Responding to Ridiculous Comments from Gamers Nexus & Hardware Unboxed on My Last Video

[@RATechYT]

*UPDATE:* Steve Walton decided to delete his tweet as well as the comment he left under that video after my response. Check out my previous video here if you haven't seen it: kzbin.info/www/bejne/pYGlm6ykbdOee7c

[@chilldudie242]

Lmao wow chill bill was right they're narcissists

[@elemkay5104]

Hey bud. I appreciate your videos and comparisons; I just want to point out something in this vid I noticed. You said that with i3 8100 vs the i7 2600, extra threads in the 2600 would mean better performance. But think about it - with games that don't use multi-threaded performance very well, the opposite would be true. Games that can not fully utilize those 8 threads would get a benefit from the same performance being achieved in just 4c4t. It's the same with games that run best with high single core clock speed; only you're applying it to 4 vs 8 threads. You get what I'm saying? Just wanted to point that out. Otherwise good stuff. Good to hear that Steve Walton deleted his comment as well. I think he's probably alright, but got the wrong end of the stick.

[@RATechYT]

@@elemkay5104 For the most part the 2600k is going to be better. Games that care about single core are going to run a bit better on an i3-8100 compared to the 2600k, but games that care more about cores are to be much more playable on a 2600k not only due to a higher frame rate, but also a more consistent frame time and extra CPU headroom.

[@shadowopsairman1583]

I stopped watching GN because of FUD spreading

[@itpugil]

@@shadowopsairman1583 FUD? What does that mean?

[@czbrat]

Although I came over from HUB's tweet, I ended up subbing and watching most of your videos lol

[@RATechYT]

Appreciate it!

[@666Necropsy]

its nice to see people who actually look into the information being presented. this community could grow really good with that.

[@drumsmoker731]

I'm with you on this. If they were right, I should have abandoned my 8350 a long time ago.

[@ConfusionDistortion]

Looking back on this, it's gotten pretty hilarious how the "big tech reviewer" scene pretty much turned on itself with drama and infighting, with none of them really working together anymore. And in retrospect, how Gamers Nexus, outside of reviews, relies on a lot of critical hit piece stories over products and companies to boost video views past 200k when their channel goes through lower viewership lulls. Meanwhile Digital Foundry, who has been making an argument for stable frame times for years, is doing better than ever.

[@N3xlow]

i don't think you should apologize for the fx video

[@ihavenoaffinity]

really lost a lot of respect for Hardware Unboxed and Gamers Nexus after this. I don't even think I would have cared whether their results were misrepresenting the FX or not, but their replies to you came across as extremely unprofessional and butthurt.

[@usmanshah8395]

You are being bullied for having an alternative methodology ultimately... They didn't like you mentioning their name and tried to defend it with some level of hierarchy science approach, trolling as they felt you were trolling them ( just pointing out alternative testing results and that not all hardware tests show the same results ) You are a good guy, a techy, enthusiast and deserve a little more credit for your efforts. I have lots of suggestions for your channel but as a person who could happily afford any new gen hardware 3 x over, I'm still pushing a FX8350 at 4.8ghz on Air with a Rx580 daily with a rock solid setup, games happy at 1080p there isn't much it can't do after all these years... Although an upgrade is probably on the horizon as the IPC for single cores has vastly improved and demands on the GPU are growing.

[@reki353]

There is a reason why scientists and the scientific method exists, Computers are not like cars where all that matters is how many G's it can pull or how fast it goes from 0-60. ANY little thing can affect performance, There is no such thing as a "different" testing methodology. There is only the right tool (in this case testing methodology) for the job.

[@usmanshah8395]

@@reki353 theory is open for discussion, as much as I respect your example and hear what you are saying, that we shouldn't necessary go to a layman for medical advice if doctors are available but then a herbalist may have the answer but be ignored, because he's supposedly not a 'doctor' . I'm not saying it is the best methodology used but certainly shows that the variance and performance can and will be affected by many variables.

[@RaimarLunardi]

It's more he is being Bullied because people misunderstood how he tested.

[@rudysmith1552]

You can’t simulate an Intel haswell dual core on a zen + architecture.

[@luisjp4782]

I recently bought an FX8350 because they sold it to me very cheap, and the truth is that after trying it in several games and taking into account that I have an R5 3600, I can say that the FX8350 has been undervalued and that it can still give a decent experience, and that it has aged very well.

[@derek8564]

@@lukatosic4447 dude...what's more important....8350 or keeping your food from spoiling? Let us know how the 8350 benchmarks :)

[@todogaming779]

@@derek8564 i do in my channel, a lot of test on my 8320 oc to 4.4 ghz

[@reki353]

there is no such thing as a bad product, only a bad price. Even the most shit computer components become good deals if the price is low enough

[@CharcharoExplorer]

I honestly lost a lot of respect for GN and HU. Their Twitter rants and drama are also annoying. Same for Tech Deals.

[@ihavenoaffinity]

Their responses hurt their own reputations far more than RATech's video did. I didn't even notice RATech's original video on this and I've been subscribed ever since I found his FX overclocking videos when I OC'd my 6300.

[@ytszazu2gmail381]

Yup tech deals was way more fun to watch when he benchmarked the Rx 560 times.

[@Sixkill3r]

I agree I am honestly surprised that gn would say it like that...as he just criticized Nvidia for being dumb....

[@benuten3907]

Celebrities are expected to act like one..seems YTubers/Tech influencers are really mature when they say they are "10 years+++" in the industry..

[@ygorbachev]

Same here, expected more professional response from both of them. As industry leaders with "10+ years" of experience it is expected that there would be more leadership and less whining.

[@derek8564]

The guys from hardware unboxed act like children whenever they get challenged.

[@Dj-Mccullough]

True.. alot of the time they are bitching legitimately though.. im not sure why both steves have such bad results on their 8370s.. ive gotten similar results to RA here with my 8350.. something isnt adding up. Isnt the 8370 just a higher binned 8350? Or are there differences?

[@longmuskox4194]

If it walks like a duck and talks like a duck, then it must be a duck.

[@RATechYT]

@@Dj-Mccullough The 8370 only has a 100 MHz higher turbo boost.

[@deaniliev7648]

Is HUB deleted their comment from original FX video?

[@RATechYT]

@@deaniliev7648 Yes.

[@AnotherLotte]

I'm just disappointed in them, none of what went down should've happened in the public space imo. Doesn't help that they announced their misunderstanding to their audience to fan the flames.

[@idogalisr]

Around two years ago, I wanted to get a new computer. I had a Athlon X4 760K, which was a bad purchase. I was deliberating VERY hard between getting a used AM3 platform (which would've allowed me to keep my RAM) and getting a new AM4 platform. I ended up getting a B350 motherboard and a Ryzen 3 1200 cpu that I overclocked to 3.85 ghz, because all of those site were telling us that the new architecture is so much better, regardless of the core count, etc etc. A short while ago I upgraded to a Ryzen 5 3600 because the Ryzen 3 just CANNOT keep up with newer demanding games. I hindsight, if I had gotten a FX 8350, it would've been LESS MONEY + no need to upgrade now(!!!). That's why the work of channels like yours is so important. Thank you.

[@RATechYT]

Thanks for sharing your experience. I'll be comparing the FX-8350 to the Ryzen 3 1200 in the near future.

[@gilbertrodriguez6053]

Heaven forbid your results differ from GN's or HUB. I dropped those two channels over a year ago. I personally am sick of their static bar charts showing avg. FPS. Show actual ganeplay numbers with avg. .1% lows. That's what we all experience here in the "real world". Yes, the content the other channels takes work, alot of work. Aren't we done with the static ppt charts in 2020/2021? RA hold your ground sir.

[@CrongTech]

Would like to disagree with your static charts for benchmarks, personally im ok with it since i am only concerned with the numbers. Id watch a different video if im looking for gameplays

[@Andriyshchurko]

...how else do you want to represent the performance of PC hardware in games? Average and low percentiles are the cleanliest way to do so, with clear apples to apples representation on how fluid and smooth game runn on different hardware. Any "hey, I'm getting more fps on this cpu in this moment than on that cpu in that moment" comparisons are easily skewed, as the fps will depend on what exactly is being processed and rendered on screen. Hence any sysntetic or ingame built-in benchmark takes the total/average across the entire run, with sceene composition and camera movement being basically the same every time.

[@Oversoulse7en]

This is not the first time that steve from HU behaves like a child and I can assure you that it will not be the last.

[@longmuskox4194]

He's a man-child. No question about it.

[@AlexM-uh7sq]

🤣🤣

[@Mistereify]

Remember him answering to every arguments in favor of the gtx1650 in its comment section. I left his channel after that. Useless video from an AMD fanboy

[@longmuskox4194]

@@Mistereify Steve ain't no AMD fanboy. He's only beholden to those who pay him, including all his fanboys.

[@link6563]

@@longmuskox4194 It is like the mafia you pay up or else

[@RafaCoprimo]

impressed by how HU and GN replied, disapointed to say the least (embarassed actually) you're right in your methodology, showing real time footage, after all we play the games, we don't play with static graphs for me no matter how big or influencial a youtube channel is, respect must be equal to all people here, i've unsubscribed from them since they can't handle constructive criticism, everybody makes mistakes sometimes, they somehow think their results are absolutely right don't worry about negative comments, you are not losing anything that's worth it from there, here in brazil i'm acostumed with the hardware/pc community having drama, and youtubers from brazil thinking they are superior to everyone else because they can use msi afterburner

[@owningdishonestshills7435]

Most big time youtubers become arrogant.

[@Stermy57HW]

This is why I decided to delete my subscription from their channels months ago. Now I can say that it was a good decision...

[@b.henriques9871]

You handled it like a fucking PRO

[@Eggwhites94]

I wasn’t expecting them to come for you that hard when you made really good points. I watched there videos on FX and I never understood why they got those results. I have a fx 9370, an a i5 2500k and my fx blows the i5 out of the water on new games. I think it’s just easy to make a hate video on FX and not have to worry about people questioning it because.. well every other big KZbinr talks about how FX was the worst thing since hitler. You gained a sub for me, keep up the videos!

[@666Necropsy]

i dont want to claim intent but it seems there is a goal of selling hardware. i have been able to use older hardware for years past what others may claim. this was such a clear example of how something isnt adding up right. i have witnessed the same results with different gear and never knew how to explain it. im sure software is playing a huge role. there are many factors to consider in this. having a huge list of benchmarks may be convenient but may not give real world results. benchmarks are benchmarks. how easily can they be manipulated. check out newer motherboard bios settings where you can improve certain benchmarks by name.

[@TheScure]

Everyone was busy commenting on the "simulated" CPU. But it doesn't make the FX numbers any different. One of the main point of your video was to test FX chips, because in GN and HU tests the performance is way too low and weird. You didn't even try to cheat with some super cooled PC with crazy clocked RAM and high-end VRM motherboard+OC. In fact you had worse stuff in all of these aspects, even GPU. You proved that the performance is much better than in their benchmarks. Numbers speak for themselves. Instead of talking they can make new benchmarks to show their results again.

[@ickyconcrete5370]

I think you have hit a nerve with them, but tbh they clearly dropped the ball and want to save face.

[@chilldudie242]

Too bad words can't be taken back :-\

[@AlexM-uh7sq]

Well done. But lets be honest here... they should be apologizing to you.

[@MrApple256]

I totally agree with you. I watch Hardware Unboxed sometimes but this kind of ignorance from them is so bad. I don't know Gamers Nexus but their comments are just stupid and I don't understand how such channels with almost "professional" videos can respond to criticism so horribly. I have an FX 8370 and it's still fine even for modern games.

[@robh9075]

I also watched the entire video and it seems to me that GN and HU were responding to comments from their own followers rather than watching your actual video, if you're "busy" enough to not watch a video then you're too busy to post a long ass comment shit talking another creators content. You all ran around flapping your arms in the air and squawked about how badly Nvidia treated HU "the little guy" well I am sorry to say that Hypocrisy thy name is Steve.

[@666Necropsy]

oh how the tables have turned.

[@robh9075]

@@666Necropsy Yep pretty dickish responses from HU and GM, I haven't subbed to either channel because I kind of felt both channels came across as more than a little conceited, now I know it's considerably more than a little.

[@Dj-Mccullough]

Part of this is that, I think steve from gamers nexus and hardware unboxed have internalized the idea that the 8350 is only actually a 4 core 8 thread cpu because of the lawsuit. Unfortunatly that idea is flawed. There is no need for ANY floating point unit to be present in a cpu core. Only an Alu is needed to be considered a 'core'. Up until the 486 released no fpu was even in the cpu silicon. They may not realize the architecture is made to use 8 cores and share an alu between groups of 2.. that is a different layout fundamentally than a cpu with 4 cores and 8 threads where Intel hyperthreading uses the same core(which in this case contains a single alu and fpu) to perform back to back operations on the same alu per clock. This allows 2 threads on one alu. Fundamentally different from having each thread operate independently in their own individual alu's x8 of them.. if a instruction then needs, it is sent to the less commonly needed fpu. The part that troubles me, is i have had several 8350s.. only one left now and its due for upgrade soon. I have never been able to reconcile the performance difference i saw in gamers nexus and hardware unboxed coverage, againt my own NOT overclocked 8350. Their performance always looked terrible to mine and i didnt understand why. Hell i even turned off thr boosting behavior on mine for cooling and it still didnt approach thr amount of suck we see in their 8370 reviews. What is the difference between these two chips? I had always thought the 8370 was just a higher binned 8350. Perhaps we should be seeing if there is an undiscovered difference between the chips? Perhaps see WHY the 8350 seems to be scoring better than the 8370. Oh, and my offer on those graphics cards is still open :) -salute.

[@RATechYT]

Oh right, it was you who emailed me! I actually replied to you around a month ago, haha.

[@mtunayucer]

I cannot believe how incredibly stupid people can be, even our favourite youtubers. People! Stop following popular ones opinions and develop your own opinions! I understood the original video like this: An 8 Core FX CPU still has more power than "modern" 2 core 4 thread parts despite having a lot worse single threaded performance.

[@Dj-Mccullough]

My 8350 didnt perform as badly as the steves' 8370's did. The 8350 now lives as my best friends wifes pc. It games circles around her old 2c4t system. Same gpu.

[@reki353]

an FX cpu does not have more power than a modern 2 core 4 thread cpu

[@xerxeslv]

@@reki353 What if modern part have very limited frequencies for example? That's how product segmentation can make your comments silly. Check i3 1115G4 (Q4 of 2020 btw).

[@kabutosjutsu]

What´s the point of comment a vídeo you didn´t watch, I mean "simulate an FX with Ryzen" it´s clear they don´t even take some time to analize it, so dissapointed of them, totally support you buddy!!!! FX is totally worth it even in 2021 (if u still have one).

[@Zorro33313]

We won "the battle of FX defense" around 2015 in my country, looks like now it's time for the eastern world to admit the reality. Your struggle won't be forgotten, sir, fight on. And you're not alone, if i can help you somehow, tell me. Truth will prevail.

[@idogalisr]

What a condescending response by Gamers Nexus. Self absorbed, disgusting and unprofessional. I unsubscribed from their channel IMMEDIATELY after seeing this. I think a lot of us are fed up of those channels, they have their infinite supply of high-end gear and have no awareness of the needs of those with lower income, or just people that don't want to invest so much money on gaming gear. Thank you for your great videos. Keep on being humble and approachable. It's sad seeing those guys turning up to be arse holes with their success and influence.

[@Turktien]

Who really cares what they have to say? The fact is they only cater to use cases of people who have tons of money to throw at a rig would be interested in. They are more concerned with being right than considering other use cases. I'm still using an 8350 OC'd to 4.73 GHz paired with a RX 580 8GB @1420MHz and it's meeting my 1080p gaming needs (for now, at least). We appreciate your videos! Alternate perspectives are great to have and we appreciate that you give us that. Steve is very full of himself and, quite frankly, that's why I don't watch his videos. To put it colorfully... he really just likes to sit around and enjoy the smell of his own flatulence and I've got other things to waste my time on. :) Keep up the good work!

[@deansigman6099]

I watched the original video and it was clear you were using a FX 8350 and not simulating it with a 1600af. Also, it was clear you were simulating the AMD 3000g and not the Intel counterpart. did they even watch the video or listen to what you were saying at all?

[@RATechYT]

Doesn't look like they did.

[@deansigman6099]

@@RATechYT this is why I've started watching more of the smaller tech channels over the big guys.

[@ElJosher]

I can’t believe they actually went on twitter rants. How immature. I didn’t see your video as an attack on them, but I guess they did and got very salty about it. I don’t really watch tech youtuber that much, but i had a bit of trust for HUB. Now it is kind of gone. I stopped trusting GN when he claimed that the ps5 is just a mid range pc from 5 years ago. I have a top mid range pc from 3 years ago and it doesn’t even come close to running games like the ps5.

[@cdiff07]

That is where you fail to understand the basics my friend. It is easier to optimize a game on a closed operating system that has the same hardware for every end user than it is for a open OS that has tons of different Mid range set ups. Not to mention, a lot of the performance on PS5 is not native resolution etc.. Meaning that all the Ps5 really is a really good 1080p maxed out settings PC, please do not mention that it can do raytracing, there are several proponents to raytracing and being able to do one with ok quality really is more gimmicky than anything, just like if you buy a mid range gaming card for PC you should not be playing with raytracing on!

[@aolson1111]

It's bullshit, no matter how you try to justify it. A midrange PC from 7 years ago would be a gtx 960, a quad core Intel 6500, and no NVME SSD. The PS5 has the equivalent of a ryzen 3700, rtx 2070, and an NVME SSD with dedicated quick decompression.

[@Flacon500Spat]

It's pretty clear they didn't watch your video. Like, yeah the later FX series sucked, but it wasn't terrible as much as HU and GN are making it out to be. Their bias is clearly showing when they initially replied to you they came in like a smart ass knowing it all and said you were wrong 100% before hearing your argument that they themselves could be wrong in their results.

[@magoid]

I like to watch HU benchmarks when a new CPU/GPU is launched, at last he doesn't talk too much as the Steve from GN. But I don't subscribe to HU. Why? Because a couple things he does that I don't agree. First, is some disdain for the lower end of the market. The true lower end. When you receive a lot of top of the line hardware for free, is easy to create a distance from the typical PC user, who doesn't have funds to equip his/her rig with the best AMD/Intel/Nvidia have to offer. Second, Steve from HU didn't wanted to investigate why the first Threadripper generation was performing badly after launch. When people on the comments said it was a Windows scheduler problem, and that he should test it on a Linux distro, he just dismissed it like it wasn't true and something worth pursuing. For him, it was easier to just badmouth AMD for something that was Microsoft's fault. I'm not saying the guy is a douche. I totally sympathized with him when Nvidia tried to force a editorial change down his throat. But people have to realize that just because you simpatize with some youtuber, it doesn't mean they are perfect and everything they do is flawless. Just enjoy the content they provide for free and learn not to blindly defend people you admire like they are infallible.

[@RATechYT]

Absolutely agree with the last part.

[@benmcreynolds8581]

I'm currently running a fx 8350 with a GTX 980 strix and honestly I'm always finding myself surprised that it can handle anything I throw at it. You just have to keep in mind that it's best suited for 1080p gaming and not maxed to the ceiling graphics settings, just for the sake of quality and consistent balanced performance. Yes your not going to get 150fps on most new games but 90-70 range is very common and shit that makes for a pretty great gaming experience. We're almost nitpicking fps stats after that point. Plus the CPU ha been great for multitasking, multiple programs running, multiple internet windows playing streams at the same time with chat and dual monitor as well. I mean for how old this thing is and how much hate it gets I'm surprised it hasn't earned the respect of "the patient diamond in the rough cpu" it's cheap, consistent balanced workhorse with great performance and can game, even with modern games. That's be like if a sega dreamcast later found out it could also run games that came out in 2021... Like come on give the tech some credit even though yes tons of other CPUs have higher stats. Not everything is a graph bar with numbers. The bad boy works well and I'm enjoying the deal of a gaming PC I got not too long ago. Am I going to upgrade at some point... Yes .. does that take away from how much respect I gained for this old cpu... No. I'm Always gonna respect this fx 8350.

[@doghunter4097]

You have to understand, no one in the world of computer sales is interested in keeping the old FX processor up and running for a long time. Marketing manipulation!! I'm using the FX8320 (OC4.2 (currently)), and I don't plan to replace it with a new processor because the FX still performs its function very well. What's more, its efficiency has increased lately ... precisely because of the strange architecture it possesses. Btw, good video.

[@joshuac837]

I haven’t recommended a dual core (or quad core without hyperthreading) since I’ve started watching your content. While some games may like or not like those lower threadcounts, an FX-8xxx will be much better over a wider spread of games than one of the simulated CPUs. Thank you for your testing and dedication, I’ll continue to watch despite the comments from the bigger guys.

[@RATechYT]

Much appreciated.

[@suiken3149]

I'm revisiting this video because Gamer Nexus is once again having another drama with another tech tuber.

[@ficklefingeroffate]

Steve Burke is more than a little full of himself.

[@TouchthisiProduction]

with Gamers Nexus and hardware unboxed calling out LMG right now. Maybe someone should call them out for this aswell then.

[@StrangersIteDomum]

I watched your fx vid because youtube recommended it and it's good to see challenging opinions. It would have been more appropriate for them to revisit the part if they were to respond, but honestly, fx CPUs aren't for sale and aren't for their target audience. Since I'm interested in refurbing cheap PC's, this was a great vid for ideas. Good job for at least bringing attention to potential resource.

[@thedeltaside5719]

As someone who has a few fx cpus they are still very capable simply a good clockspeed and a bump on the northbridge and ht link is what gains the most performance even on the first bulldozer cpus, yeh the 3770k at the time was faster in games but you're fire up something newer that can utilise the whole cpu it's a different story

[@gravitone]

It is clear from Steve's comments that they haven't touched an FX cpu in years.

[@rodrigomendes3327]

Yeap. He probably think old i5 are still better than FX8 even on modern games lol. I hate Nexus chart because it very clear that he just use some math to calculate some results, obviously not a real test. It need too much time to test on old CPU and they just don't have time to do it.

[@reki353]

Steve uses an FX cpu in his main computer for many years

[@rodrigomendes3327]

@@reki353 Surely not for the last 2 or 3 years since he's clearly outdated.

[@charlieplagueis9531]

@@rodrigomendes3327 steve from gn has a 8370 in his personal system lol

[@pharettface]

Yup I totally agree with you many people did not even watch the video or listen to what you had to say! I have said and commented on both of their videos for years saying that something did not seem right with the FX8370 benchmarks. I used one for years and never had performance that low in many of the same titles. Even now my FX8350 clocked at 4.8Ghz provides great performance for my buddies kid. In fact when I go over and watch him play I am always impressed with how the old system runs. Sad to see big tech channels dumping on your channel with out even watching the video, that really turns me off to those guys. Keep up the good work man.

[@sweaterboycutie]

I get it that GN and HWU are not interested in spending time revisiting and benchmarking outdated FX CPUs, but I think such dismissal of Your original video (they made an opinion on it before even watching some of it) is not a cool. They completely shut down any possibility of a discussion on the topic by claiming you are just wrong beacuse "you don't know what you are doing". Why is discussion on this topic immediately branded as "pointless"? Anytime someone mentions FX on the web there is an overwhelming barrage of shitty comments. Most of them are not backed by analysis or experience. The arrogance of GN and HWU just stands in the way of an open discussion on points made in original video because droves of their viewers just mindlessly parrot their tweets and comments. Not fair and not cool from both Steves.

[@kumbandit]

How do they continuously fail to understand your point? Or are they just drawing a strawman on purpose?

[@longmuskox4194]

Don't let those children bully you. They are wrong and their pathetic response to you reinforces this.

[@robertpintaric7354]

You are ok. They did something wrong, and they are to smug to apologise to community about it.

[@Zorro33313]

About core count: there is a fashion in community now to think "core count doesn't matter, only total performance does". What they prob miss is the fact total perf may be measured differently and not always translate to gaming perf linearly. Simple example: 4c i5 VS 8c FX, first have 4 alus and 4 fpus, second 8 alus and 4 fpus (1 fmac for each alu). let's say you measured both have equal perf of 1000 points in some imaginary synthetic shit and assumed they should perform equally in games. now you're playing the game and suddenly both are getting cache-miss or wrong prediction or smth that triggers the pipeline flush. for i5 the whole core (1alu+1fpu) is stalled now. for FX - only 1 alu and 1/2 fpu (1 fmac). This means i5 just lost 25% of it's total perf, and fx only lost 12,5%. You see where it's coming? We also may throw in the fact FX will most likely fill its pipeline faster due to faster clocks, tho i don't remember the exact pipeline length. So it turns out the test u used to measure the "total performance" were not triggering many pipeline flushes, while game was, which led to stutters on i5, cuz it was constantly losing a huge chunk of it's perf Surely different workloads may trigger pipeline flush different regularity and it's also heavy architecture dependent, but making any "total performance" idea pointless and certainly proves the fact having more cores *may* ensure you more stable frametimes. FX is a cluster physical multithreading arch. We shouldn't forget that. And it has benefits of SMT.

[@okaravan]

Shared FPUs can actually provide better performance than dedicated ones when the code being executed contains not only floating point operations. There is a lot of code that doesn't contain such instructions at all. When FPU is shared by two cores, it's throughput is two times bigger than that of dedicated FPU, because to fit it into the same silicon budget, we have to halve the number of transistors. So when one of the cores doesn't process FPU instructions, another one can process them twice faster than it would if FPUs were dedicated.

[@Zorro33313]

GN's reply looks like he didn't watch the video and was drunk on shrooms. It's just a 9000 symbols "look, i know fancy facts" rant that has nothing to do with the subject of the discussion. Such a shameful demagoguery, i'm highly disappointed in guys' inability to admit mistakes.

[@sandikata]

For me it's clear both Hardware Unboxed and Gamers Nexus are AMD haters since a long time. Their benchmarks are manipulated too much to show how good intel is (But it's a false positive, i've compared i7 2600k with my old FX8320 and i saw no difference at all and no any advantage of i7 2600k over FX8320, with same RAM modules same Graphics Card same PSU and etc).

[@TheImpartialTruth55]

7:35 he does not like this, because it directly exposes their lies (most likely, in those days, they well paid by Intel, who was then in "good shape").

[@rhekman]

Sigh. I'm sorry you had to make this video. Even when trying to be logical and scientific, emotions can derail a discussion, and people feel attacked. In the case of GN and HUB, they obviously put a ton of work into rigorous testing and I still trust their analysis of most tech. However I think their attention to detail can be a double edged sword. They spend so much time in the weeds, if someone comes along with a data set that doesn't fit with their conclusions, they can misread or misunderstand one or two things and suddenly they dismiss you, "AHA see this one thing you didn't do exactly like I did, your are totally wrong!"

[@jovianlitany]

I can not agree more. This kind of reminds me of back in the day, when Scott Wasson from Techreport first started talking about frame time. People flipped, called him wrong.... but here we are. I feel like some of these guys fall for the marketing just as much as anyone else. This is a bad look for HUB and GN. The difference between what you tested in games and what Steve "said" he tested is a difference between what I know as sensitivity and specificity. You tested by playing games, these were not canned, repeatable sequences and that is okay. We benchmark to game, not to repeat a benchmark perfectly, and that makes your methodology more valid, IMO.

[@glenwaldrop8166]

Let's face it, I like GN and Hardware Unboxed but in this case they're wrong. I've been arguing this point the whole time. They are biased against the FX so when they get crappy results they readily accept them. It's kinda like Anandtech GPU benches, they're good info but not directly comparable to real world benchmarks in that you'd never run a GTX 970 @ 4K Extreme settings on Wolfenstein. I understand having a baseline but it's not a 4K card. Steve is arguing in bad faith.

[@wolfwalker78]

I'm not going to argue with Steve or any of those guys but I owned, and was happy with, an 8350 and a 9590, and in a 1080p 60fps world that existed back then, they were just fine. I had an 8350 and a 4790K at the same time with various 280's, 290x, crossfire, not, etc, etc. The 4790K was faster but it wasn't meaningfully so at the time, for 1080p 60fps gaming. I've been doing this, as a hobby, since I put a dx2 in an old AT case and have as good a play-resume full of SMP and opterons and pentium pros and xeons and taping pins etc etc etc as the next guy. And while I understand the technical complaints against that old AMD chip, I never saw the practical problems with them for even a "power" user, for the money. That said, if you were going to call those big fellas out like that, you really should have your ducks completely, tightly, rigidly, in a row. No disabling cores, displaying results modeled after how they do, etc, etc. Leave not a blemish to be picked at. I appreciate, and follow you on here, because you carry the torch still for these old chips and I look back on them quite fondly and still think they got a bum rap(for actual use). But damned if I'd argue technicalities with what are basically professionals unless I had my shit wired exquisitely tight. That, and people have always been really oddly emotional, even if it's hiding under hard tech, about these chips. They were that way when they were new, and they still are. I'm amazed you don't have a mob of people with pitchforks outside your door as it is. I'm cpu and gpu agnostic, whatever works best for the money, or is the most interesting, or fun, is what I run at any given time. The FX chips, and the six cores prior, were a good time. Keep at it.

[@Russell970]

The dislikes are coming from Hardware Unboxed fans

[@xthelord1668]

they both clearly do not have any idea how to actually make a benchmark,instead they are just 2 big channels being popular GN is literally comming from paper news segment which tells you how bad he tests things or rewiews when buildzoid has to rewiew stuff for him his new cringiest test is contact between CPU and cooler which is ovbiously bad because CPUs are not flat at all meaning cooler manufactuers get slack if 3950x shows same shit for all coolers which would be legit if guy lapped CPUs to have a good baseline then HUB who has no idea on how to overclock systems properly at all and recommends 1440p to everyone even though it is useless res since who has a GPU which can drive 1440p let alone money to buy one their cringe moments are ones where they do not recommend any better kit then 3600 CL16 which screams waste of money E-die since 3200 CL14 exists at that price range while being B-die and where they do not recommend 5800x and 3800x because those 2 have bad value,like dude both cpus have best value and performance of whole stack when games and OS will suck up easily 6 cores and 12 threads which is great for 8 core counterparts because game gets to run entirely on 4 cores without jumping CCX and killing FPS while 12 and 16 core counterparts are not meant at all for gaming more for workstation but with better clockspeeds both of them fuck up benchmarks and tests left and right so they deserve to get some slack on dumb comments because: GN if you are going to actually read this without saying dumb points and claming dumb words: -please learn to lap things and re-test all possible cooler combinations to show why lapping IHS is so important since you probably have spare CPUs to show stock look of things and lapped to show modified look of things to determine does cooling side have garbage surface area or not and how bad is it actually -learn to actually do a proper hardware rewiew without using buildzoid as a guide on everything because at that point you can let buildzoid own your channel since guy did a shit ton of rewiews for you HUB if you are going to actually read this with same terms GN got: -learn what RAM revisions exist and why sometimes it is better to go with lower speed but tighter timing kit instead of faster speed but looser timing especially in similar price ranges -learn why 1440p and 2160p (not 4k which is term used by companies to make their product better when its just 2160p) are not used at all and why pepole prefer 1080p (without saying 1080p is not good if you play single player titles because that is a insult to everyone who cannot afford new or strong hardware when we live in a world where soemone will pay a hefty price for "cheap hardware" and both of channels please improve on real-time testing,that is more interesting then stupid bars and frametime charts which take more time then simple 5 runs and record the 5th one for best consistency or record them all because again you set yourself a goal where we actually get to see what is going on instead of plain bars and numbers which is what compaines do to attract audience

[@MP-ul]

It is 4:52 am i read the comments and yours ma n is so damn on the point. I have the same problems with the gn and hub, I follow those 2 channels for years and I see in every new video that they start to do it poorly, chosing games, skipping tests and so on.But is not nice to come down with huge hate towards this lad just for testing an FX processor and disabling 4 cores on a Ryzen 1600.

[@CharcharoExplorer]

I will disagree with you here on a few things. Contact between the cooler and CPU matters. Obviously a CPU is not perfectly flat, but neither is a cooler and mounting pressure exists for a reason here. 1080p is a popular resolution for laptops and lower-end gamers, true. However, 1440p and 4K arent expensive. I am from Bulgaria, a country where people make 300-400 EUR a month BEFORE taxes, and even to me a 1440p entry monitor isnt sounding too expensive. For high-end gamers with GPUs at the 3060 Ti level or above, it makes little sense to be on 1080p. Its a waste of their GPU's power.

[@yottaXT]

@@CharcharoExplorer i agree with you on the GPU-monitor part, @Gabrijel Toman just forget that GN and HUB create content for a *very* wide audience, the situation you live in your country is not the same everywhere, so making an average comparison is the better you can do to deliver to all your audience.

[@dominikkavur5373]

Also, you should really make a video using the EXACT parts that they used. I think that would make both Gamer nexus and Hardware unboxed redo their benchmarks. P.S: Love that you have the same shit as in the other video :D

[@RATechYT]

Unfortunately I don't have the luxury to buy the exact same parts they used, but I'll be using a G4560/ i3-6100 instead of the 1600 AF in one of my upcoming videos.

[@kanefriessner]

@@RATechYT They only have money because they are sponsored like hell. And this is why their benchmarks are flawed. No company wants you to think that your old FX is still good, no need to upgrade yet. They are paying for these benchmarks.

[@kingjasko]

@@kanefriessner maybe, but imo they just made a mistake and had too big of an ego to admit it, i don't think we have to come to conclusions like that

[@DeadlyRedRing]

@@RATechYT Comparing to the G4560 would make much more sense. GN's results show the 3000G much slower than the Pentium, and HUB didn't even compare FX against the 3000G. 3000G is a decent chunk slower than dual core Intel

[@misterpinkandyellow74]

@@kanefriessner I think they used old out of date benchmarks, because they are lazy

[@dominikkavur5373]

Great video, as always! Keep it up man :D

[@S己G]

8:47 Right there: The fx 8350 is OODLES better at gaming than my i5 2540m. (Even though my chip runs at 3.1 ghz most of the time, single thread performance is roughly equal if I recall.) The simulated 2 core 4 thread processor is also faster than my i5, and my i5 is basically an i3 2100 has a lower base clock, but can also boost to 3.3 ghz, (which means that i3 2100 benchmarks SHOULD be roughly applicable here.)

[@protienslv]

Recently i bought a pc for just 25€ with FX8320, 16GBRAM, M5.A97R2.0 ASUS Mob. It was my first AMD experience. After being testing it in monero mining I've been shocked how capable it is in 2022. Just look on the hashtate from my miners: I5 8400 3000H/s FX8320 2200H/s i7 3740QM 1600H/S I5 7200GM 950H/s Snapdragon 860 500h/s

[@MaxiMuM1441]

someone send this to HWU and GN

[@fuzzyface4515]

I missed out on all the drama but don't let your dreams be memes or something like that. Your videos are some of the most straight forward, honest and understandable. You go in much more detail than most, and showing the real time side-by-side results is one of my favorite things about you. Sure, others do copious amounts of benchmarks and make really big graphs but they don't give as much of a "feel" for that performance and plain graphs can be misleading without any real context like your videos often provide. Unfortunately for you, most people have the attention span of a goldfish which led to so much confusion and misinterpretation about your original video. I hope this hasn't been too much on you and doesn't bring you down because I'm always looking forward to your content.

[@RATechYT]

Thank you very much for your comment.

[@christiansonsen]

Great! congratulations. I'm an old suscriptor and I've been watching your videos for a while now, even before Gamer Nexus and Hardware Unboxed. Greetings from Argentina!!!.

[@sanchit.samuel]

Please keep delivering consistent investigations, like the FX delivering consistent frame times. :)

[@johnringgold3036]

There is no profit from promoting old "out dated" tech. Some of us cant afford the latest and greatest. Your videos help the little guy get into pc gaming/building on a budget with good price/performance. Keep up the work.

[@AdiGInDaHouse123]

Those are some decent points you are making. Myself, I've been using the FX 8350 for quite a while and haven't had nearly similar performance to what could be seen in HU and GN videos. Of course, it doesn't necessarily have to imply that their issues with those CPU's were not legitimate in any case. It is also hard to disagree that reaching the level of professionalism both of those channels have is reached by hard work and dedication. Nevertheless, we are still human and even the best can be wrong on something. There is no need to get emotional over it. Keep up the good work!

[@RATechYT]

Exactly. Many thanks!

[@PhaedrusAK]

Thanks to your overclocking guide, I'm still using an 8350 with no need to upgrade, which makes stopping watching GN and HU easy. Your video was interesting, it certainly didn't deserve innaccurate criticism, and keep up the good work. By the way, this made me realise that your presentation skills in English are better than those two channels ;)

[@BlackestCRT]

its just content, people should simply enjoy, if you dont like it, dont watch it. simple. i like FX videos but if u dont like it simply dont watch it.

[@KavorkaDesigns]

OC'd 2070 Super K.O with the FX 8350 4.9GHz Stable (Noctua U12A + a iPPC rear 3K) I wasn't being bottlenecked by the cpu, perfect match imo

[@saifullahnasir1984]

To be honest, even if your videos were considered "misleading", it's not gonna force anybody to go out and buy FX CPUs. All you're trying to do is prove a point that these FX CPUs aren't as bad as everyone thought they were, and I believe you have effectively done that in your videos. Personally however, I still wouldn't recommend getting an FX CPU for a used budget oriented gaming PC nowadays, unless you got an amazing deal of course.

[@cdiff07]

I upgraded from a 6 core AMD 1090T to a 8-Core AMD FX 8350 in 2011. I literally went back to the 1090T because of IPC etc. was better on the 1090T. It literally felt like a downgrade in gaming in every way!

[@Pillokun]

Actually the replies from GN and HU is pretty much expected, first they dont game so they dont really know the hw intimately at all, all they do is run a short 30sec to 1min test bench three times per hw and then move to the other game/hw component. I still remember when Steve from GN did not know how to put paste on a cpu correctly. And when you give them a tip on something they dismiss us even though we have been hw enthusiasts for longer then they have. They do steer their bases in the directive they chooses, and I still remember the big fight HU had with Jim from AdoredTV and his fans, that is why I feel HU never ever releases anything that can go against the big AMD crowd/fans. They are pretty toxic and even though I think you should do the testing with faster gpus because the perf can regress with a very week cpu I expected that they would try to attack you like this.

[@rodrigomendes3327]

You're right, they're wrong. Everyone that have any aknowledge should know that and its a big surprise to read what HU had wrote. It's just absurd. A Ryzen 1600 using 2/4 cores would have a very similar results than a 3000G, since the achiteture is very similar. No the same, since it's a Zen x Zen2, but surely very close. The claim about i3 better than Ryzen 1600 2/4 is just absurd. IWhen I move from 2500k to Ryzen 1600, I have tested 4/4 on Ryzen 1600 and it was at least 20% faster than Sandy Bridge on modern games like Control, Far Cry, Hitman and many others! You don't need to know much to know that, since there's a lot of real tests between 2200G x 2500k and 2200G, even being a worse CPU than Ryzen 1200, faster than 2500k@stock. So.... Really, I'm impressed how these gyus have no idea what they're talking about. Specially about Nexus, I have 100% sure that he use educated deduction to show results from older CPU and never tested them on modern games. That's why they're leading to mistakes like that. Probably they have no idea that FX is better than Core i5 Sandy Bridge or Ivy Bridge in basically every modern game today, since in the past FX was way worse. They just grab that data and use randomly. Lot of channels do things like that and it's a very similar methodology that bottleneck calculator have. It's a very poor one.

[@GregoryShtevensh]

I think it's quite simple... Your test gave the advantage to the 2 core, and yet still showed it lose to the 8 core FX. Their argument is that it didn't give the advantage, yet they support that with points stating why it has a superior architecture, more cache, and better thermals. I really don't get how the Steve's missed this mark

[@hellbound83]

Another excellent video! HU and GN need to learn how to take an ounce of criticism. The sad thing is I've seen this before from HU in the spat they have with Tom's Hardware.

[@DarkAbyss78]

Still using FX-6300 overclocked @4.6Ghz smoothly. give some love to AMD haters and keep up the good work bro.

[@reki353]

rip, i would have kept mine but on stock speeds the gigabyte 990fx motherboard i had keeps having its vrms overheat and single core performance was just wayy to low

[@DarkAbyss78]

@@reki353 im using an Asrock 990fx extreme3 and no issues with vrms overheat. recently i bought an FX8350 for 40euros and replace 6300. im in a hunt now to find one more am3+ board to put the 6300 back.

[@megaoyi1Wins]

lol Hardware Unboxed actually talking about being misleading, while being both gamers nexus and HU the MOST BIASED tech channels there are!, Keep it up man, don't let hypocrits let you down.

[@guyharris8007]

Hubris Steve and Steve..... Not cool to bash a smaller channel for calling your results out. If you (GN and HU) disagree calmly explain and debunk Ra Tech's testing. Bashing with explanations is childish and shall be paid no attention by RA Tech.

[@jotabe1984]

Man i used to follow both HU and consider his content really usefull... thanks to the share ammount of data he was able to put up and collect, while he also expressed al of his lab and overconfidence about their method. But then i faced their test with my real-life experience and for real my gaming and overall user experience with Fx CPU wasn't nearly as bad, while from my large experience using dual core intels like i3, i did know the low-core-count limits, that weren't never really shown in the Steve's test... He didn't simulated things like real-life multitasking... it is very normal that an average gamer depending on game type will have a navigator window open with a couple of tabs just to work as a combined walkthrough, and in FPS and other competitive games, maybe Macros (i'm not judging, just describing) and communication software might also be used. And that's not to mention all the "bloatware" that many people gets with time, that diminishes CPU avaiable room. i'm pretty shure that a high core count cpu manages these extra stuffs quite better than a higher speed low core count... in fact... i test that every single day... in my office i use an i3 4170 that sometimes struggles with heavy ofimatic software (i have to say that i never really reach 100% ram usage, not even 80%) while in home with my Fx8350 i have never faced any issue like that. With my Fx8350 i can (by managing core affinity) even render photos w/6 cores and leave the 2 remaining ones for the average usage (youtube and stuffs) and get good results despite being rendering with a 10 year old CPU. by having an Fx8350, i can split performance into 2 "emulated" Fx4300, or a single Fx6300 along with a Phenom 2 x2. That might seem like very few performance to begin with and the "performance splitted" results seems like underperformer CPUs... but again, we are talking about using 6 cores to take on a large task that might use 100% CPU if cores are not managed, while leaving a free headroom to navigate the web, or use an emulator to spend time. Or maybe render 2 things at the same time with a modest performance, but being able to perform both tasks... Pretty nice, huh? Well now imagine that but with a Ryzen 7 1700x, now you can split performance into an R5 1600 + Athlon G, or maybe 2 Ryzen 5 1400... That's a pretty solid baseline to divide cores and get good performance.... Now imagine that with a Ryzen 5 5700g... or with a Ryzen 7 7700... or why not a Ryzen 9? (same goes with intels CPU of course, but older gen 4core 8thread i7 could only be splitted into 2 i3, so at the end of the day having extra cores is better than having less cores with more threads)

[@DemonSaine]

i know i'm late af, but did either of them ever apologize for the misunderstanding?

[@RATechYT]

Nope.

[@petrichor1956]

@@RATechYT so besides the pinned comment, nothing arose from this? You got nothing? Not even a corporate apology?

[@RATechYT]

@@petrichor1956 Yup, there's nothing other than Walton's deleted KZbin comments & his tweet

[@petrichor1956]

@@RATechYT that sucks man. I genuinely enjoy your content and wish you the best. Hope you succeed on the platform.

[@jpeg3933]

this guy deadasss said an FX8350 is worse than an i3 WTFFF

[@reki353]

which i3 you talking about

[@zynix27]

@@reki353 THE i3.

[@Loppy2u]

Wish there was this kind of compare, contrast and debate during the pandemic. Instead of no questioning allowed.

[@O.Shawabkeh]

*Plandemic

[@h1tzzYT]

So i watched this video and read quite a bit of comments bellow, honestly, reading comments which insults HWU or GN feels wrong. On the other hand im happy to see new users who watched this video and previous one and went into this open minded and actually enjoyed the content. All i want is nice discussion from all sides but what i see is two youtubers which i watch daily, come with some passive aggressive attitude towards smaller channel and without properly analyzing video in question leaving snarky comments on twitter. So anyways, im looking forward for your next video RA Tech, i hope you wont be intimidated by big yt channels and keep creating your videos which i enjoy watching!

[@RATechYT]

Thank you for your support!

[@luizarthurbrito]

You just gained another subscriber two years later for that video! Cheers

[@Hardin9745]

Your AMD FX videos are very entertaining and because of your love for the processor you have gone to great lengths to defend its honor. YOU ARE CORRECT that that the AMD processor delivers a superior smooth experience even if and when it may fail to significantly best the frame rate of a lesser threaded, higher IPC cpu. (and I realize say it often does) There is absolutely nothing wrong with your methodology either. I have owned and benched cpus and built PCs for decades and read about the architectures extensively for years and while I don't claim to be the experts that the "Steves" claim to be, I can also say with absolute certainty that neither are they. Great respect for you for standing up for yourself and defending absolutely fine videos. I also have great love for both Steves and am a big fan of their respective channels, but really their impertinent and unnecessary replies diminish their stature.

[@chuizune]

That comment from GN might be one of the lamest, cringiest, butthurt, holier-than-thou things I’ve ever read. I’d known for a couple years now that Steve Burke had an ego bigger than Jupiter but holy shit, that was hard to read. Keep up the good work. I’ve enjoyed your content since discovering your OC guide for the fx8350 back in 2019. You have a very down to earth style and bring an interesting take on the topics you cover. Best of luck.

[@RATechYT]

Many thanks!

[@herbertvonsauerkrautunterh2513]

I still have 8300, 8350, 870k, various athlon x4s.. even an opteron 165 still running and also various ryzens up to 3700x. Absolutely nothing wrong with FX CPUs .. is just snobbery. I just don't touch Intel. Usually only in laptops.. Don't worry about the others.. maybe they are doing something wrong

[@char-mon4127]

My friend I have a question What is the difference in performance between the 8300 and the 8350? I have a fx 8300 and a friend offer me a cheap 8350, is it worth spending money on such an upgrade?

[@herbertvonsauerkrautunterh2513]

@@char-mon4127 depends on how cheap. It is a step up over the 8300.. I wouldn't pay too much though

[@chrisz2116]

I sold an old gaming rig I had a month ago and it had a 8350 and a rtx 3060 and when it comes to gaming it had absolutely no issues playing games like fh5 insurgency sandstorm destiny 2 and dying light 2 I got the same performance as I did with a 2700x so it was not a bottleneck when it comes to 1080p gaming

[@Babbages]

i bet on their tests that their vrm was overheating and throttling the cpu that's why on every fx cpu i had I always added a extra cpu fan to cool it and this was between like 3 different motherboards .

[@GiantsGraveGaming]

Buddy what are you apologising for???? they even haven't read nor watched what you tried to explain, being kind to this kind of bullying is not gonna help them being humble.

[@stepankyjovsky4180]

Maybe HWU will reconsider should your editorial direction change...

[@RATechYT]

That's a good one.

[@Chrono348]

David vs. Goliath on two fronts, someone get this man another sling!

[@AmbiguityCaptive]

I've had 8350 and do know how to o.c. it, also i know that 970-chipset runs hot and may throttle, turning your 4.5 fx cpu to 2 ghz one untill temps normalize. Not sure that GN or HW Unboxed cares about these little things

[@AlexSBedini]

RA Tech well done for standing up for the truth no matter who you are facing on the other side !!!! GN and HU might be a great value for the gaming community but that does not make them immune to mistakes!!!!

[@daetitus8140]

This is why I don't subscribe to Gamers Nexus and Hardware Unboxed. They are just too biased.

[@robertmyers6488]

You blew up their methodology. The responses are no sequiturs that do not address what the claim was.

[@RATechYT]

Yup. Unfortunately all we got from them was: "We have more experience, we know better. Your arguments don't matter."

[@robertmyers6488]

​@@RATechYT Hard to claim you have more experience as a justification when you advocate that a 12 GB card that can't do 4K without fake frames and stuttering is better than a true 4K card with 20 GBs and then refuse to review anything other than the reference model. Oh but it was out of the box.

[@666Necropsy]

i do enjoy those other channels. they have been really helpful. this is really a disappointing response from them. i wont trust there channels as much in the future from there response. if they cant investigate the claims from @RA tech properly what else are they missing. nobody wants to be mislead. who is being misleading. very nice work @RA tech.

[@chinesepopsongs00]

Ok the FX was not a good product, but i think you are right in your findings. I don't have any of those old ones around anymore, as i only held on to my old intel stuff as it just aged better then the AMD. Partly the response of more cores is not more fluid is also correct but as big channels hardly test the low end stuff properly anyway they did oversee something and that is that if you have that little cpu power it starts to count. So in general if you compare many cores to many cores (i do not give exact numbers because determine where the tiping point is is pretty hard) then faster cores start to count more then just the number of cores. But if there is no cores to spread the load to and you are already running on your toe's then extra cores how weak they may be start to help. Thats one of the whole things when you design big/little. As adding big cores is to expensive and would mean lower clocks for all big cores then you add little cores. People tend to think it is just for enegy savings but thats only part of the story. Also hypertreading (or anything that does the same but has a different name) gives roughly 25% performance gain in a different thread. You can compare it as a extra core that has a very inconsistent performance between 20 and 30% of a actual core when the total CPU is loaded above 70%. And that is exactly what is happening with the dual cores. The load in games is always (every milisecond the game runs) higher then 70%. Hypertreading is not magic and being every millisecond depended on it causes what you measured. There are only 2 shared cores so moving things around to the least loaded physical core is hardly a thing anymore there is no scheduler in the world that can help you. Back in the days from about 2nd gen intel until about 6th gen there where people who had a i7 but disabled hyperthreading for smoother gameplay. I don't say they where always right but the people who did where also not always wrong either. Short version you are right they are wrong when they try to apply their general rules that only apply to higher core counts and not constant loaded processors to the point that you depend on hyperthreading. Also i am a little bit dissapointed in their knowledge due to their statements. Calculating (predicting) workloads is a small part of my job, needed to scale stuff right in datacenters. Then you need to know about cores hypertreading the bus structure in the cpu between the cores and the memory/io to be able to to say something about what you can expect or how to scale vm's to get the most out of your spend money.

[@kylekennedy3299]

Classy response, they probably wanted you to go cry in a corner. Some people like pages of comparison graphs, some people like watching a workload.