@DrPizza @AURonline well, to be fair, I don't read articles unless they are 60fps at 99.99999th percentile.

@DrPizza @AURonline another data point. Look at these HUGE CPU DIFFERENCES in Hitman! 99fps vs 93fps for $180 cpu. https://t.co/Zpue2eoChV

@DrPizza @AURonline but since you can get a quad core kaby i5 for $200.. not sure why we're arguing here. Even though I am right ;)

@DrPizza @AURonline again you are cherry picking certain games. And I suspect a 2 core / 4 thread kaby *would* even handle those.

@DrPizza @AURonline that said I was very very happy to see Ryzen catch up to Broadwell in single threaded perf, and more cheap fast cores

@DrPizza @AURonline and this is my larger point: we have *OBSCENE* amounts of cpu power! That is NOT the problem in gaming! Not remotely!

@DrPizza @AURonline whyyyy are we even arguing.. $240 for a quad core 4.2 Ghz Kaby would fucking DESTROY hitman. https://t.co/vq7ZsFzNrt

@DrPizza @AURonline highly unlikely. But your opinions are, as always, fascinating. Too bad the data disagrees.

@DrPizza @AURonline 2 things: that's haswell (ancient) vs kaby, and 2 core / 4 thread. So yeah, I'd bet it is damn close to 60fps.

@DrPizza @AURonline We'd play at 1440p, high/ultra. Let's look at the numbers, shall we? https://t.co/KoYIKkLiRE

@DrPizza bear in mind that 5960x is *Haswell*. Not even Broadwell, Skylake, Kaby. So yeah 2 core 4 thread Kaby at 3.0 Ghz? Bring it on.

@DrPizza @AURonline I'm saying 60fps is a bullshit metric in Civilization. and even in Hitman, really. Along with 99th percentile.

@DrPizza @AURonline horseshit. In your cherry picked list of games where locked 60fps framerate is a non issue. Civilization? Seriously?

@DrPizza @AURonline also notice the 4c/4t results on 1080p medium vs 4c/8t. That's every bit of 15% faster. Proof.

@DrPizza @AURonline that said, I can agree that 4 *real* cores is definitely nice to have, and if ryzen ups the ante, more power to them.

@DrPizza @AURonline I'm absolutely certain of it. Modern hyperthreading is quite good, Intel has been at this a while now.

@DrPizza @AURonline note that 2 core / 4 thread was not shown, and I consider that a sweet spot. 2 core / 2 thread is .. very low end

@DrPizza @AURonline Enjoy your 144fps constant in Civ 6. Meanwhile, I'll be providing actually *useful* advice to others on this topic.

@DrPizza @AURonline I concede that Civ is a game about CPU, sort of, but why do you need 120fps constant in it again?

@DrPizza @AURonline or are we still talking about 99.99999th percentile frametimes?

@DrPizza @AURonline Fascinating opinion, but the data disagrees https://t.co/7ua0sLW5w5

@AURonline @DrPizza wait sorry if we consider the 99.999999th percentile frametimes

@AURonline @DrPizza gaming is a super dumb thing to harp on when the most mediocre Core i5 is already plenty good enough

@danluu I have seen Amazon fail to display order history a few times lately .. definitely off

@DrPizza which went away with the 1080 Ti. But at no point did I say "that lack of PURE COCAINE 60FPS is the reason I died."

@DrPizza well, I sometimes had the "frame rate warning" on the 1080 at 4k with BF1 and these settings https://t.co/tchLuQ0fnB

@DrPizza I am sure some games have their own quirks, but "144fps constant or your e-peen will shrink" is dumb as hell

@DrPizza I have played hundreds upon hundreds of hours of battlefield 64 player multiplayer and that is absolutely not true.

@DrPizza a few hitches during extreme battle carnage moments is expected and no big deal. 95th percentile I can maybe get behind. 99? Nope.

@DrPizza also 100x more expensive, for dumb reasons. Maybe if your goal is 144hz constant

@DrPizza ridiculous, nonsensical metric.

@DrPizza @arstechnica you really don't know anything about this, do you? Read my prev links. Bf4 multiplayer, core i3, i5, i7 nearly same

@DrPizza @arstechnica agreed, dual core is plenty for virtually all games

@DrPizza I hate to break this to you but the bottleneck in 4K is not the cpu

@DrPizza https://t.co/xJfzqyHuNX

@DrPizza https://t.co/jxXhG2URaV

@DrPizza @arstechnica I see what you're saying, though, there need to be midrange Ryzens. If you like moar cores, then Ryzen.

@DrPizza @arstechnica core i5-7500 is more than fine at $205. Spend your gaming $ on the GPU. Utterly and completely wasted on CPU.

@DrPizza @arstechnica well 4k is a whole nother tax bracket my friend. I mean for people doing avg bang for buck gaming

@tobermatt https://t.co/d37BakfDT8

@chowspecial THANK YOU I was trying to figure out where to go with that 👏

@DrPizza @arstechnica I recommend gamers buy the very cheapest Core i5 from current gen, and spend $500+ on GPU. So yep. CPU ain't no thang

@JasonPunyon not until one of us is blind, Punyon

@domenic @capajj I'm with you on that, bezels in the middle sucks

@DrPizza @arstechnica games are 99.9999% GPU. Unless you're, like, playing at 720p.

@textfiles new orleans apple // scene? bizarre

@arstechnica @DrPizza "slower at games" is literally the dumbest possible take of Ryzen, and unbecoming of Ars. IMO.

@Hooberbloob Battlefield 1

@dubya_brian I definitely know 60fps is the minimum but I am not convinced more than that is really needed.

@dehora @rbranson also reliable persistent storage is, surprisingly, hard as HELL. So the value of S3 is huge. Much greater than EC2 IMO

@dehora @rbranson at least S3 prices have declined dramatically, this is good. https://t.co/SKVameP8fF

@rbranson to be clear in general I agree with you but the situation is a bit more nuanced. Also profits on cloud are OBSCENE, and too high

@rbranson adding infinity cpu cores, strangely, does not make our Ruby code go faster. computer science looolll

@rbranson funny how single threaded performance works, isn't it? Let me introduce you to my FUCKIN' FRIEND RUBY

@rbranson I should probably blog this

@rbranson I have many thoughts on this matter

@rbranson also old cores. Granted CPU innovation isn't exactly on FIYAH these days, but Ivy Bridge vs. Kaby Lake? That's +30-40% right there

@rbranson Guess what giant cloud services tend to _not_ prioritize? They have lots of cores at thoroughly mediocre clock rates..

@rbranson with Ruby we are absolutely at the mercy of single threaded perf, and clockspeed REALLY REALLY matters.

@rbranson here's a very specific example, *literally* 2x real world perf for trivial one-time investment https://t.co/ylH5DqENq8

@rbranson it definitely gets way less defensible as time goes on, but there is one exception: max performance, if *really* needed

@balpha seems correct to me

@lizardbill whyyyyyyy

@danluu yeah my crappy responsive youtube video JS hack caused this, unfortunately. Was 83s to "start render" on 2G, now 4.3s

@danluu I see, my main problem is render blocking JS (jQuery), let me try to fix

@danluu oh my bad. I checked and you're right, it is accurate. Confirmed with Chrome Network, "regular 2g"

@danluu visual complete, even things below the fold? why would that matter? If blog posts 2,3,4 render offscreen?

@danluu this comparison is still kind of inaccurate since it doesn't reflect the visual time to load. Full time to load images below fold?

@Jason in the year of Linux on the desktop maybe

@voyageur_techno @theatgrex it is like considering footnotes the primary feature of Wikipedia

@cwodtke I don't think they appreciate how brutal of an industry it is. It will suck all the enjoyment out of games, forever

@lizardbill I get nervous if there aren't at least two computers within arms reach at any given time

@cwodtke do we? We're playing the world's biggest MMORPG right this second.

@lizardbill *never* have only one computer, that is a rookie mistake Bill

@cwodtke yeah the code is hard enough, UI is hard, then optimize for "fun" .. good luck. I always advised people to avoid the field.

@cwodtke I have long felt of all the programming fields gaming is the worst and hardest on humans. Since 1996 at least.

@HezelAnna @knguyen YES! COLA IS WINNING. That's how I know in my heart this is science

@joelmartinez @AdamWeeks you should! It's quite good!

@balsama it is an open question

@mmorri you would need to ask WASD

@johnolilly it is quite good but I can't bring myself to RT anything on Facebook proper. That nag dialog.

@johnolilly likely an intentional leak of a "good" year

@johnolilly is it just one year? I haven't had time to look

@pauldegnan OH SNAP SON NOW IT IS ON

@ID_AA_Carmack Galaga on a NES?!? sacrelige

@kevindente after trying it, we've decided threads are a complete disaster

@Gankro 🤗

@pete_otaqui press ctrl+f twice, not all replies are in the DOM at any given time, only the current view.

@Gankro not that it can't be improved, constructive suggestions always welcome!

@Goodwinmf well yes that is the point of https://t.co/0gf3j4UQkg

@Gankro https://t.co/RiNOq8lSyR

@railsonmaui I can barely fit the three 27" ones I have.. would need a far bigger desk

@wilshipley well they were at a "real" job interview, so distracted / nervous, probably

@textfiles @romero looks like someone had a rough divorce

@scottwasson you are really committing to this vr thing, man

@jarrod_dixon @Nick_Craver @SuperDalgas @sklivvz @jpobst Internet Dirver's License, baby! OpenID's problem was lack of email = identity

@pamelafox welcome back! I was just thinking about you the other day.

@MuslimIQ I want James A. Bretney to go home to europe.

@df07 good one @Nick_Craver lucky for you "the emperor is not as forgiving as I am"

@shanselman @MarkusKrogh there was no reason for me to be repeatedly mentioned. Learn to use the delete key, folks.