Hardware Unboxed

In that video we used FG to show how poor performance even with FG was, as this game lent on the technology heavily for RT support. Numerous times the disclaimer that FG was used for those graphs was given, and we used it to show that for most configurations performance with FG enabled was bad. "anything below 60 fps with frame gen enabled is bloody awful, and that makes sense. Ideally you want a base frame rate of 60 fps which roughly translates to 100 fps in this graph." This sort of thing was mentioned at least 4 times while discussing those results. We then showed three presets at native resolutions with any upscaling or frame generation and treated that as the real data. None of this fits your narrative that we are paid shills. We've been very much against using FG at frame rates below 60 fps and have heavily pushed back against marketing making performance claims with FG enabled.

@HardwareUnboxed @lil_skittl14249 youtu.be/kOhSjLU6Q20?si… Black myth Wukong, benchmark, 43 Gpus. You started off in benchmark format testing out Max settings with RT + FG. You can gaslight/make excuses as much as you want. This was titled as a benchmark video, in a benchmark format, and you used FG as part of it

Lego Batman system requirements include ridiculous 30 and 60 FPS frame generation configurations. So, it's time for a reminder: frame gen doesn't fix bad performance! youtube.com/watch?v=Dn9zNs…

Okay so you're just going to continue to lie then. We have never once benchmarked with FG enabled, outside of our FG video showing the increased latency and the fact that it doesn't really boost performance and is merely a smoothness feature. We were the first tech channel to explain that all it does is smooth the image, while introducing artifacts.

@HardwareUnboxed @lil_skittl14249 It's not about you guys being "known" for using FG in performance benchmarks. The fact is you guys have used it before in performance benchmarks, so why are you guys now crying when Devs are now using FG as a performance crutch?...makes no sense. You guys are part of the problem

Yes, we are known for using frame generation in our benchmarks, but to be fair we only do it when shilling for big tech, gotta pay those bills my guy. Anyway FG is awesome, you don't need to optimize games, it just works. Quit your bitching and just turn FG on, like we always say.

@lil_skittl14249 @HardwareUnboxed Exactly. These are paid shills when it suits them. These "tech" channels use frame generation in their performance benchmarks, yet they are crying about Devs using frame generation as a performance crutch. You had digital foundry also pushing FG ( Nvidia FG specifically).

@gigadenza1 Not sure, I never noticed. There is always drama when testing the 14900K. Anyway, that video looked pretty well done, so there you go.

Hmm, wasn’t it Danny who, not so long ago, criticised a benchmark that you conducted pitting a Core i9 14900k against one of AMD’s V-Cached Zen4/5 Ryzens - I believe it may have been the mono-died 9800X3D. Anyway, his own comparative test (in one of the most recent triple A ROP rippers) was run with the same pair of CPUs, and featured a timed sequence of actual gameplay that replicated both the detail levels and resolution you used - but showed a smaller average FPS discrepancy between both “fully loaded” chips, thereby suggesting you were running the Intel part with more conservative “stock” settings, perhaps without the MB’s revised “performance” profile, enabled, which would have raised both the CPU’s power limit and sustained multicore turbo frequency. You claimed that both your test rigs had been left configured at their standard settings - with EXPO/XMP memory profiles activated - but also expressed your intention to carry out a more detailed analysis involving both LGA1700 and 1851 MoBos, which would determine whether or not there was any variance in the performance that both their default and “optimised” BIOS presets provided when used with specific CPUs. Not sure if this ever materialised or any further drama evolved…but the whole exchange certainly engaged much of the itchy-fingered tech community’s wandering attention throughout several Earthly revolutions!

I’ve seen the quoted comment below a few times so I’ll quickly address it here. I suspect someone has taken what was said out of context, trying to make a big deal out of it for attention. The 9800X3D and all 3D V-Cache parts really are far less sensitive to memory performance and in the vast majority of games will deliver similar performance without even having to load memory profiles. Does that apply to every single game? Probably not! Does it apply to the vast majority of games? Yes! ALSO I never said the exact same performance, “You don’t go into your BIOS and load your memory profiles and you’ll get performance that is very close to what is seen in reviews, which is pretty wild because never before have we really seen a CPU where that is the case” For the most part that is correct. Non 3D V-Cache parts like the 9700X for example are much more memory sensitive and will see a much more significant performance decrease when not loading memory profiles.

@F0RDB0Y @DieHarteWare @JeemerTheMeemer ..... and you know what, we never said it wouldn't benefit some titles. It's almost like we were talking about performance across a range of titles, like a 14 game average or something. I will mute you and move along though.

@HardwareUnboxed @DieHarteWare @JeemerTheMeemer I'm saying that it is more beneficial in some titles. What is so hard to understand about having the extra performance when the situation arises. And even if it's only 10% in some titles, combined with PBO and GPU undervolt it adds up. Take your own advice and move along then.

Yeah stupid reviewers, everyone simulates complex fluid behaviors, it's like a mainstream hobby. Stupid reviewers. Better luck reviewing the 9950X3D next time!

@DieHarteWare @F0RDB0Y @JeemerTheMeemer He's using an outlier to disprove a general statement, and worse still, he's well aware. These people aren't worth the time of day, ignore them and move on. I made the mistake of trying as well.

@F0RDB0Y @JeemerTheMeemer @HardwareUnboxed that's a simulator game known for hammering the cpu. obviously it loves ram tuning. completely worthless for general recommendations. and again tested with low settings. finally learn how cpu/ram benchmarks are done.

@stepheneast That is my written content.

@HardwareUnboxed Hi i think there maybe a website copying your benchmarks techspot.com/review/3110-cp… they also had the x3d one u did

Would you say it is reasonable to make the general statement that modern overclocking results in a "negligible performance increase" and is therefore not worth the time and effort? Based on the polling data below people seem to agree. In extreme examples you can boost gaming performance by 20% but that is very much dependent on silicon quality and hardware configuration, it's far from guaranteed and even then most people don't want a tuned 14900K with +20% performance to maybe match the 9800X3D in certain games, the power usage still sucks and you have to worry about stability long term.

@F0RDB0Y @JeemerTheMeemer @DieHarteWare The comment I made wasn't in the review. It was a general comment about an advantage of X3D chips not being memory sensitive. It was taken out of context by idiots.

@HardwareUnboxed @JeemerTheMeemer @DieHarteWare Timings still have a bigger impact on 1% lows than speed does on X3D. The issue is your messaging could be taken as not to even bother enabling EXPO by people who don't know any better. Which is dumb.

@JeemerTheMeemer @fyslalq89876404 @Ashraf_Medhat93 faisal's time is of such low value that you should invoice him per interaction.

@fyslalq89876404 @Ashraf_Medhat93 @HardwareUnboxed tuning is negligible gains, you will not get a 20% gain in real world scenarios. and memory tuning is not free, not sure about you, but my time is valuable, not gonna waste a week testing memory OC for 3-5% real world gains in games.

@JeemerTheMeemer @F0RDB0Y @DieHarteWare It's funny how he just proves what I am saying, overall, for the most part, the eprformance you're going to see without loading EXPO will be similar to what reviews showed. Yeah, there will be a few outliers, there always are, but overall it's not going to be miles different.

@F0RDB0Y @DieHarteWare @HardwareUnboxed you've just shown that it doesn't. less than 4% increas vs JEDEC standard. 6000 C30 would be within margin of error of the EXPO score in that chart.

@Pk_419 Our reach is impressive. Especially on all that content convincing people all they need is XMP.

@HardwareUnboxed You've spent the past 10+yrs convincing people that all they have to do is enable XMP. Of course they aren't going to think it's worth it. Spend 1yr showcasing performance improvements and I bet the opinion would change.

@Charcharo I think there's at least 100 more interesting topics that we can cover.

@HardwareUnboxed There are no hardware releases of major interest coming up. Might as well do a guide for CPUs both AMD and Intel.

@asdtyuv @davepl1968 Are you talking to me?

It depends on the CPU. Simply turning on BIOS optimizations takes a 96-core Threadripper from 350W to 800W with a significant jump in multicore. So no, it's not reasonable.

@davepl1968 Skill issue! (now I sound like a real overclocker)

@HardwareUnboxed Well, I did run 100 copies of Doom on it at once. But I found it overwhelming to play them all simultaneously.

@fyslalq89876404 @Ashraf_Medhat93 It's impressive because it's out of the box, and uses very little power. It was also impressive because the rest of Zen 5 sucked, so we were more talking relative to expectations. Anyway, I am truly done with you, you're a waste of time.

@HardwareUnboxed @Ashraf_Medhat93 "Truly impressive over what was already the world fastest gaming cpu" it was 12% , mean while people with avg bin cpu getting 20% with all cpus, see steve your words dosent make sense is 12% impressive or negligible 🤔

We were having a free flowing conversation, our opinions on this subject aren't as black and white as you seem to think. There are multiple reasons for why we don't focus on overclocking and why the vast majority of PC enthusiasts don't care for it. I'm not sure why you're so hung up on the "negligible performance increase" comment. When discussing PC hardware in general for the most part the OC gains that can be had are negligible. Again this is why most people aren't interested. Back in the day when you could get 50%+ from a basic overclock that content performed extremely well. These days there is almost no appetite from OC content, it just feels like it brings out the most toxic people in the community.

@HardwareUnboxed @Ashraf_Medhat93 Oh so now its changed from "negligible performance increase" to it depends on the chip its insane dealing with a person with this mentality lol, steve its totally okay to say hey wrong about this but changing the goal post endlessly is so crazy to me

"misinformation" is such a cope. I disagree with this person, who happens to agree with most other people, therefore misinformation. The misinformation is telling people they will gain 20% more performance by tuning their 14900K, as if so say all CPU and configurations will achieve those gains and there's no string attached. Meanwhile we're showing you the performance AMD and Intel guarantee. Get a grip.

@Ashraf_Medhat93 @HardwareUnboxed My problem isn't with xmp btw, its the misinformation if you follow conversations I had with him it would be clear to you, he said negligible performance increase a modern chips while people having 10%+ from tuning for free, but spending 500 just for 8% is totally fine