|
shrike82 posted:RE the AC:Origins benchmarks, the 4C/8T 7700K being 50% faster than the 4C/4C 7600K is pretty crazy. Word is that AC: Origins is using multiple types of CPU-heavy DRM simultaneously, it may not be representative of modern game engines in general: https://torrentfreak.com/assassins-creed-origin-drm-hammers-gamers-cpus-171030/ You need an 8 thread CPU so that 4 threads can run the DRM while the other 4 run the game. I know this is just rumormongering at this point, but performance has improved in other games when Denuvo was removed so it's not that outlandish.
|
#
?
Nov 1, 2017 15:10
|
|
|
|
| # ? May 16, 2024 13:31 |
|
|
Twerk from Home posted:Word is that AC: Origins is using multiple types of CPU-heavy DRM simultaneously, it may not be representative of modern game engines in general: I find that very hard to believe, but if so it's ridiculous. Using half the computer's power just for DRM?
|
|
#
?
Nov 1, 2017 15:22
|
|
|
128 lanes of pcie gen4? Holy balls. 
|
|
#
?
Nov 1, 2017 15:46
|
|
|
feedmegin posted:I find that very hard to believe, but if so it's ridiculous. Using half the computer's power just for DRM? It's a Ubisoft game, so it's pretty believable
|
|
#
?
Nov 1, 2017 15:47
|
|
|
feedmegin posted:I find that very hard to believe, but if so it's ridiculous. Using half the computer's power just for DRM? The rumor is slowly working its way to slightly more reputable blogs: https://www.techpowerup.com/238377/cpus-bare-brunt-of-ubisoft-deploying-vmprotect-above-denuvo-for-ac-o
|
|
#
?
Nov 1, 2017 15:49
|
|
|
We should be able to see "clean" benchmarks once someone cracks it.
|
|
#
?
Nov 1, 2017 15:52
|
|
|
Plus it shouldn't be that hard to just profile it and see where it's spending cycles. But let's rumormonger instead.
|
|
#
?
Nov 1, 2017 15:52
|
|
|
shrike82 posted:We should be able to see "clean" benchmarks once someone cracks it. Probably not, these VM-based DRM systems mangle the executable so much that it's practically impossible to reverse them back into the original binary. They usually get cracked by leaving the DRM in place but spoofing the environment around it so the licence checks succeed.
|
|
#
?
Nov 1, 2017 16:00
|
|
|
lol neither the crack scene nor Ubisoft are trustworthy but the latter just released a statement about thisquote:Ubisoft: AC Origins’ DRM On PC Has No “Perceptible Effect” On Game Performance
|
|
#
?
Nov 1, 2017 16:10
|
|
|
Oh, because THAT'S supposed to be so much more reassuring. "Noooo, it's not the DRM that's doing it! Our game just runs like absolute poo poo on its own!"
|
|
#
?
Nov 1, 2017 16:47
|
|
|
shrike82 posted:RE the AC:Origins benchmarks, the 4C/8T 7700K being 50% faster than the 4C/4C 7600K is pretty crazy. The cracker community is saying that most of that overhead goes to calling their DRM library hundreds of times per second, which kinda implies that once they crack that poo poo the cracked version should run a ton better on more modest hardware. Hasn't stopped the AMD fanbois from proclaiming this to be the future of gaming titles  edit: apparently I am a page behind.
|
|
#
?
Nov 1, 2017 17:10
|
|
|
Lmao that the reason we need more multicore power in the future is because we need to have loving 120 threads for some obtuse as gently caress DRM while 8 run the software.
|
|
#
?
Nov 1, 2017 17:41
|
|
|
i guess that's one way to make games scale with cores...
|
|
#
?
Nov 1, 2017 17:41
|
|
|
FaustianQ posted:https://twitter.com/CPCHardware/status/925348852147908608 Paul MaudDib posted:The cracker community is saying that most of that overhead goes to calling their DRM library hundreds of times per second, which kinda implies that once they crack that poo poo the cracked version should run a ton better on more modest hardware. Combat Pretzel fucked around with this message at 17:53 on Nov 1, 2017 |
|
#
?
Nov 1, 2017 17:51
|
|
|
Combat Pretzel posted:Or maybe just double the CCX on the package? Then again, that would get awfully cramped. Nah, for 64C on current Summit processors you'd only have 128MB of L3 Cache, and to have 256MB you'd obviously have 128 cores or 16 Summit processors. The other hint is that it's still only Octochannel memory. Still wondering at the doubled L3 though, like aren't there diminishing returns on larger L3s and wouldn't it be better to hit a more optimum size and go for better latency characteristics?
|
|
#
?
Nov 1, 2017 18:07
|
|
|
AC:O's "anti-tamper" solution is to literally run the DRM in a non-hardware assisted, software-based x86 virtualization wrapper. Even if that's using some kind of magical dynamic translation engine that's still going to eat cores like a hot drat.
|
|
#
?
Nov 1, 2017 18:09
|
|
|
Kazinsal posted:AC:O's "anti-tamper" solution is to literally run the DRM in a non-hardware assisted, software-based x86 virtualization wrapper. Even if that's using some kind of magical dynamic translation engine that's still going to eat cores like a hot drat. That's how most (somewhat) effective DRM systems work, although the VM is often a custom design rather than based on x86. The trick is to only VM-ify parts of the code that aren't in the hot path so performance doesn't go to poo poo. Doom 2016, MGS5, Mad Max, etc all used VM protection and ran fine because the developers weren't dumbasses and only wrapped performance insensitive code.
|
|
#
?
Nov 1, 2017 18:16
|
|
|
Doing full software virtualization and throwing the hot path and friends in is absolutely mindblowingly retarded though. It's not quite "they just threw Denuvo in a QEMU instance" but it's drat close.
|
|
#
?
Nov 1, 2017 18:20
|
|
|
FaustianQ posted:Nah, for 64C on current Summit processors you'd only have 128MB of L3 Cache, and to have 256MB you'd obviously have 128 cores or 16 Summit processors. The other hint is that it's still only Octochannel memory. The rule of thumb is that doubling the cache size reduces your miss rate by sqrt(2), or ~1.41x. So yeah at some point it's good enough and you will start crowding out other stuff or increasing latency. It sounds to me like they are going to 8-core CCXs.
|
|
#
?
Nov 1, 2017 18:45
|
|
|
repiv posted:That's how most (somewhat) effective DRM systems work, although the VM is often a custom design rather than based on x86. The trick is to only VM-ify parts of the code that aren't in the hot path so performance doesn't go to poo poo. According to people on Twitter stepping through the assembly, it goes to the VM every time the character moves, so NBD I'm sure. E: the VMs plural because I guess there are two layers of it and one(?) is 'fractional' so it spawns a bunch of them... somehow? IDK how something like that would work at all unless what they're seeing is two VM layers per thread Munkeymon fucked around with this message at 18:58 on Nov 1, 2017 |
|
#
?
Nov 1, 2017 18:52
|
|
|
Munkeymon posted:According to people on Twitter stepping through the assembly, it goes to the VM every time the character moves, so NBD I'm sure. Ubisoft never fails to disappoint 
|
|
#
?
Nov 1, 2017 18:55
|
|
|
Paul MaudDib posted:The rule of thumb is that doubling the cache size reduces your miss rate by sqrt(2), or ~1.41x. So yeah at some point it's good enough and you will start crowding out other stuff or increasing latency. Well yea, but even then they're still doubling the cache per core, so 4MB instead of 2MB per core. It's the one part of the"leak" that I think is bullshit, everything else is moderately believable, even a theoretical 64 PCIE lanes per die. Actually, gently caress are there even motherboards that'd be capable of utilizing a full theoretical 128 PCIE lanes?
|
|
#
?
Nov 1, 2017 19:01
|
|
|
FaustianQ posted:Well yea, but even then they're still doubling the cache per core, so 4MB instead of 2MB per core. It's the one part of the"leak" that I think is bullshit, everything else is moderately believable, even a theoretical 64 PCIE lanes per die. You can do up to 7 slots in an EATX/SSI-CEB form factor, PCPP says there was one 8-slot board on LGA1155 that used an ATX-XL form factor.
|
|
#
?
Nov 1, 2017 19:10
|
|
|
repiv posted:That's how most (somewhat) effective DRM systems work, although the VM is often a custom design rather than based on x86. The trick is to only VM-ify parts of the code that aren't in the hot path so performance doesn't go to poo poo.
|
|
#
?
Nov 1, 2017 19:12
|
|
|
If all else fails you could route them to a shitload of U.2 ports (4 lanes each).
|
|
#
?
Nov 1, 2017 19:12
|
|
|
Combat Pretzel posted:So the point of the VM is to have the DRM implemented in an unknown instruction set, so that it can't be easily reverse engineered? http://vmpsoft.com/ Pretty much: VMProtect posted:VMProtect protects code by executing it on a virtual machine with non-standard architecture It can do this over and over: VMProtect posted:Each time you protect the application, VMProtect generates a completely different set of virtual machines, so even if a cracker finally understand an architecture of the particular virtual machine, he has to start from the very beginning for the second protected procedure of the same file.
|
|
#
?
Nov 1, 2017 21:06
|
|
|
repiv posted:If all else fails you could route them to a shitload of U.2 ports (4 lanes each). This is probably the biggest use case. Light up a storage server with 32 full-speed U.2 ports for fast enterprise SSDs with a single CPU.
|
|
#
?
Nov 1, 2017 21:13
|
|
|
HalloKitty posted:http://vmpsoft.com/ "You asked me to design a next-generation piracy countermeasure. It's not my fault it requires next-generation hardware to work!" - an engineer
|
|
#
?
Nov 1, 2017 21:13
|
|
|
Combat Pretzel posted:Gotta admire their tenacity about this DRM thing. The more complex versions don't delay pirated releases significantly. They might as well just drop in some token copy protection scheme to make casual copying impractical, because that's what the complex ones still only do, and leave it at that. Not to defend DRM that just wastes cycles but AC:O is apparently uncracked a week on while Wolfenstein literally got cracked two days before the game actually launched with their CTO whining about piracy. Watch Dogs 2 was another Ubisoft release which managed to delay a crack release buy a couple months. There was an interesting article on arstechnica a while back that delaying a cracked release by a couple months results in a material impact on sales.
|
|
#
?
Nov 2, 2017 00:22
|
|
|
shrike82 posted:Not to defend DRM that just wastes cycles but AC:O is apparently uncracked a week on while Wolfenstein literally got cracked two days before the game actually launched with their CTO whining about piracy. Watch Dogs 2 was another Ubisoft release which managed to delay a crack release buy a couple months. yeah what with how heavily AAA game sales are weighted around time of release it doesn't need to be permanently uncrackable, just uncrackable while there's launch buzz. it's just a shame that people who actually bought the game get treated like criminals. i mean ubisoft could try making more interesting games with a more sustainable business model and... who am i kidding; 42 nested vms and biometric authentication it is Generic Monk fucked around with this message at 00:38 on Nov 2, 2017 |
|
#
?
Nov 2, 2017 00:32
|
|
|
Watchdogs 2 took months to crack? Wouldn't have thought that.
|
|
#
?
Nov 2, 2017 00:34
|
|
|
Game released Nov 15, 2016. First crack was Jan 18 apparently.
|
|
#
?
Nov 2, 2017 00:37
|
|
|
Combat Pretzel posted:Watchdogs 2 took months to crack? Wouldn't have thought that. yeah as drm goes denuvo is pretty loving effective at the purpose it was designed for. obviously as a concept the whole thing remains a stupid loving idea, but it is slightly less openly contemptuous of your customers than lying to them about why the singleplayer game they paid money for needs a constant internet connection a la simcity
|
|
#
?
Nov 2, 2017 00:46
|
|
|
Too bad denuvo gets cracked in a few hours now
|
|
#
?
Nov 2, 2017 00:51
|
|
|
PerrineClostermann posted:Too bad denuvo gets cracked in a few hours now There’s only a few people who know how to break it, depending on the engine. This year’s Life Is Strange sequel was easier than many games because it uses Unity for some reason I’m not hardcore enough to understand. Also, a few high profile games that remove it after launch month get so much attention that it ignores the big list of games that have been broken but haven’t removed it anyway. There was a big old list of games that weren’t cracked until someone came on the scene with a new identity, cracked a lot of them, and then disappeared back into the unknown.
|
|
#
?
Nov 2, 2017 01:12
|
|
|
PerrineClostermann posted:Too bad denuvo gets cracked in a few hours now obviously just needs more vms fam
|
|
#
?
Nov 2, 2017 02:24
|
|
|
It'd be hilarious if AMD does a "Powered by AMD" technical partnership with game devs to drive multi process utilization by spamming DRM VMs.
|
|
#
?
Nov 2, 2017 02:51
|
|
|
Wouldn't be that different than Nvidia and gameworks tesselation
|
|
#
?
Nov 2, 2017 14:08
|
|
|
Crashing Ryzen update: I took the computer apart and dusted off all the components - which were clean anyway, since I thoroughly dusted them off a few months ago when I first built the system, but figured that it can't do any harm. Then put everything back together, zerofilled the drives and reinstalled everything from the ground up.... and the system is now bulletproof. I don't get it. I know we've come a long way since the DLL hell days of Win9x, but sometimes the guy's gotta wonder, as I have no rational explanation for this. Hooray I guess? 
|
|
#
?
Nov 3, 2017 06:52
|
|
GAY 4 ORGANS
|
|
| # ? May 16, 2024 13:31 |
|
|
It's just how computers work these days. Doing the exact same thing over and over is not guaranteed to produce same or even similar results. 
|
|
#
?
Nov 3, 2017 07:01
|
|
