JAYZ ON RYZEN

GET HYPED FOR ANOTHER MASSIVE DISAPPOINTMENT

i thought jayz was a nigger
what happened

>>59189755
See:
>>59189181

>>59189774
Also, DRAM is the true last level cache in Zen, and its absolutely fast enough for it

Ryzen will be shit and all the madcucks will commit sudoku

>>59189755
>Snake oil salesman who actually attempted to make an argument for AMDs failed APUs strikes again

Don't care, him and all his e-celeb friends will say whatever lines their pockets.

File: 1488232597795.png (66KB, 206x369px) Image search: [Google]

66KB, 206x369px

Shock?

File: ryzen_test_change - FISHY.png (65KB, 631x471px) Image search: [Google]

65KB, 631x471px

>>59189755

Pretty fishy

>>59189755
He meant: go back to /b/ you e-celeb shitter.

Watching Gamers Nexus' new episode seemed like Steve was kinda wary of the results garnered by his benches for common use. But who knows, really since he couldn't really expound.

>>59189840
Uh oh

>>59189765
No you idiot you're thinking of the engine

>>59189774
I don't know what any of that means aside from 60ns of latency

File: C52BpObWYAICp-Y.jpg (95KB, 1024x768px) Image search: [Google]

95KB, 1024x768px

>>59189840
>>59189914
well...

>>59190933
This is a feature on high end intel boards too. Its nothing new.

>>59190962
Hell, it came on my bargain bin Z170 too

>>59189765
He discovered /pol/

>>59189840
intel about to get BTFO'd really really hard

>>59189755
>e-celebs

>>59190933
Well that's fucked up

Literally rigging benchmark results

Explains why they do so well in CineBench and other synthetics, and then fail in real world tests

Reminder that youtubers are even below tech journalists in the grand scheme of things.
Intel trying to bribe all of them to benchmark 5 year old games at ultra low resolutions

>>59193524
lower resolutions tax the CPU more
and what are you talking about 5 year old games

>>59193728
iranian "review"
it got shreded to pieces though due to being on stupid ram, ES sample and ES motherboard

>>59193728
>lower resolutions tax the CPU more
please explain your reasoning, I'd love to understand that.

changing resolution won't affect the cpu performance in most cases, but lowering it certainly won't tax it more

>>59189755
boy howdy can't wait for these youtubers to test how it performs in Hitman, Crysis 3, Rise of the Tomb Raider, and GTA V. Those are the games I play!

>>59193790
Lowering the resolution yields a higher framerate meaning the CPU is taxed more

>>59193818
>AMD game
>ok
>recent
>still relevant
what games do you play

>>59193494
to be honest, Bulldozer was decent for games and pretty good for other things for the price. Even today you can play games at 60 fps with it

>>59193790
You are retarded, try it yourself. Lowering the resolution removes the GPU bottleneck, which means the CPU is tested more.

>>59193821
>>59193840
What?
Why?

A review using a game should use the settings most people would use,and I don't tend to run my shit at 800x600 because it's not 1998 anymore.

If they want to test raw CPU compute performance run synthetic benchmarks, that's literally the reason they exist. If you want to give a practical performance benchmark run the fucking game at 1920x1080.

This is the dumbest shit I've ever heard. They're just ignoring the point of a benchmark on a game. If you run with the same GPU the difference, if any, will be marked by the CPU.

>>59193818
>AMD sponsored
>Low CPU usage
>Medium CPU usage
>High CPU usage
What's the issue here?

File: lol2.gif (195KB, 200x150px) Image search: [Google]

195KB, 200x150px

>>59193839
>Bulldozer was decent for games

>>59193862

Its almost as if production code is different than code meant to artificially stress the system....

>>59191002
Prove it.

>>59193862

moron

>>59193862
Read some more CPU reviews, this is common practice

It isolates the CPU gaming performance.

If you don't do this, all you are getting is a GPU-limited benchmark.

>>59193862
Calm down
It's the way CPU benchmarks have been done for a while. Running lower resolutions shows CPU bottlenecks in gaming performance, not synthetic scores

File: William-Intel-Director-of-Creative_Innovation-240x200.png (68KB, 240x200px) Image search: [Google]

68KB, 240x200px

>>59189755
>Jay Z

There's only one rapper who's tech opinions I care about.

>>59193862
No it shouldn't we went over this shit after the shitdozer release

>>59193862
Your a fucking idiot or a troll

Synthetics test nothing, just made up shit


They've literally been testing in low rez for cpu performance since the mid 90s


U a straight busta me n my boys gone ride and u gonna get clappd CLACK-CLACK-CLACK

>>59193887
My point is, the test doesn't make any sense then. If a game running at 1080p, which is a minimum resolution for anyone gaming I'd say, doesn't stress the CPU enough for it to be a bottleneck, then what's the point of using that to test the CPU?

Again, the testing methodology is just dumb. The reason I look at game benchmarks is because I want to know how well the game would run if I had a specific part. If it ran the same way at a realistic resolution as another CPU, then the information being given to me isn't very useful.

>>59193928
It let's you know how the CPU performs in a non synthetic scenario in gaming

>>59193928
Is this the first CPU review you've ever read?

You are seriously new and need to educate yourself.

>>59193866
Intel was better for games, sure. But Bulldozer was good enough. Again, those processors still run games at 60 fps today if you pair them up with a good graphic card. That's 6+ years of good service

File: 1488335165972.jpg (66KB, 568x612px) Image search: [Google]

66KB, 568x612px

>>59193969
>But Bulldozer was good enough.

The everlasting cry of the AMDtard

"it's good enough!"

>>59193981
Oh well, you're trolling. I though we were talking.

>>59193839
>Bulldozer was decent for games

Even after all these years we still have AMDrones full in denial mode.

>>59193955
Well then just transcode a standard video file and give me a time comparison. That's a non-synthetic, CPU bound metric. That's a realistic scenario of something CPU bound as well.

My point is that if a CPU just does not make a difference in a game unless you're running at 800x600@700FPS, then it's not really realistic to say that one is better than the other because of this. However, if CPU A transcodes a standard video file 10 seconds faster than CPU B, that's a useful fact that I could take into account when making a purchase.

>>59194015
Good for you

>>59194008
Trolling? It's the truth, you're in denial. Bulldozer was good enough for games like a McDonald's happy meal is good enough for your health.

>>59194014
Dude. I bought an Intel cpu back then because i like to play at +90 fps. But what i say it's true; check the games benchmarks, those processors still play games at 60 fps. Enough for more than 90% of people out there.

>>59194015
How is transcoding helpful when determining gaming performance? Are you trolling? This is literally how CPU gaming benchmarks are performed, you lower the resolution to make a CPU bottleneck. This isnt controversial at all.

>>59193887
Gaming at 1280x720 is just as synthetic as as any other synthetic benchmark.
I don't want to run Skyrim at 250fps or 300fps. I want to know how processor choice influences current games at WQHD or 4K

File: 1445297301531.jpg (78KB, 816x840px) Image search: [Google]

78KB, 816x840px

>>59194059
>gaming at a low resolution is synthetic

>>59194059
4k cpu benchmarks are worthless since your GPU will be the bottleneck in 99% of games.

>ITT: synthetic benchmarks are mistaken as real world performance and low resolution gaming is unrealistic and useless

>>59194015
You just don't get it.

>>59194059
You are seriously delusional and have no idea what you're talking about. Go read some more CPU reviews to understand how it's done.

>>59194059
>I want a CPU benchmark where every single benchmark has the exact same FPS no matter what CPU is being tested

Be retarded somewhere else please

>>59194081
Games that are limited by the GPU anyway are not relevant to my CPU buying decision.
There are multiple current games that are influenced by CPU choice even at 4k (see: Battlefield 1). Those matter.

>>59194121
Wrong.

>>59194057
A transcode obviously isn't a measure of gaming performance, that wasn't my point.

What I'm saying is that if you have to increase framerates to 300fps to start seeing a divide between CPUs, then the metric isn't realistic. What was the last time you ran a modern game at 300fps at 1080p? I'm willing to bet never. If CPUs aren't a factor while gaming unless you're pushing retarded framerates at frankly unplayable resolutions, then gaming isn't a very good CPU benchmark.

>>59194095
You're right, I don't understand why playing a game at a resolution no one runs is a viable method of reporting to a consumer what they should buy.

>>59194143
>yet listens to synthetic benchmarks that are not indicative of real performance at all
>finds scores in benchmarks a viable method of reporting to a consumer

>>59194143
Go back to /v/, holy fucking shit.

>>59194143
You are goofiest idiot I have seen here in along time. Thanks for the chuckles.

>>59194143
playing it at sub normal resolutions is only good if you show a bunch of resolutions.

So like 1280x720, 1600x900, 1920x1080, 2560x1440, 3840x2160 just to show at what point the it's GPU bound vs CPU bound.

>>59194143
Are you new to PC gaming? Because this is how CPU benchmarks have been performed since the advent of discrete graphics accelerators. I find it bizarre that you would think this is Intel paying reviewers.

>>59189755

That testing PCs is beyond his comprehension.

But then he owns a 350Z... He coulda had a V8.

>>59194143
>then the metric isn't realistic.
OF COURSE IT'S NOT REALISTIC
THE IMPORTANT THING IS THAT THE METRIC IS USEFUL, AND GPU LIMITED BENCHMARKS TEST THE GPU, NOT THE CPU.
WHICH FOR A CPU BENCHMARK MAKES THEM EXTREMELY UNUSEFUL.

IS YOUR BRAIN FULL OF CHEESE

DO YOU HAVE PROBLEMS FIGURING OUT WHERE TO PUT YOUR HEAD WHEN WEARING CLOTHES?

>>59194168
He is seriously out of his mind. This is probably the first CPU review he has ever read.

>>59194168
I don't even care about the Intel v AMD thing, my point it that even if this was the way they've been doing it for 100 years I think it's not very logical nowadays.

I'm not saying they should use a regular resolution for testing while gaming, I'm saying that they shouldn't use gaming as a test at all. If by your own admission running a game at, say 1080p, doesn't push enough frames to stress the CPU to the point where it makes a difference, then why force the situation into something that isn't realistic in order to stress the CPU?

Just use a different test. Do a rendering workload, or an encoding workload. Something that stresses the CPU enough so that there's a marked difference between different products, but is still a realistic scenario someone in the real world would do.

>>59194207
You don't know anything about CPUs in general, so why would anyone care what you think is logical? Your logic is retarded, that's why it doesn't make any sense to you.

File: zm8n5fp3loiy.jpg (288KB, 1920x1280px) Image search: [Google]

288KB, 1920x1280px

Does anyone know when exactly the reviewers like Linus will publish decent benchmarks?

I use After Effects and Premiere, need to see the benchmarks before selling my 5960x system to build a new 1800x system.

>>59194207
Please stop before you hurt yourself

File: full-retard.jpg (80KB, 585x503px) Image search: [Google]

80KB, 585x503px

>>59194207
>I'm saying that they shouldn't use gaming as a test at all.

Congratulations, you've now gone full retard.

>>59194220
>watching linus reviews
HAHAHAHHAHAHAHAHA

>>59194220
I heard tomorrow 9am

>>59194220
>Linus
half the hardware industry is on his sponsor list. He has never published an unbiased review

>>59194220
>Linus
>ever

>>59194220
I have it on good authority that it rendered a 320x240 video CD in 38 seconds.

>>59194220
while the 1800x is better than the 5960x in general I would really question how it will do with adobe
AMD CPUs are historically bad with the Adobe suite

>>59194254
>>59194263
I'm more waiting for Paul's Hardware (and maybe JayzSeveralCents) also did Kyle get one?

>>59194280
Yeah does Linus even do technical reviews?

Personally I'm waiting on techpowerup and anandtech.

>>59194272
Exactly why I'm impatiently waiting for benchmarks!

>>59194269
>>59194263
>>59194254
Well, fuck Linus. I just gave him as an example.

Out of curiosity though, which source is best for me to check tomorrow for the most unbiased reviews?

>>59194280
Why do you watch any of those shitty channels
literally no redeeming qualities about any of them
they consistently break hardware and fuck up all the time, literal label readers

>>59194207
You seem to be completely at odds with logic. I state this seriously. It was fun and a good laugh at first, but now I am concerned for your wellbeing.

>>59194287
>I have never watched Paul's Hardware

>>59194284
Oh and guru3d is usually pretty reliable

>>59194286
find the average from multiple sources to avoid bias

>>59194292
Paul is a moron as well

>>59194300
Could be... but I'll still trust him more than most of the others

Man's got integrity

>>59194286
guru3d or some germans if you can read that

Intel fired whole PR department in Europe except UK, so everyone from EU except UK is fine

>>59189755
>JayzTwoCents
He somehow manages to be worse than Linus

Official list of people/sites you can kinda maybe trust tomorrow:

Steve Burke - GamersNexus
Paul's Hardware
Anandtech
Tech City (Aussie dude who isn't that technical but is honest as fuck)

People to avoid:
Linus
Jay
[H]ardforums

Feel free to agree/disagree or add more.

>>59194348
He is just linus in 10 years when children grow up a bit.

>>59194272
>historically bad
>x was bad once so it will always be bad
>you cant change anything
So wouldnt this then be a reason to change it with a new product?

File: 1421136595643.jpg (222KB, 852x939px) Image search: [Google]

222KB, 852x939px

>>59194350
add to avoid:
toms hardware

>>59194350
GN is surprisingly not bad, I don't like their methodology for games(buy FCAT already, god) but they seem honest

>>59194300

The thing with Paul, is he isn't afraid to admit he was wrong. He documents his fuck ups in his builds, and says "this shit is hard".

Watch his hotbox video and see how he made a shitty loop, wasn't happy, and went back at it.

He's relatable, has no fucking ego, and doesn't stir the drama lot like Eugene and Linus and Jayz. That alone makes him one of the best on youtube.

>>59194350
Meant the hacks at HardOCP, not necessarily the forums even though the forums are like an echochamber of reddit tier faggotry.

>>59194363
Agree on Tom's. They've been shit for about 6 years or so.

>>59189755
Oh wow here we go Intel wins again after AMD lied and schilled harder than a motherfucker. This reminds me of the Samsung controversy a few years back with synthetic benchmarks.

>>59194376
>HardOCP
did you miss the news? editor in chief appeared on stage AMD event, it was hilarious

>>59194350
>[H]ardforums

Not true, they are good. Their reviews are not as in depth as some others, but they are legit.

>>59194386
Kyle was on stage at an AMD event? Kek, would have liked to see that.

>>59194350
>Tech City
this kangaroo bro is pretty cool

>>59194350

Jury is still out on tech city for me. But I don't hate everything he does.

Gordon Mah Ung at PCWorld is becoming one of my favorite nerds on youtube. He's so San Francisco Bay Area it hurts though.

>>59194250
I agree that gaming is diverse workload that can be very hard on CPUs. It's certainly different from something like a rendering workload for example, and is an interesting case to look at. Not arguing against that. This is what I'm getting at, I'll present 3 cases:

1. I render a video at 1080p, 60FPS, 7500 kbps VBR.
That's a reasonable thing a media creator might do. It is CPU bound, and will show a clear difference between different CPUs in render time, all else being equal.

2. I bench a modern DX12 game at 1080p, and get around 60-70 fps stable.
This is a reasonable situation that most gamers would be familiar with. It is GPU bound mostly, but the CPU has a marginal influence. This test will not show any great difference between different CPUs.

3. I bench a modern DX12 game at 240p, and get around 600-700 fps stable.
This is a reasonable situation that most gamers will never encounter. It is CPU bound pretty much completely. This test will show appreciable differences between different CPUs.

Alright now, why would I chose to do 3, a test of CPU computing performance basically, when I could do 1 and get arguably more "pure" CPU benchmark? Add to that the fact that 3 is a situation no one will ever find themselves in, and as such isn't very useful to someone who wants to know if buying this CPU will make a difference to his gaming performance. Meanwhile, 2 shows a situation that an average person might find themselves in, but will show no appreciable difference in gaming performance. As such, gaming performance isn't greatly affected by CPU choice, within reason of course, at normal resolutions. Therefore, testing CPU performance this way is not useful for the person reading the review. Ergo, it's not necessary.

>>59194386

Praising VR, NOT AMD. He's still a sad sack of shit.

>>59194412
*for 3; an unreasonable situation

>>59194412
>gaming performance isn't greatly affected by CPU choice

This is just 100% false.

>>59194413
he actually praised Ryzen , kind of, for having and effect of 8 core on games already

>>59194398

yes, he even shook hands with roy taylor on stage.

>>59194426
Well of course if you're comparing a dual core i3 and a quad core i5 the difference is quite obvious. That being the case, if you're comparing these 2, then just run the game at a normal resolution and the difference will show. That's valid. I'm simply arguing against the stupidity of running a game at retardedly low res to compare CPU performance in an unrealistic scenario.

>>59194433

The AMD section of their forum exists solely to shitpost about AMD. Why do you think Juangra ended up there after he finally got banned from S|A?

>>59194350
I really liked Tech Citys video on that Asian mall, I wish I had a place to buy cheap older computer stuff around here.

>>59194452
Dude stop posting. Youre shitting up every Ryzen thread whining about people benchmark CPUs in gaming, it was funny at first but its getting annoying.

>>59194465
we have stores like that here in slavlands, I use them mostly to sell old stuff quickly

>>59194482
I'm literally only in this thread. I'm not an amd fanboy, I have intel shit on both my pcs. I just think the methodology for testing using gaming is dumb. No one asked you to read shit if you don't like it.

>>59194412
Just wow. Please stop.

>>59194502
Great argument.

I'm not even trying to be confrontational. If I'm wrong just tell me why that is.

>>59194453
Well there is a lot to shitpost about AMD

>>59194535

Like how ryzen won't break 3ghz and has sandy bridge IPC?

>>59194426
And you're 100% a child. The impact a CPU has on standard gaming performance is like the impact Mussolini's invasion of Greece had on WW2. It seems like it matters, but in truth Russia (the GPU) is what did the most of the work to give the end result.

And inb4 fringe benchmark where CPU matters but not really because the fps was over 60 anyway.

>>59194498
You have no understanding of the topic. Why on earth would you look at a cpu benchmark for actual real-world numbers to expect in your daily life?

You need to get a handle on what a benchmark is intended to present and then open your retarded mouth.

>ITT idiots who subscribe to retarded dogma just because it's always been the way it was always done
720p low is synthetic. It's as worthless as superpi.

>>59194554
Lowerinh the resolution makes it easier to tell the difference between CPUs
Do you even understand the voncept of benchmarks. How CPUs scale across games is interesting and worth noting even if GPUs are the primary bottleneck. Seriously, just go back to /v/.

>>59194514
Car a goes 0 to 100 in 11 seconds
Car b goes 0 to 100 in 10 seconds

Which car is better for you to get to work in?

>>59194014
I have a 6 core bulldozer, plays Darks Souls 3 and Battlefield 1 perfectly with an R290

Also Zen was designed by Jim Keller, not pooinloos

>>59194577
Nerd rage harder retard. I can see you you can barely type words now.

Keep believing in your indoctrination. You are only limiting yourself.

>>59194550
Holy crap you are an idiot.

Yes CPU performance has no impact on gaming, keep telling yourself that.

File: 1397599464304.gif (985KB, 157x157px) Image search: [Google]

985KB, 157x157px

>>59191413

by cheating

It is OK when AMD does it. R-right guys?

>>59194608
Did you type this and actually think it was a good idea to post?

>>59194607
No it wasn't you fool. Do some research.

>>59194585
the cheapest one

>>59194585
how does this relate to what he's talking about at all?

>>59194553
So a becnhmark isn't an indication of numbers I should expect in real life? Is that what you're saying?

Sure, if you want to know how good a CPU can perform when placed under a workload you'll never see irl, be my guest. I still think a render speed comparison would be more impressive but that's just my opinion.

If a reviewer wants to provide useful information however, then they should tell me what I can expect when I buy the product. That's their job. Sure, if they also want to tell me about how fast I can run a game at 240p, but that shouldn't be the main focus.

>>59194585
Why do you think car reviews usually do a little more than saying the 0-100 times? I'd rather buy a nice comfy mercedes that I can enjoy every day, instead of buying a stripped out racecar that is faster but I can't deal with day to day. The information most consumers care about regard their own use cases, not a ridiculous scenario no one would ever find themselves in.

>>59194550
Watch this video and then kill yourself

https://www.youtube.com/watch?v=mSU0UWFCa1Y

>>59194631
Maybe you'd learn something today. But then again, you don't seem very bright. I mean, it's almost like you never had an original thought in your life. I pity you.

>>59194660
You seem like a winner.

>>59194647
Seriously?

>>59194577
>using video games as benchmarks for this specific thing is dumb
>lel go back to /v/

it's like telling a christian to stop being dumb and go to a mosque, they're complete opposite stances you git. You people want to keep using games as benchmarks even when they are not very good at it. You go back to /v/

File: received_1878215459115456.png (430KB, 627x627px) Image search: [Google]

430KB, 627x627px

>>59194633
Are you retarded? They brought Keller back to do Zen, he also designed the original Athlon series which were ridiculously good

>>59194649
You have answered your own question from earlier. The gaming benchmark is but a part of the test. Is it important in the abstract? Not really, but it helps in providing a complete picture to the consumer.

>>59194673
you made no attempt to relate your subject to what he was saying, made no attempt at all to address his points directly, and have generally made an ass of yourself in this thread.

>>59194676
Did you even read the post? Games are good benchmarks of CPU performance once you remove the the GPU bottleneck, which is why they lower the resolution.

>>59194693
Well, not in my opinion. I think it's mostly useless. Here though, I can agree to disagree. At least you presented a counterargument. Congratulations.

>>59194679
And you are wrong. Keller was brought back after Zen was designed. Go read the actual amd info about the design team.

>>59194697
Retarded /v/ logic.

>>59194703
Well its a good thing review sites don't care about your opinion.

>>59194695
Not too observant are you?

>>59194712
And it's an unfortunate thing that you weren't able to able to accept my point. Still, good chat.

File: cpu-gaming-performance.png (46KB, 618x537px) Image search: [Google]

46KB, 618x537px

CPU HAS NO IMPACT ON GAMING PERFORMANCE!!!!!!11111111

>>59194709
Again, games are a good way to compare performance among CPUs once you remove the bottleneck. This isnt a new concept at all.

File: zen.png (8KB, 279x95px) Image search: [Google]

8KB, 279x95px

Amd already win for price/performance, intel shill are desperate

>>59194726
Well if you misrepresent the point sure; they start making a difference. Someone buying an i3 isn't cross-shopping a 6700k.

Also notice that the testing there is done at a reasonable resolution, so the differences shown there are actually useful to a consumer.

File: ryzen-doa.jpg (386KB, 1301x739px) Image search: [Google]

386KB, 1301x739px

>>59194751
Sadly they don't.

AMD is getting beat by Intel processors that cost $100 less.

>>59194759
Backtracking hard I see.

>>59194390
Their reviews are a inconsistent fuckmess, their GPU reviews try to gauge at what settings will certain equipment give playable framerates never keeping the same methodology for two reviews
And that's not even taking into account the butthurt of Kyle and his biasing

>>59194759
What this shows is that that gpu is poor value for a person with an i3 cpu and they should consider a new cpu or a gpu that will not outpace the i3

>>59194735
Friend, almost anything can be a good way to compare performance among CPUs if you skew it enough.

The issue is that creating an unrealistic scenario in which the CPU is stressed and using that to compare different products is basically the same as a synthetic benchmark. It's not applicable to the real world. It's a neat fact, but it's not all that useful

>>59194790
100% false. See >>59194726

File: 1488403807086.png (99KB, 650x380px) Image search: [Google]

99KB, 650x380px

>>59194767
>using shit ram

>>59194788
Definitely agree on that. The test is run at a normal resolution and as such is a realistic depiction of what a customer can expect.

Again, I don't have an issue with using games to compare CPUs in this way. My issue is with the comparisons featuring the games ran at 240p, which aren't realistic

>>59194795
The test you refer to is a good comparison of CPU performance because it is run at a normal resolution. My issue is, once again, with the folks running these tests at ridiculously low res and claiming it is a useful test.

Whatever gayming benchmarks all the reviewers are running I hope they at least have the sense to test the r7s both with and without hyperthreading in every gaym considering there are so few that can utilize more than 8 threads in any scenario.

>>59194821
There is no difference between running at a lower resolution vs. using a high-end GPU - both of them reduce the GPU bottleneck.

CPU performance obviously has an impact on gaming and to deny it is lunacy.

>>59194726
>most games are over 60 with some going up to 120+ fps on the i3
What a great way to make your point senpai

>>59194795
How many people who buy a $700+ GPU and a $300+ CPU play games in FullHD?
Makes more sense to benchmark high end CPUs at resolutions people actually use.

>>59194790
You really don't understand the concept of a benchmark, do you?

>>59194841
Unless the reviewers run the first Win10 build or something old as fuck there shouldn't be any problems between having SMT enabled and disabled, unless they're OC'ing

>>59194846
1. That's average FPS. They will dip below 60 many times which is unacceptable for those who want smooth performance.

2. 144Hz monitors do exist and even the highest end processors are not hitting stable 144fps

>>59189755

You're about to be shocked by sandy bridge levels of single thread performance in 2017.

>>59194855
Those games aren't even hitting 144fps, which means they can't even max out a 1080p monitor at 144hz

>>59194871
Shouldn't you be in your designated shitting street?

>>59194855
We now have 240hz monitors yet an i7+Titan XP cant even hit 144fps at 1080p maxed out. Yeah, those numbers are relevant.

>>59194845
different CPUsand games have different effects though.
Skyrim @ 1280x720 is a test of single core performance.
DOOM @ WQHD or 4k is a test of multicore performance
the latter is actually relevant to real world usage. The former might as well be a synthetic benchmark

>>59194887
>Skyrim @ 1280x720 is a test of single core performance.
>DOOM @ WQHD or 4k is a test of multicore performance

You have no clue what you're talking about

>>59194845
Again, yes CPU performance makes a difference. The important thing you're missing is that running a very low res and using a high end GPU are not the same. They are similar, but not equal. If I wanted to I could use a Titan X with an i3 and have a reasonable gaming experience. I can't however have a reasonable gaming experience if my resolution is retardedly low. That's my issue.

>>59194861
If you want a figure that allows you to rank CPUs , I'd rather have that figure be based on something I might do in real life. Going back to that car analogy, 0-100 times are quoted as benchmarks because it's a common scenario someone with a car might encounter. Running a game at 240p is not a common scenario

>>59194901
>The important thing you're missing is that running a very low res and using a high end GPU are not the same.

For benchmarking a CPU, they are.

All you're trying to do is reduce the GPU bottleneck, it does the same thing.

>>59194887
This is nonsensical.

If you can get a video card to render the 4k without being the bottleneck, both tests will have a similar value in determining which cpu is the winner in that test. No more, no less.

>>59194926
Similar method, different results. Can you say with 100% certainty that running a game at 240p with a 1060 is equivalent to running a game at 1080p with a Titan X?

Probably not. If so, then why go with a method which has unlikely circumstances (240p when most monitors nowadays start at 1080p), instead of one whose circumstances at least seem possible? You're introducing an unknown factor into this, and it's just making shit more complicated for no reason

>>59194901
You know, ignore the specific numbers and just look at the percentage delta of the competitors. Forget about the actual test parameters. Get a few of these tests together, decide what price you can afford, and make a decision.

>>59189761
GET HYPED FOR A MAGNIFICENT DEAL

>>59194962
A for instance. I need the best cpu for heavy asio usage. Most benchmarks don't specifically target this metric, but I know from experience which tests simulate it.

You see...???

File: gapestation.jpg (8KB, 184x184px) Image search: [Google]

8KB, 184x184px

>>59194706

>>59195006
Alright, now we're getting somewhere.

Your use case is a good example of why there should be a more useful metric for CPU performance. Most of the sites reviewing stuff are gaming focused, with a few exceptions. I'm advocating for providing a wide variety of useful, realistic metrics for measuring CPU performance. Whether it be gaming at full resolution, rendering video files, decryption work, compression workloads, etc...

From this bunch of tests, you can gleam the information which is important to you, as well as finding out how good the CPU is at completing certain tasks which are more common.

>>59194706
>In August 2012, Jim Keller returned to AMD, where his primary task was to design a new generation microarchitecture[4][10][14] called Zen.[13] After years of being unable to compete with Intel in the high-end CPU market, the new generation of Zen processors is hoped to restore AMD's position in the high-end x86-64 processor market.[2][12]

>>59194962
Stop being retarded. Eliminating a GPU bottleneck means you can easily determine how much faster on CPU is at a typically single-threaded application like gaming without running into margin of error problems. Just because no one plays at 720p anymore doesn't make the numbers useless, you can still see percentage increases in performance.

>>59195072
And as clearly demonstrated earlier by this guy: (>>59194726) that can also happen at 1080p. SO WHY NOT JUST RUN THEM AT 1080p? You're not making the test any worse really, and now you're giving me numbers that are actually useful.

>>59194046
>doing stupid analogies
As a offended I'm intelfag citizen.

>>59194800
There was a BIOS update for increased memory bandwidth as well.

>>59194075
It also isn't real world, because nobody will play a AAA game in 80 columns.
Then what is it?

>>59189755
He meant that it's a good idea to make money on retarded gaymers.

File: 1452454852899.jpg (86KB, 487x460px) Image search: [Google]

86KB, 487x460px

>>59189755
Intel is still the king of laptop CPU's
>>59195133
>>59195133
>>59195133

>>59189761
This.

/g/ had learned nothing from the past.

>>59195247
The word you were looking for is "has" my dear Pajeet. "Had" makes no sense in the context of that sentence. Ask your tutor about it during your next English lesson.

>>59195275
aren't you suppose to sleep, cucklifornian?

>>59195220
and that's because amd didn't release zen cpus for laptops yet.
this argument is so dumb it hurts

>>59195286
"Supposed" to sleep, Sanjay. These past/present participles seem to be a real weakness.

>>59194350
paul is a cuck i dont trust cucks

>>59194350
The people who bought it (consumers) that benchmark it with their rigs to give more accurate results with varied hardware (non-optimal hardware/situation).

desu

>>59195247
It's standard AMD-shitposting, it's a /g/ tradition.

We all pretend AMD will wipe the floor with Intel until the release, and then afterwards we all laugh at our previous selves for thinking AMD could ever hope to compete.

>>59189840
Probably last minute BIOS updates to.fix something buggy, if it was anything else I think AMD would have prepared reviewers in advance.

>>59195354
Too bad Intel is infinitely shittier

File: sanic.png (96KB, 900x636px) Image search: [Google]

96KB, 900x636px

>>59190898

>>59195390
B-but it's good, goyim. Intel is best. Buy Intel.

>>59194220
>selling my 5960x
Why would you sell your 5960x? A 1800X is going to to be Maybe 3-5% faster at stock, just overclock if you need more performance.

>>59194220
Linus dropped the motherboard and bricked it. So maybe in a week.

>>59195494
Is he a literal retard?

>>59194350
Anandtech has been so lazy lately.

They're fairly unbiased and testing methodologies are good, but they're lazy as shit and hardly do any work now days.

File: muh watts.png (66KB, 650x363px) Image search: [Google]

66KB, 650x363px

>>59195411
You forgot to mention that it's also going to use 100W+ less power. Intel and Nvidia users have been claiming that power efficiency is extremely important for years, so you'll all be upgrading to Ryzen to enjoy that massively-superior perf/watt ratio, yes?

>>59195512
Yes. He breaks things all the time.

>>59195527
Holy shit

>>59195527
INTEL HOUSEFIRES (hopefully this meme catches on)

>>59195589
Dude p4 existed lmao.

>>59195589
>hopefully this meme catches on
spirals of history

File: Screenshot from 2016-10-28 18-59-04.png (1MB, 1920x1080px) Image search: [Google]

1MB, 1920x1080px

>>59194143
Testing CPUs in a GPU limited scenario is like trying to determine which car has the highest top speed on a 70MPH highway, and not allowing the cars to exceed the speed limit.

Would you also stipulate that GPUs should be tested with Vsync on capped at 60Hz, when they are all getting >60FPS in games tested? In a case like this it would be easy to compare a RX460 and a 1080Ti and determine that the 460 is the same speed, and that the 1080Ti is just massively overpriced, costing 5x the price for 1x the performance.

>>59195527
Uh that's only a 15W power difference, while the 7700K is both cheaper and faster.

>>59195733
The 1700x has better perf/watt tho :>)

File: worthlessfuckingshills.png (535KB, 2560x1080px) Image search: [Google]

535KB, 2560x1080px

>>59194350
Avoid Hardware Unboxed at all cost.

>>59195733
DOUBLE THE CORES
LESS POWER
>>>>OVERPERFORMS

>>59195773
But the graph is right. 1070 is better than Fury X in most titles.

>>59195773
why do those look around 10 percentage points off?

How'd they make the Fury X perform so bad?
That looks more like Fury/Nano vs 1070 performance.

>>59195790
Maybe, but those numbers are waaaaayyyy off compared to http://www.techspot.com/review/1182-nvidia-geforce-gtx-1070/page2.html

>>59195826
Fury series looks pretty decent after they fixed shitty drivers.

>>59193839
LOL what?
bulldozers biggest weakness was games due to absolute shit IPC

>>59195858
Yep. You can get an R9 Fury for $235 that's 80-95% the performance of the 1070

And that benchmark isn't even doing DX12/Vulkan

>>59189755
If Ryzen is shit I'll finally build an Intel machine

>>59195733
The question was in relation to the 5960X, you fucking retard. Nobody cares about corelets and their childrens' games.

>>59195963
Corelets, when will they learn?

>>59193862
Because in each frame the cpu has to do a lot of calculations for physics, ai, etc.

You fucking idiot

>>59195963
Look at the chart you posted yourself. The 7700k is cheaper, faster, and only 15W more power.

>>59196015
Corelets, when will they learn?

Official list of people/sites you can kinda maybe trust tomorrow:

Steve Burke - GamersNexus
Paul's Hardware
Anandtech
Tech City (Aussie dude who isn't that technical but is honest as fuck)


People to avoid:
Linus
Jay
[H]ardforums
HardwareUnboxed
Tom's

Maybes:
PCWorld
Anandtech

Feel free to agree/disagree or add more.

>>59195071
You left out the important bits. He was brought back after the design of the Zen was done. His job was to add elements to the design that we will end up seeing in the mobile variations, etc.

Nuance... it's handy to understand it.

File: Interesting.png (98KB, 887x499px) Image search: [Google]

98KB, 887x499px

>>59196093
Guru3D

Hilbert is a bro.

BIG PIMPIN

>>59196143
Damn Intel is really playing multi-dimensional chess. Tbh they aren't wrong.

People who own 4c/8t Intel CPUs have literally no reason to upgrade to Ryzen, people who own Ryzen will have no reason to upgrade for years to come. It really is a dead market where people who have 5 years old hardware are good to go for next 5 years to come.

>>59196143
We would all do ourselves a favor and look at Intel's recent acquisitions.

Also, will AMD be able to unload the old ATI division before it sinks them?

>>59196236
The graphics division is going to help them dominate a large portion of the server market in a few years.

>>59196178
>It really is a dead market where people who have 5 years old hardware are good to go for next 5 years to come.

The market is dead because Intel basically had no competition and have been rehashing for years.

>>59189755
PROCESSOR DRAMA

WHICH DOES THE BEST PROCESS?

>>59196302
I didn't say graphics division.

I still don't feel the need to upgrade from my 4770K with lousy overclocked DDR3 @ 2100. Especially with memory prices as they currently are.

when is this freaking NDA ending?

>>59196446
In about 4 hours.

>>59189755
buckaroo

>>59194286
GamerNexus

>>59189840
ryzen sucks with high speed DDR4 RAM

>>59196752

1800x review copies were sent with 3000MHz DDR4 RAM you fucking nigger

>>59189755
DOA

Hype was a meme

ETC

>>59195773

>Ignore test results I don't like. . .

>>59196752

> Running DDR4 above 2133 MHz

: DDDDD ebin!

>We think Ryzen is a great gaming CPU, and you’ll test that for yourself – we’re not going to win every head-to-head, but if you think about gaming do you want theoretical performance or do you want the CPU to be good enough to showcase your GPU? I think what Ryzen allows is those folks to do something more than just gaming. So your gaming CPU might only uses four cores, but if you are doing video editing or streaming it will do a lot more. So I think what we’re trying to address is maybe the forward thinking users, not just the today gamer.

So Lisa Su pretty much says already that Ryzen won't win 1 vs 1 against Intel, but the amount of cores / price should give it more value than intels 4c/8t processors

>>59197205
It's nice to have if you don't care about ECC

>>59197221
That's the point also, I wouldn't be surprised if they fix some BIOS stuff. The thing about a consumer is that it should always be price and performance, not just high prices.

>>59197221
Having eight real cores will certainly help if you're streaming.

>>59197221
Good thing I am not a fucking gameshitter underage tard

>>59197247
Yeah. I usually have 2 screens on, so I might play some grand strategy on another and do work/study with other. On the other hand my 3570K doesn't seem to be struggling with this, but we'll see. I honestly think the biggest obstacle for cpu+motherboard upgrade is the extremely high ddr4 price

>>59197252
I play games here and there (5-10 hours a week) and I don't care about 5 fps difference. I run VMs and unzip a lot which means the 7700k is not even an option for me. I don't understand these gayming numale kiddies. They're willing to give up 4 cores and 8 threads for muh 3 fps, plus they're paying almost the same price for a much weaker CPU.

>>59197289
>I honestly think the biggest obstacle for cpu+motherboard upgrade is the extremely high ddr4 price
what

>>59197326
RAM prices are starting to rise and the only RAM he probably has are the DDR3 he's using for his 3570k

>>59197322
Gotta get that optimum FPS if gaming is only thing you care, which is true for probably majority of young males

>>59197322
>much weaker CPU
>I7 beats Ryzon in almost every benchmark available

>>59197392
>i7 beats Ryzen in SuperPi
ftfy

>>59197392
>I7 beats Ryzon in almost every benchmark available
Not in this timeline that's for sure.

>>59197441
>>59197447
>Almost competitive with Sandybridge
>AMD is superior!!

>>59197392
It only wins by a few % on st benchmarks, but it loses by up to 100% in mt benchmarks. You have to be braindead to buy a 7700k at its current price.

>>59194550
i have a fx6350 at 4.3 ghz and a rx 480
i can't reproduce any of the fps result in rx 480 benchmarks
guess why

>>59197495
>You have to be braindead to buy superior technology

I'm sorry you're poor Pajeet

>>59197495

Maybe I like to watch Netflix at 4K. Too bad AMD never thought of this.

>>59197701
>Maybe I like DRM. Too bad AMD never thought of this.

>>59194350
I'm waiting for Tech City's review. Brian always makes really good videos. He has just recently started taking things a lot more serious with mainstream parts since his wife left him last October. I feel bad for the poor guy.

>>59194295
Better find the mean.

>>59189774
shit means nothing when it cant beat intel with 4 fps difference.

literally unplayable!

>>59194767
>monster hunter
what?

>>59194767
3 resolutions tested, 1080p is highest of the 3.

What third-world shit is this?

>>59196843
>CONFIRMED 1800X CANNOT RUN HIGHER THAN 2666Hz RAM

Sorry what was that, fucking nigger?

File: lol.png (27KB, 687x694px) Image search: [Google]

27KB, 687x694px

>>59189755
I GUESS WE ALL KNOW WHAT HE MEANT BY THAT NOW

AHAHAHAHAHAHAHAHAHAHAHAHAHA

AAAAAHHHHAHHHAAHHAAAHAH

>>59204114
Try again, kid
http://www.fudzilla.com/news/processors/42986-amd-debunks-ryzen-ddr4-ram-issues

File: 0826 - Tbd1nKx.jpg (8KB, 200x200px) Image search: [Google]

8KB, 200x200px

>>59204140
>fudzilla
HAHAHAHAHAHAHAHA

>>59189755
https://www.youtube.com/watch?v=BXVIPo_qbc4

We got all those graphs with numbers showing Ryzen behind everything but when I see it in action it trade blows with intel price counterpart pretty fine.

https://youtu.be/9AVZ_x64hg4

>>59204168
It has a tweet by AMD with ryzen + 3400Mhz DDR4. But we didnt saw it in any review that is out. Most of them got 2133, some 2400 and the max was from Gamers Nexus with 2933

>>59204114

https://www.youtube.com/watch?v=GA7x_ftav9Q&t=4m26s

>>59196093
>Maybe
>PCWorld

Gordon is the GOAT

>>59197221
Except Doctor Su knows that not everyone is video editing and/or streaming on Twitch.

>>59202340
Literally iran

>>59204170
https://youtu.be/sciuiEcrnzg

File: bingo.png (63KB, 874x771px) Image search: [Google]

63KB, 874x771px

>>59197223
Ryzen supports ECC, just need a Mobo that does that, but it seens there is none.

>>59204454
>board partners fuck up
(lol

Now watch Asus charge $499 for """workstation""" grade features like ECC, more SATA ports, etc.

File: 1487801576360.png (205KB, 583x612px) Image search: [Google]

205KB, 583x612px

>>59197554

>>59193790
>CPU writes instructions, sends them to gpu
>CPU writes next instructions, but has to wait now because GPU is completing previous instructions.
>lower resolution means less work for GPU, but about the same work for CPU since the CPU doesn't work on a per-pixel basis.
>thus, a lower resolution will be more taxing on CPU since its waiting for the GPU less often.

>>59204424
Is this the most boring review video ever? That guy is like a bad salesman.

inb4 shills review gaming performance with bulldozer-optimized drivers and then claim ryzen sucks for gaming.

>>59196178
The problem I think, is a bit more complicated.
Even after they were pretty handily beating AMD they kept improving their architecture and getting more and more ahead. Now it seems like they really have hit a wall in terms of performance (evidenced by how IPC-wise even Ryzen isn't really there, and how similar this new architecture is to Intel's overall). So Intel could dump billions into R&D and try to find the next big thing, or since they already had a huge leg up, they could just keep refining the formula they have.

At this point to get much better than Intel's x86 CPUs it may be that we have to move to an entirely new architecture, and nothing else is showing promise so far as its successor.

With AMD finally catching up though we may finally see Intel start dumping money into finding it now.

>>59204615
This is his mojo.

>>59204633
They can't into 10nm yet and GloFo, SMTC and Samsung are alreadly moving to 7nm. Unless they give up on the tri-gate finFET and go for something else, they will be stuck at it. Worse part is, they reduced their workforce in late years, they're spending way less in R&D and basically rehashing cpus. Intel got sloppy.

>>59194175
>testing something that is completely retarded and not something people would ever do is useful, but an actual workload that people may want a CPU for (but isnt exactly gaming) is pointless because its not gaming.

>>59204695
moving to a smaller process isn't magically going to boost performance, just lower power requirements and TDP. I know they've been rehashing, but I have a feeling with Ryzen here and already being so close to Intel's best in its first iteration, it's going to light a fire under their ass to go on a hiring spree to find some fresh R&D talent. Otherwise AMD is probably going to catch up real quick with Ryzen's successors.

>>59193839
Vishera was good. Bulldozer was definitely shit.

>>59194291
>is too stupid to understand why they benchmark
>yells autistically at someone who asks why they bother
>"lol your odds with logic bro"

just admit you're retarded and dont know why they do it either.

File: 1479618237612.png (325KB, 522x593px) Image search: [Google]

325KB, 522x593px

>>59204615
It was boring as fuck but I couldn't look away.

It was like walking into a time-share presentation.

I don't think it was fair that he didn't compare 1080p BF1 performance. But the fact that the 1800x gets higher minimum framerate in the majority of games, at least that he said he tested but didn't include them all and is saving for a two hour long video, at 1440p when GPU bottlenecked still seems like a really big deal.
I'm always more interested in the higher minimum framerate over average.

>>59204633
Nope all AMD has to do is keep doing what they're doing.
Intel won't have 7nm until 2021-2023 while AMD will be on it in 2019. Possibly even early 2019.

If they improve their virtualization and other things for servers, Intel is going to get fucked.

File: 1440946449347.gif (904KB, 427x240px) Image search: [Google]

904KB, 427x240px

>>59189755
>theire

>>59204878
>all AMD has to do is keep doing what they're doing.
What they're doing is making an architecture which is similar to Intel's Core architecture. If intel has hit the wall with Core, then chances are AMD isn't going to be the one to break through. Maybe there is some untapped potential in the architecture, but all signs point to it being some new architecture that will be our next big performance boom in the PC space.

I'd love to be proven right or wrong, so whichever way it goes is great, long as it goes.

>>59204863
Going down nodes will give you less energy usage and more room for transistors, what I see is that Core uArch is done, there is no more room for improvements, and AMD just got at them at first strike, it is time to Intel move to its next uArch, wich will take time and while they do this, AMD will be refining Zen.

File: 1469404476527.gif (846KB, 245x194px) Image search: [Google]

846KB, 245x194px

>>59193981
>"it's good enough!"
This is hypocritical at best. Then tell me why the argument on most of the AMD bashing threads is:
>Stick with your old intel 4 core processor. It's good enough.
>If you're only gaming get a pentium instead. It's good enough.

>>59193981
(you)

>>59194767
>1fps/nofps difference at 1080p
>like 5 fps difference at third world resolutions

LOL AMD IS FINISHED. HOW WILL I ENJOY MY NEW 500 DOLLAR CPU ON MY OLD 20 DOLLAR CRT?

>>59204952
Except it's like 1.5x more power efficient under 3.3ghz.

Put that on 7nm, 2 years before Intel gets to 7nm, and they'll have trouble selling a single CPU to servers which is their largest market.

>>59204615
I enjoy boring but informative videos like this