lynux3 2,052 Posted April 20, 2019 Share Posted April 20, 2019 (edited) 12 hours ago, Remij_ said: I think I understand exactly. 3dmarks physics test scales with cores dumbass. It will almost certainly lower his physics score... @The Mother Fucker Now enable all the cores, and run using only game mode and post the results. I want to shut this dumb bitch up Like I said, you have no idea how it works, dumbass. Glad The Mother Fucker was able to shut your dumb bitch up. Edited April 20, 2019 by lynux3 Quote Link to post Share on other sites
Remij 4,670 Posted April 20, 2019 Share Posted April 20, 2019 4 minutes ago, lynux3 said: Like I said, you have no idea how it works, dumbass. Uh... he proved my point fucktard. The fuckhead who tried to argue that overclocking improved IPC thinks he knows shit about CPUS. ROFL God damn you're dumb 1 Quote Link to post Share on other sites
lynux3 2,052 Posted April 20, 2019 Share Posted April 20, 2019 6 minutes ago, Remij_ said: Uh... he proved my point fucktard. The fuckhead who tried to argue that overclocking improved IPC thinks he knows shit about CPUS. ROFL God damn you're dumb No, the result was completely opposite of what you thought was going to happen. Quote Link to post Share on other sites
Remij 4,670 Posted April 20, 2019 Share Posted April 20, 2019 Just now, lynux3 said: No, the result was completely opposite of what you thought was going to happen. No ROFL And to top it off... you posted asking to see Game Mode performance as if it would improve the score Quote Link to post Share on other sites
lynux3 2,052 Posted April 20, 2019 Share Posted April 20, 2019 1 minute ago, Remij_ said: No ROFL And to top it off... you posted asking to see Game Mode performance as if it would improve the score Quote Link to post Share on other sites
Remij 4,670 Posted April 20, 2019 Share Posted April 20, 2019 10 minutes ago, lynux3 said: "Did you even put the threadripper into game mode?" Dumb bitch.. Game mode performs WORSE than disabling half the cores... ie Legacy Compatibility Mode... He has a physics score of 20K with Game Mode and 29K with LCM... It performs WORSE. We know Game Mode limits the CPU to a single die with the 2990WX... we know that the Physics test scales with cores. 8 cores vs 16 cores... it's not hard to figure out. Apparently it doesn't scale much past 16 cores though.. Intel's 28 core processors don't even score better than 30K. So yes.. I was right. Eat shit Quote Link to post Share on other sites
Remij 4,670 Posted April 20, 2019 Share Posted April 20, 2019 Quote Tests that are sensitive to clock rate and IPC throughput, such as VRMark, were a challenge for AMD's first-gen Threadripper processors. But we saw a big improvement from Threadripper 2950X compared to the previous-gen 1950X, which was expected due to the more aggressive multi-core turbo bins. The 2990WX's Game Mode reduces overall core count, but it also keeps bandwidth-starved cores from hurting performance. Nevertheless, Ryzen Threadripper 2990WX fell to the bottom of our chart due to lower per-core performance. Enabling PBO helped push it up to the middle of our test field. 3DMark typically scales well with higher core/thread counts. But the Threadripper processors, including the 32C/64T 2990WX, lagged Intel's line-up. The 2950X did enjoy a nice speed-up compared to AMD's older Threadripper 1950X. However, the 2990WX was hobbled by its Game Mode setting that turned it into an 8C/16C CPU. Both Threadripper 2 models realized solid gains from enabling Precision Boost Overdrive. 1 Quote Link to post Share on other sites
The Mother Fucker 26 Posted April 20, 2019 Author Share Posted April 20, 2019 that's essentially it. I wouldn't recommend 2990WX for anyone looking to game. My benchmarks shown here is the reason. AMD markets this CPU for creators and innovators, which I plan on to eventually doing once I get Sony Vegas and Photo Editing apps migrated over to it. The best AMD CPU bang for your buck without a doubt is the 2950X Threadripper. 16c/ 32 threads. Only goes for $839 now. My other benchmark shown here is the reason. Quote Link to post Share on other sites
Remij 4,670 Posted April 20, 2019 Share Posted April 20, 2019 Just now, The Mother Fucker said: that's essentially it. I wouldn't recommend 2990WX for anyone looking to game. My benchmarks shown here is the reason. AMD markets this CPU for creators and innovators, which I plan on to eventually doing once I get Sony Vegas and Photo Editing apps migrated over to it. The best AMD CPU bang for your buck without a doubt is the 2950X Threadripper. 16c/ 32 threads. Only goes for $839 now. My other benchmark shown here is the reason. Yea in the case of the HEDT segment anyway. 16 cores is the sweet spot price/performance-wise. Ryzen 3000 should be pretty awesome for gaming and really give Intel a run for it's money. Intel is going to be assaulted on all fronts very shortly. Ryzen is already doing that, but being able to take both price/performance and top performance crowns will push people even farther to AMDs side. Not to mention AMD has a clear strategy for their entire lineups in every segment and is continually improving while Intel looks to fumbling with getting past their current issues. Quote Link to post Share on other sites
Remij 4,670 Posted April 21, 2019 Share Posted April 21, 2019 Throw up your Timespy scores Quote Link to post Share on other sites
The Mother Fucker 26 Posted April 21, 2019 Author Share Posted April 21, 2019 When I get home tomorrow morning, I'll do post a new TimeSpy score. got a 11 hour shift tonight atm. Quote Link to post Share on other sites
The Mother Fucker 26 Posted April 21, 2019 Author Share Posted April 21, 2019 Looking forward to comparing a Timespy score I pulled early this morning right after I did my first Fire Strike with half Threadripper, to my workstation's current state. I think I stumbled upon some mining hardware I have that inadvertently provides a boost to gaming. Quote Link to post Share on other sites
lynux3 2,052 Posted April 21, 2019 Share Posted April 21, 2019 (edited) 2 hours ago, Remij_ said: "Did you even put the threadripper into game mode?" Dumb bitch.. Game mode performs WORSE than disabling half the cores... ie Legacy Compatibility Mode... He has a physics score of 20K with Game Mode and 29K with LCM... It performs WORSE. We know Game Mode limits the CPU to a single die with the 2990WX... we know that the Physics test scales with cores. 8 cores vs 16 cores... it's not hard to figure out. Apparently it doesn't scale much past 16 cores though.. Intel's 28 core processors don't even score better than 30K. So yes.. I was right. Eat shit He put Threadripper into Game Mode and got his best score. You're a dumbass. Edit: I'm a dumbass too apparently. I misread his results. Looks like we were both wrong. It doesn't scale with more cores and it doesn't perform as well in Game Mode either. Edited April 21, 2019 by lynux3 Quote Link to post Share on other sites
lynux3 2,052 Posted April 21, 2019 Share Posted April 21, 2019 3 minutes ago, lynux3 said: He put Threadripper into Game Mode and got his best score. You're a dumbass. Edit: I'm a dumbass too apparently. I misread his results. Looks like we were both wrong. It doesn't scale with more cores and it doesn't perform as well in Game Mode either. ie. it doesn't scale with more cores like you thought it would with Threadripper. Too much latency with infinity fabric, which is expected, when it comes to these benchmarks. Quote Link to post Share on other sites
lynux3 2,052 Posted April 21, 2019 Share Posted April 21, 2019 (edited) 2 hours ago, The Mother Fucker said: that's essentially it. I wouldn't recommend 2990WX for anyone looking to game. My benchmarks shown here is the reason. AMD markets this CPU for creators and innovators, which I plan on to eventually doing once I get Sony Vegas and Photo Editing apps migrated over to it. The best AMD CPU bang for your buck without a doubt is the 2950X Threadripper. 16c/ 32 threads. Only goes for $839 now. My other benchmark shown here is the reason. Typical for the way Threadripper is designed. Infinity fabric, as badass as it is, doesn't scale well for this type of workload. Beast CPU for tasks that you're looking to tackle though. With the possibility of Ryzen 3000 coming with a 16c/32t option (all but confirmed with missing die on the package) and the increase in IPC in general with Zen 2 architecture that'll probably be the best bang for your buck... that is unless you absolutely have to have an HEDT CPU. Edited April 21, 2019 by lynux3 Quote Link to post Share on other sites
lynux3 2,052 Posted April 21, 2019 Share Posted April 21, 2019 16 hours ago, Remij_ said: I think I understand exactly. 3dmarks physics test scales with cores dumbass. It will almost certainly lower his physics score... @The Mother Fucker Now enable all the cores, and run using only game mode and post the results. I want to shut this dumb bitch up Enable all the cores, Game Mode will lower teh physics score. Not only did Game Mode increase his physics score, he got his best graphics score in Game Mode. Only enabling legacy mode upped his physics score which you weren't even aware of, but subsequently lowered his graphics score. You don't understand how Threadripper works. Quote Link to post Share on other sites
Remij 4,670 Posted April 21, 2019 Share Posted April 21, 2019 2 minutes ago, lynux3 said: ie. it doesn't scale with more cores like you thought it would with Threadripper. Too much latency with infinity fabric, which is expected, when it comes to these benchmarks. No.. it absolutely DOES scale with more cores.. in general, and on Threadripper. His post PROVES that. It simply doesn't scale past a point which has nothing to do with Threadripper and everything to do with 3dmark itself. The original Firestrike CPU test apparently doesn't scale past 10 cores. With Timespy Extreme their CPU test was updated to support more than 10 cores. You saying saying "Did you even try game mode" means that you thought it would perform better. You apparently knew game mode would reduce the CPU to a single 8 core die... right? Which is why there was barely any difference between that an the full 32 core test. Game Mode makes it perform worse... straight the fuck up. Quote Link to post Share on other sites
roflpwnedz 289 Posted April 21, 2019 Share Posted April 21, 2019 why the fuck do you have 2 vega 56'es sell me one Quote Link to post Share on other sites
lynux3 2,052 Posted April 21, 2019 Share Posted April 21, 2019 9 minutes ago, Remij_ said: No.. it absolutely DOES scale with more cores.. in general, and on Threadripper. His post PROVES that. It simply doesn't scale past a point which has nothing to do with Threadripper and everything to do with 3dmark itself. The original Firestrike CPU test apparently doesn't scale past 10 cores. With Timespy Extreme their CPU test was updated to support more than 10 cores. You saying saying "Did you even try game mode" means that you thought it would perform better. You apparently knew game mode would reduce the CPU to a single 8 core die... right? Which is why there was barely any difference between that an the full 32 core test. Game Mode makes it perform worse... straight the fuck up. No one is arguing that it doesn't scale with cores. Game Mode not only performed better in the physics benchmark, it gave him his best graphics score. It has everything to do with Threadripper and infinity fabric, which you weren't aware of and that's okay. There are flaws in its fundamental design for synthetic benchmarks like this which is why AMD even patched in a Legacy and Game Mode. And it DID perform better in Game Mode. Game Mode: Graphics: 36,276 Physics: 20,821 All Cores (Like YOU Suggested): Graphics: 21,533 Physics: 17,121 Quit being a dumbfuck for once. Quote Link to post Share on other sites
Remij 4,670 Posted April 21, 2019 Share Posted April 21, 2019 42 minutes ago, lynux3 said: No one is arguing that it doesn't scale with cores. Game Mode not only performed better in the physics benchmark, it gave him his best graphics score. It has everything to do with Threadripper and infinity fabric, which you weren't aware of and that's okay. There are flaws in its fundamental design for synthetic benchmarks like this which is why AMD even patched in a Legacy and Game Mode. And it DID perform better in Game Mode. Game Mode: Graphics: 36,276 Physics: 20,821 All Cores (Like YOU Suggested): Graphics: 21,533 Physics: 17,121 Quit being a dumbfuck for once. You're so fucking stupid.. god damn. I was obviously referring to his first bench.. which is all cores. First bench (All cores, non game mode) Physics: 19,192 Game Mode Physics: 20,821 How does his physics score drop from 19K to 17K?? Throughout his benches he's got physics scores ranging from 14K to 19K with all cores... He's been messing around with settings. The ONLY difference there... being that the CPU was actually overclocked 200mhz more.. to 4.2GHz.. lol Either himself overclocking, or PBO is boosting differently and affecting the scores. (probably due to thermals being affected, cpu being warm already..ect) Not fucking game mode. and The "combined" score is worse with Game Mode no matter which "all cores" test you want to use. Meaning that game mode isn't doing SHIT. The only reason why All cores scores aren't better is because the actual fucking physics test isn't able to scale beyond 10 cores... not Infinity Fabric you clown Which is why I want to see him post some Timespy Extreme scores. Game mode on and off too of course Quote Link to post Share on other sites
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.