I don’t quite believe it just yet, but I will find out first hand next week, when we patch our sql boxes and other servers… 
if it’s really the fault of meltdown patches… seems they were running those “services” awfully close to capacity to start with
What a bad time to be Intel… I guess it is time to start investing in AMD.
Yeah, that seems very likely that they could’ve been just under full load without the meltdown security hole and the patch may have still brought overall CPU performance slightly down.
Sounds like rockstar and the GTA servers to me XD
I saw the original post earlier, didn’t read through it yet. But some thoughts when I saw that graphic…
I kinda doubt this is just the patch. I mean think about it.
They are probably not patching all the servers at the same time (duh.), so quite some server will be offline at a time (considering it’s cloud servers they’re using, not their own server farm). Servers/nodes offline = less total resources = less optimal load balancing across nodes = more load per node.
I mean yeah there will probably quite some performance impact on their side, but I doubt it’s that strong…
What was the hardware spec of the servers in the first place?
Does anyone know?
Doubling utilisation on a potato isnt as impressive (well, impressive is wrong word… should be horrifying) as doubling utilisation on a beast cpu.