she used to see how good I can get it to look.
-
-
A few million pixels Retweeted sarah@ganbaranai
damn. that is still really good thoughhttps://twitter.com/winocm/status/1276037359503466497 …
A few million pixels added,
2:20sarah@ganbaranai @winocmWell, I got it running for the most part. Here’s Bad Apple on taskmgr on a very large computer. SMT had to be disabled to ensure that the image displays properly as the kernel scheduler handler those logical threads differently than physical cores. pic.twitter.com/QMchANZy9WShow this thread0 replies 0 retweets 3 likesThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.