Ryzen Gaming explnation

I can barely understand this guy. It's a chore listening to him. Scottish bastard

>>59269537
No way he's Scottish. Sounds like a Pajeet

>>59269537
Really? I'm an Ameritard and I understand him just fine.

>>59269514
>adoretard
endlesstrash.gif

>>59269588
lol u serious?

>>59269514

Yes.

>ryzen is great for gaming when games have gpu bottleneck
>ryzen is great because it might be better than Intels worst cpu's in 5-6 years
>Joker's video with underclocked 7700k

He used to make nice videos, now they are Chris Chan cringe tier stuff

>>59269514
So this person is saying that and 1800x will be as good as or better than current intel chips in 6 years for gayman? Or am I missing something?

Yeah, I'll just wait onther 6 years, then I can show all you retards that it was a good idea for games.

>>59269654
His videos were always shit but you can't expect unbiased opinion from extreme amdfantard.

>>59269662
His point is that running the game tests at 1080p to show future longevity is stupid.

You bench at whatever people play at today and that's it.

If you really want to see the cpu bottleneck, check out the fact that Intel's all are running at basically 100%. What headroom is left?

god damn that scottish accent is so thick it's hard to understand him as an amerilard

>>59269514
>is he right?

Yes.

>>59269866
The key to understanding foreign accents is to listen to them more.

>>59269866
get the burger out of your mouth fatass

>>59269830
In the future when gpus are powerful enough ryzen will shit the bed compared to 7700k but in 4k.

>>59269514
He has a point. People who says he is a AMDfag, just watch his video shitting on Vega, and there is one of him shitting on Polaris.

>>59269942
>sandy E
>xeons at 2.2Ghz
>heck even bulldozer

all work fine at 4K now.

>>59269654
You missed the point completely.

No one's claiming Ryzen is better than Intel in 1080p/1440p/4K gaming.

The diehards from Intel are claiming Intel is much better at Ryzen in 480p. This is supposed to show Intel's power over Ryzen.

The diehards at AMD are counter-claiming the difference is much smaller at 1080p/1440p/4K given that those are real scenarios.

The scottish guy is claiming the GPU bottleneck test with 480p low settings are not real world scenarios or future world scenarios as it simply distorts reality.

>>59270042
meanwhile i5/i3/i7 of same time period suck at 4k

>be me
>6600k
>water cooled
>never oc'ed
>idles at 40c

wtf

>>59270114
your investment is reasonable.

a stock cooler from intel would have done the same thing though

>>59269678
>you can't expect unbiased opinion from extreme amdfantard
He literally posted a video saying Vega is probably going to be total shit

>>59270073
How are they being "intel diehards" if they want to see tests that actually measure how cpu performs to see how good a cpu is?

>The scottish guy is claiming the GPU bottleneck test with 480p low settings are not real world scenarios or future world scenarios as it simply distorts reality.

well of course you cannot test for the future scenarios because you don't have applications and games that there will be in the future, this is why they try to simulate it by removing gpu bottleneck (this is assuming that user buys new gpu of course)

File: bulldozer_cores_not_important.png (164KB, 1261x709px) Image search: [Google]

164KB, 1261x709px

>>59270178

>>59269942
Testing a cpu's performance at 1080p gaming is equivalent to testing a car on a straight track.

No one drives that way, no one games that way.

>>59270178
it boils down to compromise

lower but smoother fps or higher and slightly hectic fps

if you want smooth but don't care about 144fps get ryzen
if you want 144fps but with drops get i7

>>59270178
Why not just look at cpu utilization percentage? That's a better indicator of future cpu performance.

>>59270178
The charts he uses shows that these simulation of "removing gpu bottlenecks" don't reflect reality. This indicates the testing methodology/reliance on low res benchmark is not working.

If you want to see pure CPU benchmarks, there's synthetics that show it truly. This completely removes the GPU bottleneck. Why are they not valid? Why is "by removing gpu bottleneck" by doing 480p even realistic over pure cpu benchmarks which actually removes GPU bottlenecks? Ofcourse pure CPU benchmarks have their own issues like intel optimizations or single threaded synthetics. There are ofcourse synthetics that focus on pure single-thread over multit-hreads. These all come down to developers and what they want to show in their work.

Just a meme.

>>59270201
It's not done at 1080p because people are supposed to game at that resolution. It's to ensure the CPU isn't waiting on the GPU, and is saturated, rather than twiddling its thumbs at 4K.
Dismissing 1080p benches because "I don't game at that resolution" is you misunderstanding. It's not the purpose of the test. And as the resolutions of tomorrow become the resolutions of today, you'll find your CPU is getting crushed sooner than the alternative which costed less at the time of purchase, despite all the benchmarks showing them neck-and-neck at what was high-res in 2017.

>>59270260
Synthetics aren't real-world gaming workloads, and they never have been anything but a rough approximation which can fall apart under real-world tests. Actual games never suffered that problem. Hence why, from time immemorial, synthetics have been taken with a grain of salt. A score on a low-res game is more representative of the performance you'll see IRL than a synthetic.

>>59269514
The guy is kind of correct.

The CPU is perfectly fine, but suffers from a symptomatic AMD problem - lots of resources but poor utilizatio. It will most likely increase somewhat in performance as bioses and windows support matures. What happens now is teething problems. Early adopters are going to be shafted as usual.

Moreover. The only route to more performance these days is through multithreading.

The gaming benchmarks are indicative of the thread scheduling problems. Games suffer from very uneaven loading of threads. Especially DX11 titles, because they off-load most of their work on the Core 0. If the scheduler fucks up and puts both heavies threads to Core 0 Smt 0/1 then the entire game will suffer.

This is not visible in rendering, cinebench or whatnot, because threading on those tasks is nigh symmetrical.

Also - /g/aymurs need to finally understand that HEDT chips perform worse in games then top clocked i7 series ones, and Ryzen is somewhere between HEDT and top desktop CPUs. For pure gaming 7700K is superior to just about anything as of yet. This will change as threading becomes utilized even more (especially with Vulkan/DX12 and console ports), but even in 5 years 7700k will do very, very well.

I intend to bamboozle my employer to get me a 1800x for my workstation once support matures, because I am a programmer and I compile a shitload.

>>59269866
I can understand him just fine. Are you sure that English is your native language?

>>59270461
Did you watch the video? He shows how these 1080p/low res benchmarks do NOT indicate future performance.

>>59270571
>1080p/low res benchmarks do NOT indicate future performance
They don't even reflect the reality today. Every new big game is being made with 8 threads in mind because both consoles use AMD 8 core processors and only way to make game engines run nicely on those systems is to optimize the games for 8 cores.

We don't even need to look at Ryzen to see this, just compare 6900K to 7700K and it's obvious that in games like BF1 on multiplayer with higher graphical settings that ought to bottleneck the GPU a four core processor starts to showing its weakness with occasional stuttering

no game benchmark I've seen to date has yet measured CPU usages during benchmark runs, which SHOULD be measured, because if Ryzen's low/average/high FPS is lower than Intel's but the cores aren't fully used, then it's not a CPU bottleneck at all.

95%-100% CPU usage is a sign of CPU bottleneck in games, anything less and it's a GPU or Software(game) issue.

>>59270806

There is a snapshot of BF1 in the video that does, in fact, show the 7700k pegged to 90% across all 8 threads at that given time. In the same scene the ryzen chip is sitting at around 35% per thread.

>>59270461
If synthetics don't show the power of cpu, then how the fuck is low-res games supposed to show it? How is it more representative?

At best, both benchmarks would be out. However synthetics like 7z/Photoshop/etc have real world applications. Meanwhile 480p low res shits have no application.

So your whole argument about low-res being more representative falls apart because the synethetics actually have real-world applications.

>>59270880

>synethetics actually have real-world applications.

*when actually affordable Intel chips are winning.

Its why the review done by Serve The Home is so educational and given the results they got I can see why they are salivating over ryzen.

File: intelfags.jpg (282KB, 752x548px) Image search: [Google]

282KB, 752x548px

>>59269514

File: AMD-Ryzen-7-1700-Sysbench-Single-CPU-Benchmark.jpg (77KB, 858x490px) Image search: [Google]

77KB, 858x490px

>>59270907

https://www.servethehome.com/amd-ryzen-7-1700-linux-benchmarks-zen-buy/

Fuck it, /g/ needs to learn why Intel should be afraid and its not because of muh vidya. Ryzen is a direct assault upon the areas where Intel makes most of its money.

>>59270540
English is not my native language and I understand him perfectly.

Maybe this >>59269537 guy is pretty stupid.

File: amdronetimeline_0.jpg (2MB, 700x5000px) Image search: [Google]

2MB, 700x5000px

>>59269514

>>59270571
His argument isn't that it doesn't indicate future performance. Only that we should take into account developers working multi-threading more extensively into their programs. Benching at low-res is still useful for determining performance, he's arguing that shouldn't be the only factor.

The question then becomes: would you be willing to purchase a weaker CPU today, in the *hope* that 4-5-6 years from now, devs *may* take greater account of its cores in the future? Even if you're still using your computer then, even if you intend to keep it around for another year or 2 at that point, you're probably starting to think about upgrading, whether because of CPU performance increases, or chipsets, or whatever.

And that's the clincher. You don't keep your CPU forever. It needs to be useful when you use it. Nobody cares that the 1700 is faster than the 7700K in 10 years. It matters today, not then.

>>59270880
>How is it more representative?
Because it's a game. You want to know how well the CPU runs in a game. So you use it. In the game.
The same draw calls and AI computations at 1080p at the ones happening at 4K.
This isn't rocket science. Again, reviewers have always told audiences that synthetics aren't necessarily useful for determining gaming performance. They never said that same thing about low-res game testing. Because one is a real-world game load on the CPU, the other isn't.
>However synthetics like 7z/Photoshop/etc
A synthetic designed specifically for Photoshop might be useful for Photoshop. But it's not necessarily representative of performance in Doom.

>>59270512
>because I am a programmer and I compile a shitload.

literally the same reason I'm pressuring my employer. It helps fucking a lot.

>>59269514
It's not like Intel could release 7th gen CPUs for 6800k, 6900k. Literally ryzen just came to a late party that's going to be outdated. AMD should've released competition for i5 i3. Let's be honest that $60 pentium is amazing for 400$ budget builds

>>59270952
Intel makes most of its money from their consumer CPUs not Xeons. Look up Intel market segmentation in financial reports.

>>59270985
>devs *may* take greater account of its cores in the future

This is something I wonder about too. With Ryzen you have to make a wish, gamble, that devs optimize games for more than 4 cores. It comes to gratification too, people want to feel good about the things they purchase especially when they cost 400€+ immediately, not in few years

>>59271030

>Intel makes most of its money from their consumer CPUs not Xeons.

Only because they lump the mobile stuff in with the desktop chips. Mobile and server chips are Intel's bread and butter. In practical terms Intel only make twi tiers of chips - mobile focused parts and xeons - everything else is an overclocked/cut down version of those.

>>59271020
what kinds of stuff are you guys compiling ? It never takes me so long to compile that I would factor it into a CPU purchase.

>>59271028
Sure Intel can do that but they aren't worth the premium $1000+ that Intel has been tagging them with.

>>59271062

Now if I can count if a shiny new processor saved you 30 seconds per day compile/render time over the course of a year with a chip running 24/7 thats an additional 7 days worth of processing available to you.

Naturally most cpus won't ber run like that even within this context but the time savings sure add up. Power consumption savings are also potentially a considerable factor. Its why for servers the cost of the chip is basically meaningless - its the cost to run the fucking thing that matters.

>>59270985
>>How is it more representative?
>Because it's a game. You want to know how well the CPU runs in a game. So you use it. In the game.
>The same draw calls and AI computations at 1080p at the ones happening at 4K.

This is not necessarily correct. As resolution increases, the amount of detail you can cram into the same space increases - what was ~200 pixels 100 metres away in 1080p might be ~800 pixels at 4k. So to compensate you need to increase drawing distance. This directly increases draw calls.

Since we tapped out single thread performance for the near future, to increase draw calls you need threads. Since there is a push for 4k, you need more detail to fill the void.

8/16 CPU is actually as future proof as possible. 4/8 cpus are going to struggle in ~1-2 years, just like dual cores were starting to struggle in 2010.

>>59271062
EE software suite. Full compile of my project on our distributed compile farm takes about 5 minutes. Full build around 1.5 hours.

File: gta-v-cpu-1080tx-vh.jpg (51KB, 806x532px) Image search: [Google]

51KB, 806x532px

>>59271146

>just like dual cores were starting to struggle in 2010.

LIES. /g/ told me the anniversary pentium is the best cpu ever.

>>59271119
Sure but in my development environment I am also using shit that is single thread bound. So while compile time might be fast actual work speeds would slow down.

>>59271174
Funnily, the only reason I could play DA:Inquisition was because I bought core 2 quad in 2007 :) That mofo served me for 7 years - precisely because it was a quad.

>>59269514
>hurrdurr it will be faster just wait
I've heard that shit so many fucking times when it comes to AMD products, I want to believe it, but I can't help but think 'Yeah I've heard that before'

Should you really buy a product that 'might' be faster in 3 years, or should you get a product that is fastest now? and might still be fastest in 3 years?

>>59271119
Replacing my 4790k with 1800k would halve local compile time and would benefit entire compile farm and automatic testing farm.

I reckon It would yield me 20-30 minutes/day.

File: AMD.png (265KB, 872x493px) Image search: [Google]

265KB, 872x493px

Since I have been shitting on Ryzen so much felt obligated to post this. Finally found a game where it performs better.

>>59271248
It performs worse than the 6900k though

>>59271248
I wish they did tests with popular older overclocked cpu's as well, like 2500k & 3570k

>>59271257
>$1000 proc

>>59270201

this meme again. Mostly everyone plays at 1080p, and that's only very recently started to change.

Keep pretending we all can afford an Asus IPS 4k 144hz monitor though.

>>59271286

Whats more interesting is the huge gulf between the 6600k and 7600k - something not shown for their respective i7 brethren.

File: intel-mobile-chart.png (29KB, 725x655px) Image search: [Google]

29KB, 725x655px

>>59271056

But didn't they merge mobile and desktop because mobile was literally dying too fast for them too handle. i know pic related is years old, but I doubt that they've turned that shit around.

>>59270201
>>59271318

Isn't a better analogy testing ryzen at 1080p is like on normal roads, where as testing at 4k/144hz is like testing on Silverstone.

>>59271247

Wouldn't that save you over a hundred days in a year? I cannot into maths.

>>59271382
Higher clockspeed

>AdoredTV
>Pajeet
>Thinks all the press is wrong and lose the plot
>not shilling

Fuck off, the people working for the credible tech press know what they are doing compared to some random people on Youtube or on the internet.

Do you really think I care for your opinion when I see the hard facts presented to me?´

Think again, you wannabe tech guys.

>>59271416

The 7700k is clocked higher than the 6700k. Though I guess you mean a higher delta in clockspeed.

File: 614acf17877b72f2c4795136812e81e2.jpg (74KB, 300x450px) Image search: [Google]

74KB, 300x450px

>>59271412
>Wouldn't that save you over a hundred days in a year? I cannot into maths.
>browses /g/
>can't do elementary math

/g/ in a nutshell

>>59271421
>the people working for the credible tech press know what they are doing
kek

File: intelie.jpg (245KB, 1024x768px) Image search: [Google]

245KB, 1024x768px

>>59269588
>>59269537
I'm Russian and when I heard his voice, It sounds pretty pajeet to me. Well, now I know what scotts and pajeets have in common.

>>59269514

Wasn't the Ryzen supposed to be for Workload, 1440p applications?

I've seen tell that they've been busting reviewer balls about being critical of it's unintended 1080p performance or some shit.


I'ma wait till their 6 cores are out, maybe they'll be more up my ally. I also don't like the Bios instability they've experiencing apparently.

>>59269514
intel on suicide watch

>>59271257
that chip is twice the price
now you know why AMD puts the r7 against it
they can always fall back on the price argument, and rightly so

>>59271472
>kek

Okay

sandy bridge confirmed trash chip

intel on suicide watch

File: 20170306211034_1.jpg (324KB, 1920x1080px) Image search: [Google]

324KB, 1920x1080px

R7 x1800
GTX 970
2x8 GB Gskill Ripjaws 3000 15cl


i know thats not a good screenshot but the fps is steady troughout the game. pretty shure im gpu bound

>>59271473
Is this image real?

>>59270114
LOL

>>59270201
/o/ here
Bad analogy. There are cars built only for strait line performance

>>59270512

The scheduling issue doesn't stop there. At the moment it's passing work from CCX1 (first 4 cores) to CCX2, for which there is a performance hit along with L3 issues doing that, so it's a double performance hit. All instructions should be confined within each CCX module and its relevant SMT threads. At the moment Windows is throwing them all over the place.

Of course this isn't happening to all instructions, but it is happening and will hopefully be fixed soon. I really want a reviewer to disable one entire module/half the chip to 1) test the theory, 2) give us an idea of upcoming smaller Ryzen chips, 3) see the affect on OC headroom/power/temps

>>59271793
pretty cool that it gets such a consistent rendering timings even though apparently draw calls are shit

no idea how it works, too much unknown about whole thing and nobody bothers to check actually relevant data like this
they can shove avg fps up their rectum

>>59271973
it was tested, 60% performance increase
seriously.

>>59271973
So, essenitally, this chip is an on-die dual Numa with a common memory interface? No wonder it gets hammered so hard if Windows throws threads all over the place. The mere rescheduling of a thread to a different CCX if done too often will just about murder performance.

If this is the case, and gets fixed, then this is going to be a monster of a cpu.

>>59272042
source pls

File: 1461100431529.png (14KB, 800x401px) Image search: [Google]

14KB, 800x401px

>>59272101
it went from 30% behind to 5%

http://www.zolkorn.com/news/report-an-error-for-ryzen-7-1800x-vs-core-i7-7700k-mhz-by-mhz/

>>59272043
L3 is shared which makes things even worse for gaming under such windows behavior.

File: donald_dick_gif_by_funnydank-d52ntgy.jpg (81KB, 567x461px) Image search: [Google]

81KB, 567x461px

>>59269514
All of the tests I've seen that I know is not biased (mostly Scandinavian ones). The Ryzen CPU's are better when you up the resolution to 1440p and beyond. At 1080, Intel is slightly better.

But who aint using at least 1440p on their gaming setup today?

>>59269866
Must suck being a burger.

>>59272130
RAM is bottlenecking it desu.

>>59272201
A bunch of key clocks in the Infinity Fabric set up are tied to memory clock speed. In instances where data is passing between CCXs the chip can heavily benefit from faster memory.

>>59271473
cмeeшьcя дpyжe?
пoгляди IT crowd и Black Books для бpитбoнгoвcкoй тpeниpoвки cтaнeшь пoнимaть хaггиcoв тoк тaк

File: average-gaming1.png (72KB, 601x830px) Image search: [Google]

72KB, 601x830px

>>59269514
>AdoredTV

Actual AMD shill, fuck off

>>59272174
L3 is broken for the ccx's its not shared

the mapping of windows currently has all the physicals+logic as one resulting on a 138mb l3 virtual space
>>59272176
seriously now? when i logged in on march 2 i could have sworn EVERYONE was using their awesome 4k setup to run shit on 720p and 480p because "it shows the true strength of the cpu"

File: [Onizuka Naoshi] Emotive Ch. 1-3 [English]_P010.jpg (433KB, 1177x1146px) Image search: [Google]

433KB, 1177x1146px

>>59271421
>the people working for the credible tech press know what they are doing
>retarded spacing

>>59272176
Can you show us some screeniees of those viewies ur talking about?

>>59272269
posting the same image for days

still doesnt get it that despite all the OS problems the cpu is just 5% behind the muh 7700k

to be fair i thought intel shills had a bit of iq i guess i should have know better since they own intel parts

>>59272269
>10 fps difference is huge
If they were all on 30 or some shit, then I would be worried, but they're practically all in the 100 range. This isn't even bad considering Zen doesn't do gaymen well for whatever fuck reason and annihilates everything else.

File: 1458585747492.png (5KB, 518x230px) Image search: [Google]

5KB, 518x230px

>>59272269
post real benchmarks

File: ryzen-cpu-value.png (26KB, 773x371px) Image search: [Google]

26KB, 773x371px

>>59272310
It's fucking 20% behind the 7700k, it's awful

It's also twice the price if you buy an equivalent chip.

Ryzen has both horrible price and horrible performance, a complete disaster.

>>59272333

You know they'll never make a similar chart for workstation workloads right? Not only is it harder to quantify as there is no unifying performance metric it also has ryzen curbstomping Intel into the floor.

>>59272329
What a surprise, the 7700k wins in that one too

Ryzen is a complete garbage gaming CPU

>>59272130
How does this even make sense? Despite wanting to believe, I call bullshit.

>>59272353
how does 20% increase in w7 makes sense? same thing
wintel going strong in 2017

File: ryzen-autocad.png (135KB, 711x533px) Image search: [Google]

135KB, 711x533px

>>59272349
>workstation workloads

Please call it what it really is, rendering and encoding.

It's basically server workloads, the kind of crap you do on a a dedicated machine.

In everything else, Intel destroys Ryzen, every other fucking app.

>>59272333
>its fucking 20%

seriously did you ever made a single mathematics lesson in your life?

1800x is 105 +20% is 126 fps 11% more than the 112 of the 7700k

seriously you need to RMA yourself this is getting embarrasing for you

>>59272350
i'd agree with you if it was more than one frame of a difference, its not

>>59272353

Its the same sort of shit bulldozer suffered from as well has Intel's early hyperhtreaded chips. The scheduler just isn't allocating threads correctly and is slamming headfirst in the chips weakpoint. Its why for years hyperthreading really did give negative scaling in a lot of workloads.

/g/ has the memory of a goldfish and has long since forgotten Intel and MS spent considerable time tuning both the hardware and software to make hyperthreading work.

>>59271793
i had more fps on i5 2500k + gtx 1060, ryzen is so fucking shit lmao

File: 8GfYX.gif (920KB, 500x377px) Image search: [Google]

920KB, 500x377px

>>59272396
AMD master race

File: adoredtv.png (19KB, 981x106px) Image search: [Google]

19KB, 981x106px

>AdoredTV

He gives up on one AMD product and shills for another.

>>59272405
You realize the fucking 7700k is $160 CHEAPER than the closest Ryzen.

Literally no reason to buy Ryzen for gaming, it's slower and more expensive.

>>59272396

>47% increase in performance from an 18% increase in max turbo clock speed

Seems legit family

Also nice cherry picking of a clearly unoptimized application

>>59272418
https://www.youtube.com/watch?v=gYAI6zcL7tI

did you now?

>>59272400
No, that's only for the 1800X, it's much worse for the 1700

But go ahead and compare the 1800X too, it's horrible.

It's $500 for WORSE performance than Intel's $340 7700k

>>59272396
im sorry but correct if im wrong how this benchmark is actually done?

last time i checked autocad doesnt let you anymore to render anything on 2d since 2015

that is so stupid

>>59272445
He does favor AMD, but calls out bullshit on any side. I've remembered him praising Maxwell multiple times despite Nvidia being terrible for the consumer. His recent Vega video compares it to Fermi, being as fatalist as possible.

>>59272130
>obscure site as always
>first claim 60%
>then 25%
>still no proof from other site than random turkish fag with no explained testing methodology and specs
>2400mhz ddr4 ram

lmao, other guy tested it and he get max 6fps which was still out of reach i7 7700k

File: 1photoshop-ryzen-2.jpg (79KB, 701x479px) Image search: [Google]

79KB, 701x479px

>>59272463
I can post this shit all day

These are not unoptimized applications, these are real world applications people use all day

>>59272471

By your logic, the $1700 6950X is WORSE performance than the $340 7700K

>>59272501
you do it all day, fuck off already

>>59272494
God I fucking hate AMD fags so much

1000 sites will all show the same benchmarks

Then these morons will scour the internet for that one foreign site written by an actual paid shill that has different results, and they'll try to pass that off as legit

>>59272494
>I will believe only my favorite eceleb everyone else is lying

do you hear yourself? if anything non english press is more reliable

>>59272514
It is, 6950X is a horrible gaming CPU

You would never buy that for gaming, just like you would never buy Ryzen for gaming

>>59272526
go brush steve's hair, don't forget a lotion

>>59272466
nice trash video, dude's gpu is literally melting idk what the fuck is happening there and i don't care

https://www.youtube.com/watch?v=zX5MzWcoM-w

>>59272353
In some instances Window's load balancing was needlessly passing workload between threads. This has a notable performance impact when you're crossing CCXs. The two CCXs can access each other's L3, but with a much higher access time, so anything timing critical shouldn't behave in such a way. Utilizing only one CCX negates this issue.

>>59272466

Y'know, looking at this I think (not 100% certain) that my 8320e at 4.7ghz and stock clocked 290x in vulkan with everything to ultra (but no nightmare settings) actually outperforms what is shown in that video. The big fps killer in that particular section of the game is when you step out of the elevator and look over the cliff with the burning wreckage. I'm fairly sure my setup doesn't dip to below 70fps ever.

That doesn't sound right as there is no way my lowly 290x even under vulkan matches a 1080 but I do remember testing it a fair amount to see if a gpu overclock made any difference (in my case it didn't as running at over 200fps for most of the game means i've hit a cpu wall).

File: qrvnKy1.jpg (246KB, 1600x1000px) Image search: [Google]

246KB, 1600x1000px

>>59272471
its so horrible..meanwhile lets see what you are on about..

amd compares 1800x with the 6900k it says that it gives similiar perf for half the price

amd is literally 4.7% behind 6900k on your own picture FOR 500 bucks

do you remember how many prizes the 6900k got because it was "muh workstation cpu" "muh the best all around value" ?

but that isnt the good part.. amd later on said that the 1700 and the x is going against 6800 and 6700

but on /g/ (sorry i meant /v/) somehow everything is around 7700k anyway i mean we get it you are crying you are furious but at least try to be sane a bit

in case you are wondering this is one of the first re benches with the new bios on the aorus

it isnt too late to repent

>>59272501

As if we couldn't go on all week showing the scales tipped in the other direction, even with the architecture in infancy

just

>>59272543
>6950X is a horrible gaming CPU
assuming you can afford one effortlessly its amazing gaming cpu that smooths out any fps drops you can imagine

File: reminder.png (114KB, 1270x1546px) Image search: [Google]

114KB, 1270x1546px

Your daily reminder that this $329 chip with a TDP of 65W outperforms every single consumer Intel chip out there right now.

>>59272572
>posting simulated ryzen benchmarks

This is how bad it has gotten for AMD fags

They have resorted to literally posting ACTUAL ADMITTED MADE UP BENCHMARKS

THEY ARE LIVING IN A MADE UP WORLD IN THEIR HEADS

>>59272572
Stop posting this edited chart, kid.
The source is right here.
http://www.zolkorn.com/reviews/amd-ryzen-7-1800x-vs-intel-core-i7-7700k-mhz-by-mhz-core-by-core/4/

>>59272501

So are we to take that tom's magical 6900k has over 100% scaling?

>>59272561
oh wow, all those 27 and 30ms render times in an empty room

>>59272581
>assuming you can afford one effortlessly

Yes assuming all of us have infinite money, everything looks like a great buy!

File: wow.jpg (26KB, 600x610px) Image search: [Google]

26KB, 600x610px

>>59272501
>application that doesn't scale with threads at all runs best with the fastest single thread processor on the market

File: Disappointed.jpg (50KB, 330x327px) Image search: [Google]

50KB, 330x327px

/g/ is one of the worst boards on the website.
Release the 1080ti already so there's more shitposting.

>>59272599
do you even know what simulated cpu benchmarks are ?

jesus fucking christ why you do bother to reply if you are so tech illirate?

>>59272594

Get out of here with your synthetic nonsense

>>59272635
It's literally fake

>>59272581
If value didn't matter, you'd go with the 6950 for gaming anyway. So it's a moot point. There's no gaming scenario where Ryzen is attractive, infinite funds or no.

>>59272681
>There's no gaming scenario where Ryzen is attractive

THIS

>>59272676
its literally fake..

simulated is when you force a multicore cpu into a specific combination you ape

>>59272543

>Horrible
In terms of price:perf, maybe, but not actual performance at all.

>You wouldn't buy Ryzen for gaming
Stop pretending that Ryzen = 1800X only

Try doing absolutely anything in the background whilst gaming on a 1700 vs 7700K. 7700K is good for badly threaded games in a bare OS and that's fucking it desu

File: 1477976632948.png (45KB, 630x424px) Image search: [Google]

45KB, 630x424px

>>59272681
>If value didn't matter, you'd go with the 6950 for gaming anyway.
I can afford comparable smoothness for 329 right now.

>inb4 coomputerbase has fake bnechamrks

>>59272704
It's not an actual CPU you dumbfuck, it's SIMULATED, it's an ACTUAL FAKE BENCHMARK

File: exUPcT9VENor6YqqPhfVbc-650-80.png (267KB, 650x766px) Image search: [Google]

267KB, 650x766px

>>59272269
i always love to debunk that BULLSHIT SITE

literally 6700 being 5% faster than the 7700k three months before

keep using that intel shill site it brings me so much pleasure to see them doing shit like this

File: streaming-obs.png (129KB, 1440x1080px) Image search: [Google]

129KB, 1440x1080px

>>59272723
>Try doing absolutely anything in the background whilst gaming on a 1700 vs 7700K.

OK let's try doing some stuff in the background.

How about streaming?

OH OOOPS INTEL WINS AGAIN SORRY

File: 1462359397669.jpg (57KB, 600x798px) Image search: [Google]

57KB, 600x798px

>>59272737
anon, are you okay?

>>59272737
oh god please stop talking about things you dont know go back to /lgbt/ or whatever else board they vomited you out

File: 16-game-average.png (61KB, 1306x1646px) Image search: [Google]

61KB, 1306x1646px

>>59272753
Oh yes it's totally bullshit

That's why all the other sites show the exact same fucking thing

>>59272776
Sorry you can't handle the truth

>>59272329
Lmao well presented data
t. Intel Employee

>>59272726
>time spent over 14.1ms
Ryzen = 0
7700K = 34618

>>59272792
says the person that thinks a simulated run on 4c is fake without understand the purpose of a locked chip

>>59272760

You literally just showed that Ryzen takes MUCH less of a performance hit than the 7700K when running OBS.

Did you even look at it before posting lol?

>>59272848
Jesus christ, you need to learn to read

Look at the performance

The 7700k is the clear winner for streaming and playing games at the same time

>>59272863
Its actually not. Ars used totally unrealistic low bitrate settings on their test. If you recorded that footage at their settings and looked at it you'd see a stuttering mess with artifacts everywhere.

>>59272353
go back to /v/

>>59272566
>CCX
I get a car when I search this term

>>59272863
>The 7700k is the clear winner for streaming and playing Dota 2 at the same time

fixed that for you

>>59272863
>you need to learn to read
you first, numbers seem to confuse you, let me help
ryzen didn't lose performance with obs
7700K did lose performance with obs
with retardedly low bitrate settings nonetheless

File: 1.jpg (271KB, 720x540px) Image search: [Google]

271KB, 720x540px

>>59272920
CCX = core complex.

>>59272889
>>59272942

>lower FPS is better

AMDtards truly do live in an alternate universe

>>59269514
The biggest thing people are avoiding talking about is that the new Ryzen CPU's seems to have solved the "stutter" problem that has plagued every AMD/Intel CPU up until this point INCLUDING the 7700K

So on max settings at 1080p you have a 7700k getting 120fps and a ryzen chip getting 112fps but the Ryzen chip is smooth rendering and without the stutter.

why the fuck arent more people talking about this? apparently Ryzen has a radically new memory subsystem designed that has never been done on an x86 cpu before and its drastically improved the rendering consistency.

seeing the goddamn stutter big finally solved means i will be exclusively be buying Ryzen chips until intel has a new arch that solves the problem too.

Jim Keller is a fucking god.

>>59272778
none of those can drive a signal to 144hz, so whats your point?

>>59272863

7700K
With OBS
-11.8% Average FPS
-20.7% Min FPS

Ryzen
With OBS
-2.7% Average FPS
-3.4% Min FPS

Back to school Jamal

>>59272965
>The biggest thing people are avoiding talking about
its hard to explain, try explaining how 144hz monitor is better than 60hz
people that never used it can't understand

>>59272991

Not to mention frametimes, I'm sure that they would be fun to look at scenarios like that :^)

>>59272962
name a streaming site that allows over 60fps

>>59272965
At equal clocks Zen has just as much L1 bandwidth as intel's core arch with the same latency to boot.
Zen also has more L2 and L3 bandwidth than intel's cores, by a very significant degree.
The caches in Zen are pretty impressive. More reviewers would cover it but they tend to stick only to what they understand, and that usually isn't much. The number of big reviewers actually qualified to understand the coherent Infinity Fabric might only be 1 or 2. Anand himself doesn't even write any more.

>>59273013

Would you rather stutter at 144hz or be smooth at 120hz? The fps isn't even that big in many games and will only get smaller. Ryzen is the CPU to get for gaming and real world use imo.

>>59273043
draw calls are lower though, why?

File: ryzen-r7-1800x-bench-bf1.png (120KB, 806x580px) Image search: [Google]

120KB, 806x580px

>>59273064
>Ryzen is the CPU to get for gaming
It gets lower FPS than similarly priced or cheaper Intel chips. For gaming, it's a bad buy.

>>59273043
>>59273080

As covered in some previous posts, Zen's L3 is currently being crippled by Windows for quite a few instructions that incorrectly get thrown between CCX units.

>>59273102
>still clueless what people actually talk about

>>59273130
If stuttering is your primary concern, note the 0.1% lows.

File: 1488664232472.png (84KB, 653x726px) Image search: [Google]

84KB, 653x726px

>>59273102
y-yeah, a bad b-buy

g-get intel

>>59273102

>Day0 benchmarks

Come on lad..

>>59273146
if they publish full graph I will
they did not
one dot in time has no meaning

>>59273043
What irritates the shit out of me is this low resolution fps testing to gauge the "gaming performance" of the cpu

it doesnt not represent max settings performance at all and in the cases where intel does deliver better fps at realistic settings at 1080p/1440p I have to circle around to the fact that zen can deliver over 60fps on 4K which is all that matters for the vast majority of people. It can deliver acceptable frames for a 120hz/1440hz screen as well and the next gen consoles will be running zen which will further push game devs to utilize more than 4cores on game engines going forward.

Jokers video was a watershed moment. Both the 7700k and the Ryzen 1700 delivering 136+fps on max settings on battlefield 1 but the 7700k was red lining at 95%usage and the zen was only at 44% usage.

that displayes how big of a deal this 8 core zen chip is going forward.

>>59273025
>over 60fps

are you this retarded?

AT 60fps stream i7 7700k is clrea winner vs 1800x

File: 1_16000.png (20KB, 1174x452px) Image search: [Google]

20KB, 1174x452px

>>59273080
Impossible to say without doing an indepth analysis of the chip's performance while its running the bench.
The program could be handling the threads in some aberrant way.

Though looking at results other people have supplied, it seems this test is utterly worthless.
2.6ghz Core2Duo on par with a 3.1ghz Sandy Bridge i5. A 3.5ghz Skylake i5 barely being any faster. Dual core Haswell at 4.3ghz totally murdering a 5.1ghz Sandy Bridge i5.
This test objectively seems worthless.

>>59272396
>CAD
What else, you going to post Adobe shit too?

>>59273234

Look again chap, the GPUs are all random as fuck

>>59273261
>>59273234

Also varying OS and probably software revision.

>>59273261
Is this not supposed to be entirely CPU bound?
Unless the GPU drivers themselves were having a meaningful impact, I don't see how it should matter. Its just measuring the CPU passing draw calls to the GPU, not measuring what the GPU can render since the entire bench is very low utilization.

>>59273234
I was refering to data that 3dmark gives, ryzen has twice lower drawcalls on same gpu. Doesn't affect score it seems, but still can it be only L3/w10 fault.

>>59273301

I'm not sure desu senpai, just seems like a red flag to me

>>59272943
Thanks
Now to figure out these chip naming schemes.

>>59269514
this guy is such an amd shill holy fuck i remember him shilling 480s and how amazing they will magically become in 5 years when they are obsoleted by 75w nvidia cards.

>>59273339
>>59273301
its purely synthetic, but relys on drivers
so data is from pre gcn has no meaning, it also only works on amd

>>59273362
480 is amazing right now fyi, 7 months after proper partner release

>>59273362
He shits on Polaris and Vega in two videos tho. His pc is intel/nvidia.

>>59271200
Yes if my mobo have not have crapped I was going to get zen now. Too bad I had to buy i5 I don't feel like replacing 1 year old system. Probably will get Zen next year when there are better mobos around.

>>59271421
>Pajeet

Are you retarded? How do you not know what an Irish/Scottish accent sounds like? Seriously, do you have some form of brain damage?

Virtually no games today will be updated for Zen performance.
Most games moving forward will not target Zen performance.

If you don't have a real purpose for 8 cores then you should probably go Intel, unless the 4-6 core Ryzens turn out to be competitively priced and you're a poorfag.

>AMD
>Here is a new piece of technology that has infinite possibilities it all depends of the developers now to make use of it
>Intel
>MOAR VOLTS MOAR ENERGY MOAR MHZ

And this is how the market works.

>>59272599

simulated means they locked the cores on the cpu to simulate a 4/8 cpu you retard.

i mean ffs, are you dumb or just a shill?

>>59273421

If you think irish and scottish sound similar you equally have brain problems.

>>59272681

except the 6950 is actually shit for gaming and costs 1500$.

the 1800x is a simliar performer for 500 and the 1700 (which imo is the best buy since its basicly a 1800x downlocked) gets close to it when oced and costs 330$.

its a no brainer.

>>59273453
Not equal, but definitly similar.

>>59273421
Irish pleb detected and triggered

>>59273462
1700 is not only downclocked it has lower tdp of 65W.

>>59273462
>and the 1700 (which imo is the best buy since its basicly a 1800x downlocked) gets close to it when oced and costs 330$.
And is soundly beaten by the 7700K in both average and minimum framerates in games.
Looks like you can have your value and your performance too. But not with Ryzen.

>>59273477

that might be because of the lower clock? but still thats damn fucking impressive.

im just waiting for a decent mini itx.

gonna make a 5litre 8core with a 1080/vega in it and hook it behind my monitor. just for the lulz.

>>59273503

except it isnt beaten. once you tread to the 1440p territory it is pretty fucking equal. sure it beats it in 720p and 1080 on low, but on ultra it is even or outperforms it. and sure you faggots might say gpu bound or w/e. but all that matters for me is performance at the higest quality settings.

File: spurdo.png (117KB, 1024x749px) Image search: [Google]

117KB, 1024x749px

>>59272396
>cadalyst

I upgrade a cpu every 4 years and the gpu every 2.

Ryzen will be better than 7700k in half or at most a year. Plus i use my pc realistically with multitabbing playing multiple games etc.

>>59273527
>once you tread to the 1440p territory it is pretty fucking equal.
Even at 1440p, Ryzen sees less performance for equal money.
You can move to 4K if you like. But GPUs will get more powerful, and you'll likely replace your GPU before your CPU. When 4K becomes today's 1440p and the bottleneck disappears, you'll end up with a CPU that chokes sooner than a 7700K, requiring you to upgrade sooner than you might prefer. You can account for the possibility that future games support more cores in the future, but that's your call, and introduces greater risk in the investment.

>>59272501

>Present relative data
>Origin at 100%

there are a number of things the CPU must do to run a game, regardless of whether or not frames are being drawn. How much of the CPU is actually typically used in the process of drawing frames, vs how much is used in these other activities?

Does drawing a frame use the full IPC potential of the CPU?

If you reduce the GPU load to the point that the CPU fills in its available clocks just drawing more frames, isn't this a less efficient usage of clock cycles than a more a complex game engine at a reasonable frame rate?

What would be the result of running a GPU bottlenecked game, while encoding video in the background under lower process priority. Which CPU could do more with the remainder of their resources?

File: _________.png (1KB, 375x34px) Image search: [Google]

1KB, 375x34px

I'm off to bed lads. Keep up the good fight; whichever side that you may or may not be on :^)

>>59273689
(You)s were a mistake even though I lost my shit when they were (...).
Here's one extra, just for free.

>>59273387
My pc is intel nvidia but thats literally due to the lower performance of FX chips and the fact that AMD didnt offer a GPU in the power class of my GTX 1070.

thats it. If AMD offered a 490 that matched the 1070 i would have bought that because AMD crimson drivers are comfy as fuck and geforce drivers suck massive asshole.

Zen is good enough in single thread to keep up with intel and offers more cores for less money which boosts longevity of your system. Also offering a B350 board that can overclock your CPU but does not tack on the extra cost of bullshit SLI compatibility is a huge plus.

The 1700 chip at 330$ can OC up to 3.9ghz and a cheap b350 means this is a rock solid workhorse chip set that can do literally everything well. The i5/i7's are just good for vidya gaming and not much else.

>>59273632
the "bottleneck is meaningless because its only running on older engines/API's that are using 4core optimization. Its painfully obvious we are in the middle of a major transition to 8core optimization and all of the bugs of this brand new Gen1 design have not been ironed out with BIOS updates yet.

You are complaining about the equivalent of a first gen Nethalhem core i7 processor.

>>59273422
Games are going to use more threads moving forward.

-AMD octacores in both consoles
-both big game APIs, Vulkan and DX12, feature expanded multithreading support
-4k resolution and the resolutions required for good VR gaming are tapping out 4c8t parts
-signs that Intel is reaching theoretical IPC limits of the Core architecture and/or x86; performance gains can only be had by added more cores (this was not true when AMD built Bulldozer)


>Virtually no games today will be updated for Zen performance.
Part of the problem is from Windows itself, and Windows will absolutely get patched.

>>59271964
American car.

>>59270985
AMD is working on a fast track for Ryzen. 300+ devs already signed up with 1000+ by years end targeted. This is unprecedented for AMD. If they follow through things could change real fast. Not years down the line. They have put this in place for a while now and is part of AMD's plan which is consoles, Vulkan/DX12 and server workloads.

>>59272602
It was not edited. They silently updated the charts after this was posted. Check the archive site to see for yourself. It still performs close to the 700K on same cores and clock speed though. Once the patches roll in it will probably move ahead.

File: 1488425477998.png (146KB, 500x515px) Image search: [Google]

146KB, 500x515px

>Playing video games.

File: Screenshot 2017-03-06 at 8.25.58 PM.png (80KB, 758x196px) Image search: [Google]

80KB, 758x196px

>I-I swear AMD isn't made by Pajeets, for Pajeets

>>59272130
>kabylake having about an 8% IPC advantage

hope that's still because of optimisation. i thought ryzen was supposed to be a lot closer.

>>59277725
epic

>>59277795
Why did you think that?
Even by AMD's own Cinebench tests Kaby lake has a 6% IPC advantage over Zen.

>>59277816
most of the reviews i've read of skylake had it within about 3% of broadwell in production benchmarks and actually behind in video games, and amd portrayed ryzen as having roughly equivalent ipc to broadwell.

File: Screenshot 2017-03-06 at 8.30.09 PM.png (99KB, 1226x878px) Image search: [Google]

99KB, 1226x878px

>>59277812
PLEASE SIR DO THE NEEDFUL AND BUY R7 RYZEN CPU AND RADEON GPU. AMD OFFERS SUPERIOR VALUE WHICH IS IMPORTANT FOR GAMERS LIKE US.

@59277862
>p-please reply to me, i'm very lonely

File: AMD-Ryzen_4-1140x641.png (772KB, 1140x641px) Image search: [Google]

772KB, 1140x641px

>>59277850
With totally equal clocks Zen is about 2.7% behind Broadwell in Cinebenc 15.
IPC isn't one thing because every architecture has its nuances, and performs better in certain areas while worse in others.

>>59277877
PLEASE BUY AMD. MY FAMILY HAS TO EAT. I HAVE 6 SONS. YOU ARE LETTING THEM GO HUNGRY IF YOU BUY INTEL. YOU ARE LIKE BRITISH.

>>59277892
yes, i realize now that blender and handbrake were outliers for AMD and were misleading benchmarks. i thought that blender at least would have been roughly representative. the situation appears even worse than cinebench in many applications. that's my point. it is disappointing.

>>59277925
They're not outliers at all, nor are they misleading. Performance is not derived from IPC alone. Multicore scaling, and SMT throughput, and frequency all factor into ultimate performance.

You have to be a shitposting shill to actually type all that out and think anyone would take you seriously.

>>59277946
>Multicore scaling, and SMT throughput, and frequency all factor into ultimate performance.

right, and as it stands it's losing out to both broadwell-e and kaby lake in all those areas.

i'm not shilling, you're just taking my observations on its ipc too personally. that's your problem.

>>59272396
>Please call it what it really is, rendering and encoding

>what is science

File: CB15.png (37KB, 960x540px) Image search: [Google]

37KB, 960x540px

>>59277963
Why are you trying so hard to shill when you're obviously so far out of depth here?
You keep making entirely fallacious posts and expect them to be taken at face value.

Stop acting like a kike.

>>59272543
Whoa whoa whoa, cool down the retard a little. The 6950X is literally the best gaming CPU there is. The only flaw with it is the price.

>>59278007
>cinebench isn't misleading, look at this other benchmark that is also cinebench.

uh huh.

i don't know why you're so mad about this but have fun with it.

File: Average SMT uplift.png (36KB, 960x540px) Image search: [Google]

36KB, 960x540px

>>59278037
You're a little shill talking out of your ass.
The 1800X has higher base frequency, all core turbo, and peak single core turbo than the i7 6900k. So again your statement that Ryzen is losing to Broadwell-E in that metric is entirely fallacious.
In an average of over 20 different metrics AMD's SMT implementation yields a higher throughput uplift than intel's HyperThreading. Even with BIOS issues at launch, and Ryzen having bugged performance in Linpack, AMD's brand new arch has higher SMT throughput than intel's latest.

If you're not a little faggot child shill why are you so dedicated to outright lying in all of your posts? A normal healthy adult can act mature and stay within the realm of facts. For some reason you can't seem to handle that.

When do the Zen server chips drop?

>>59278105
Sometime in Q2

My job at the dick sucking factory is to use winrar all day, ryzen will be a day 1 purchase for me

>>59271888
No? Why the fuck would any mainstream media give a fuck about Intel or AMD? Also, the "LMAO AMD" in the corner should have tipped you off.

>>59278082
>The 1800X has higher base frequency, all core turbo, and peak single core turbo than the i7 6900k.

that's nice, but it only means that's it's clocked higher out of the factory. I've yet to see a single Ryzen 7 with a stable overclock above 4.0 ghz. Gamers Nexus for instance have their 6900k overclocked at 4.4ghz trouncing the 1800x in just about every single benchmark, whether it's synthetics or real-world applications, and decimating it in gaming. that AMD has better SMT yields in some cases is a nice academic aside. It's even less relevant than IPC as a measure of absolute performance.

To recap: As it stands, Ryzen is significantly technologically inferior to both Broadwell-E and Kaby Lake. It does better than Kaby Lake in some - certainly not all - multithreaded applications, which you should hope so considering it has twice the number of cores and thread.

I will never understand why normal healthy adults let themselves get so emotional about CPUs, Suck it up, whinge and cry and stamp your feet if it helps, but those are just the facts.

>>59270114
What's your ambient, nigga? Where the fuck do you live?

>>59269588

Are Americans really, truly this fucking retarded?

>>59278218
Nope, Only Intel shills.

File: specs.png (37KB, 660x523px) Image search: [Google]

37KB, 660x523px

>>59270114
stock cooler

>>59278203
Oh look, the little shill makes it blatantly obvious.
I've always wondered: Do you kikes organize on reddit, a Skype group, or what?

You outright lie about frequency.
Then stock frequency doesn't matter.
You outright lie about SMT uplift.
Then SMT uplift doesn't matter.
I suppose you'd claim that energy efficiency doesn't matter either, considering the 1800X only uses 78% of the power of a i7 6900k. Being even 10% behind in total performance while consuming almost 30% less power is a huge advantage.
Though I'm sure you'll just keep acting like the childish shill you are, little subhuman kike.

>>59278203
6900k is twice the price and less power efficient
SMT is a problem with the windows scheduler

>It does better than Kaby Lake in some - certainly not all - multithreaded applications,
>some

absolute bullshit

>>59270114
You put the sticker on before the thermal paste, didn't you

>>59278278
>You outright lie about frequency.

No, I was just alluding to a different figure to you, ie the average overclocks. This is a technology board, I generally give people credit that they know how to overclock their CPU. You want to compare everything at stock for some bizarre reason that could only mean anything to absolute laymen. If you don't know how to overclock then watch some youtube videos. Granted, I can't promise anything won't go wrong on your Ryzen CPU given the state of the AM4 UEFIs. I can promise you you won't hit a stable 4.4ghz on all 8 cores.

>You outright lie about SMT uplift.

No I didn't, in fact Ryzen blatantly loses performance in many applications when SMT is running, particularly gaming.

That's neat about its energy efficiency, but if energy efficiency were the be-all-end-all we'd all be running shitty ARM tablets. I'm referring specifically to its absolute performance against the competition. It comes up short. That may change with firmware updates and better support in Windows and with software updates. Ryzen still has a good value proposition IF those get fixed.

Sorry you're so upset about a computer part, but I've got better things to do than argue more with angry little fanboys.

>>59278406
>loses performancein applications when SMT is running
heavy, symmetrical loads actually show more improvement with SMT than HT. In applications where the OS has to dynamically balance the load, the performance takes a hit. This has more to do with windows shitting itself than it is about SMT being """""inferior""""".

Holy shit you gaymurs are such retarded cancerous children. Video games can be made to use any kind of CPU in different ways. There's a reason Ryzen does very well and even beats a lot of Intel's best offerings in objective non-gaming benchmarks. Faggot video game makers have been optimizing for Intel CPUs for years.

If some developer only makes games run on 1 fucking thread, does that mean some little snot nose's i3 is better than my 18 core xeon?

Fuck off. Game developers are retards who took Intel money.

File: akQ5oUg.png (129KB, 807x906px) Image search: [Google]

129KB, 807x906px

>>59273503

>soundly.

thats 3 fps there chief. and thats not taking into account the 7700k stuttering.

the biggest problem for ryzen isnt performance. its the fact that it cant OC for shit.

>>59273597

thats wrong on every points tho. the ipc coupled with more cores makes the ryzen future proof.

in 6 months time it will shit on the kaby lake. intels only response now is to bring in cheap 6/8 cores to compete.

the higher res you get the better ryzen does.

File: ______.jpg (100KB, 640x640px) Image search: [Google]

100KB, 640x640px

Is there any competent reviewer out there that has a LIVE review with incremental additions for each BIOS update, relevant windows patches etc?

Also memory speed scaling. Also with GPU boost 3.0 disabled and a static frequency on the chosen graphics card FFS (Joker and many others have fallen prey to this one). Also with a min/avg/max thread uti% and some kind of calculation to scale the theoretical 100% performance ceiling of each CPU.

All of these rushed, incomplete day0 review cuckposts need to stop.

>>59279835

got to agree. so many things that factor in are just laughable. the only person that comes to mind is adored. but he hasent done his review yet. the guy levels the playing field and will hopefully give a decent and fair review.

File: 1488067634728.jpg (18KB, 267x297px) Image search: [Google]

18KB, 267x297px

>>59272269

File: _______________.jpg (73KB, 750x604px) Image search: [Google]

73KB, 750x604px

>>59279889

There are issues in being a purely video reviewer in that you can't go back and fix details or make small updates (without the convolution of multiple videos), but yes; in general he's very good and finds correlations and angles that others do not. His >optimism< may be biased, but when it comes to the raw data his opinion is not.

I was especially interested in his point about bulldozer eventually pulling ahead of the 2500K, albeit taking 4+ years. Now that the 8 core Ryzen CPUs are in consoles the turnaround for WineFine should be massively reduced, along with the fact that threading utilisation in general is growing exponentially. Even budget phones are octa+ now. AMD really have the groundwork and headroom here. Intel is losing its foundation and the end of their runway is in sight.

>>59279652
>the higher res you get the better ryzen does.
You don't understand the charts, at higher resolutions than 1080p you usually hit GPU bottleneck which is why they're at similar FPS rates. Both CPUs are waiting on a GPU.

Ryzen doesn't get better at higher resolutions, GPUs just get worse.

>>59280128
>Now that the 8 core Ryzen CPUs are in consoles

wtf are you talking about?

>>59280143

alright fair point. my point tho is that if the gpu bottlenecks and the cpus perform the same at 1440p or whatever res. then the whole debate regarding which cpu is better is pointless.

this of course is related to one instance and thats gaming, so if they are identical in gaming at high res, then other factors should be considered like more cores that can do more things at the same time.

>>59280143

I think the point is something along the lines of the Ryzen being able to feed the GPU what it needs at a higher rate, whilst doing its normal low-thread game CPU work at a slightly lower rate (mainly due to optimisation of a fresh architecture and raw MHz). So it loses in 1.3K but pulls ahead in 4K

>>59269588
Are you retarded?

>>59280203
I do agree with you in some ways, I find it ridiculous to say that Ryzen isn't good for gaming. How much performance do you need for a fucking game, it's 20fps difference for games that already run at 100+ fps, it's not that big of a deal.

Plus, at least you're not dropping frames like you do with 7700k. No stutter

>>59280195

I think that he meant "coming* in consoles", or that AMD in general is in consoles.

Scorpio is pretty much confirmed for Zen, and of course the PS4/PS4Pro is at least on Jaguar

how is CPU utilization correlated to future?

how true is that 7700K stays at 95%+ utilization but Ryzen at 50%? how will it translate into future proofing?

>>59280282
Zen should be in the PS5. PS4 Pro is Jaguar.

>>59271048
its not a fucking gamble
devs are FORCED to optimize for 8 cores since the consoles have a crappy jaguar 8 core, and it will even get better if Scorpio has a Zen based core.

>>59280328

Hard to really say for sure, but in an ideal scenario a 200fps@95%uti 7700K has theoretical headroom for 210.5fps, whilst a 170fps@50%uti Ryzen has theoretical headroom for 340fps.

Obviously it's a lot more complicated than that and very much down to the API and game coding, but on paper the Ryzen architecture is the one to watch.

>>59280380

It could also be down to Ryzen being choked of L3 and not being able to top out, due to the CCX instruction crossover issue in Windows.

>>59280451

Linux gaming benches when?

>>59280467
Doesn't phoronix have some?

>>59271888
>AMD LMAO
are you fucking retarded?

>>59280328

a workload created for 4 cores will use a maximum of 4 cores or 50% of an octacore.

if the software supports full 8 cores, then you would see the zen utilizing 100%, but that wont happen soon because why would a game use 8 cores? what would the job needed to be to fully do that. the game would need an insane amount of physics or something to do that and i dont see t happening anytime soon with the GPU being the limiting factor in games.

>>59280467
there are some

http://www.phoronix.com/scan.php?page=article&item=amd-ryzen-gaming&num=2

Minimums are much higher than 7700K.

>>59271174
>disabling HT gives more performance
ok ?

>>59280513

In many cases it seems that Linux has issues of its own on Ryzen, but they mention that at the end and that they will be trying different compilers etc soon.

>>59280491
>if the software supports full 8 cores
battlefield 1 does.

it's a fucking stupid point to make as a reason to buy 1700 instead of a 7700k as a pure gaming cpu.
Bc by that logic you should buy fx-8350 instead of a 2500k in 2012 bc if youwait 5 years it manages to pull 10% lead. Have fun waiting.

That is assuming his number even match reality.

>>59280656

battlefield will not use all 8 core of a cpu to its fullest.

>>59280660

thats not the point. the point is that the 1700 overclocked performs similar to a 7700k overclocked in gaming when the graphics are set to ultra or w/e its called.

AND the 1700 is futureproof.

and the 7700k is a 100c cpu, and is more power hungry. and only a 4core.

>>59280660

The difference is that the optimisation turnaround will be shorter, its performance is already well up there and well ahead in multitasking, no stuttering, temperatures and power are good, etc etc

>>59280513
>minimum
You sure mean power consumption, right?
Too bad Phoronix couldn't turn off SMT, or we would see whether disabling it would give gaming perf increase or not in Linux.

>>59271119
does it help with webpack compiling? I know this sounds stupid but I use webpack and reload the same page 10,000 times a day while developing and have to wait 500ms-2000ms each time just to load a page from localhost.

>>59280677
bf keeps 7700K at 99% utilization on 8 threads.

>>59280719
nah, I just seen that in CSGO(the one that must run 500fps at 640x480 or you can't compete)

7700K on linux gets 52 minimum fps
and ryzen on linux gets 69 minimum fps

>>59280282

Don't forget Apple.

Basically it's in every sftw developers interest to support multi-core multi-thread cpus. AMD is tiny compared to Intel, but basically for Intel it's all downhill from here. EIther it loses market share or profit margins. AMD comes from such low base that even a small increase in market share and margins will be a big deal.

>>59280749

alright. i stand corrected.

from a reviewer.

Results: (min, avg, max)
6 cores/12 threads: 195, 136, 103
4 cores/8 threads: 188, 130, 88
4 cores/4 threads: 128, 85, 47

>>59271174
lol, not every game, but many games just cant deal with 2 threads only anymore, they are moving to 4, but that's where the line is going to stop, because so many i5's and so many 2 core i7's exist, they can't drop 4 core support unless the game literally can not boot otherwise.

>>59280526
microsoft's scheduler is known to be dogshit, especially windows 10 apparently, they did something bad, win 7 performs better with ryzen then 10 does.

>>59271404
how much of that is due to contra revenue?

>>59269987
He is interested in new tech, and amd constantly does that, successfully or not is up in the air.

>>59271449
an 8 hour work day, assuming sunday off it would save you 6.5 days a year.

>>59269514
>buy intel play gayme
>no fuck die alone
>buy amd get work done
>insert benis + money

File: 1.png (486KB, 700x574px) Image search: [Google]

486KB, 700x574px

Intel got owned, time to take out nVidia with Vega

>>59281253

>inesert cpu into anus

indeed.

>>59269514
>Taking Adored seriously

He is literally "Every benchmark that makes Ryzen look bad is part of Intel smearing program" tier of retarded.

>>59270114
>2500k with nhd14
>1,36vcore
>26°C

>>59269942
But that's not what this video showed. It showed that the 7700k is already getting maxed out by current GPU's and games, whereas ryzen isn't. More powerful GPU's are going to be held back by the 7700k compared to the 1700/1800.

>>59278390
Do you think I'm stupid, I put the sticker on top of the thermal paste

>>59272475
>last time i checked autocad doesnt let you anymore to render anything on 2d since 2015
What did he mean by this? Is this considered English?

>>59277532
>underclocked
hmmmm

>>59281578
It's comparing 4.0ghz to 4.0ghz

>>59281587
It's retarded though, ryzen has troubles hitting 4ghz while most intel chips can easily hit 4,8ghz

>>59271226
Because it won't be, intel is moving forward with 6 cores to the i7 lineup, the i5's may get smt finally, and the i3 may become the 4 core no thread, and with amd being competitive, intel may move back to solder instead of tmi.

either way, you got programs that will move toward more core, or you have programs that will instance out to multi core, either way, you are better served with a 1700 than a 7700

the memory hiccup is an issue that will be sorted out in time, windows scheduler... while it is an issue, can be sorted out in the mean time by disabling smt on 10

>>59281600
You're fucking clueless jesus

>>59279889
Intel shills hate him not only because they believe he shills AMD (He does but he also criticizes them too). But because of his thick Scottish accent that Amerilards have trouble understanding (plus he always sounds a bit smug).

I am guessing he is waiting for some updates and patches to fix the various issues before trying to do a review. It's fairly pointless to review right now with all the bugs. They show it is a beast for 'most' multi-threaded applications. It only falls short in a few outliers that rely heavily on Intel optimizations and the first few cores and it can only get better in games once the bugs are fixed and games get patched (and newer games will sort this out anyhow). Windows is obviously a big iiissue here since MS are known to be in cahoots with Intel anyhow. It's probably why AMD did not get MS to fix it before release in case they 'fixed' it if you catch my drift. Best to get it out there and let others discover MS needs to patch their shit up.

>>59272176
decent 1440 cost over 500$ and oleds are soon, rather hold over on my 1200p monitor then get a shit in between one.

>>59281600
it's an interesting comparison. not every chart needs to be showing one side trying to blast the other other in the ass for 0.5fps in resolutions that were relevant 10 years ago

>>59271793
I have the same performance with a fucking intel 2400 and GTX 970 and 2x4GB of ram.

>>59281646
It has zero real world value though

Does anyone here even own a cpu or are we all just shitposting on some old ebay trash in our studio flats?

File: 34hju5gfk3i54.jpg (16KB, 480x360px) Image search: [Google]

16KB, 480x360px

>>59281692

>>59281471
>taking seriously 1 screenshot from debunked joker's benchmark
only on /g/

>>59281692
>c2d refurbished laptop
yup just for teh lulz

>>59281754

"Debunked" how?

File: amd fags and streaming.jpg (283KB, 2493x1005px) Image search: [Google]

283KB, 2493x1005px

>>59272991
Seems this same excuse is used by amd fags in every thread.

>>59281856
>desperately clinging on to a benchmark that is known to have dubious parameters

>>59281863
>lets interpret the numbers in % terms to skew the data into AMDs favor
>OH NO we have been proven wrong
>claim it's dubious
AMDfags trully have no depths they wont sink to, huh?

>>59281620
OLED monitors are not coming. They have image retention issues and Dell found color accuracy problems with theirs so cancelled its release.

>>59281879
There's literally no details on the streaming settings other than 3.5mbits/s bitrate. I would at least expect the 6900k be on a huge lead against the 7700k.

>>59281901
they are coming

>>59281908
No they are not unless they fix image retention (which is unlikely as LG TV's have all kinds of shit to prevent it happening). Maybe you are thinking of QLED.

File: 1488881569669.png (45KB, 630x424px) Image search: [Google]

45KB, 630x424px

>>59279642
>and thats not taking into account the 7700k stuttering.
uuuuuuhhhuuuuhhhhhh?

>>59281920
thinking oleds, tvs are already out, will be on monitors soon enough, or ill get a 40~ inch 4k oled, don't really care which. qled is doa as far as im concerned because its yet a fucking nother need backlight panels

I want my fucking blacks back.

>>59281879

nigga, we know nothing about the benchmark or how it was run.

for all we know it could be running at 720p where recording and streaming has no effect on performance.

also you faggy shills complain about amd users only cherrypicking benches, but you post that chart in every thread.

>>59281930

controlled tests always reveal every single problem.

https://www.google.no/webhp?sourceid=chrome-instant&rlz=1C1DLLB_enNO678NO678&ion=1&espv=2&ie=UTF-8#q=7700k+stuttering&*

>>59281685
you can argue OCing has zero real world value because not everyone gets same OC

>>59281944
Not true QLED. Shamsung are using the term on their new TV's but they are fucking lying. It's pure marketing bullshit. True QLED is self emitting.

Shamsung are lying cunts!

http://en.wikipedia.org/wiki/Quantum_dot_display

Also OLED TVs have rubbish response times compared to computer monitors.

>>59281930

Time spent above 47.1ms:
Ryzen 0
Intel 1

INTEL BTFO