So Ryzen is fucking awesome after all?

So Ryzen is fucking awesome after all?

Other urls found in this thread:

youtube.com/watch?v=nsDjx-tW_WQ
cpubenchmark.net/compare-test.php?cmp[]=2970&cmp[]=1190
arstechnica.com/gadgets/2017/03/amd-ryzen-review/
anandtech.com/show/11170/the-amd-zen-and-ryzen-7-review-a-deep-dive-on-1800x-1700x-and-1700/23
translate.google.co.uk/translate?hl=en&sl=de&u=https://www.golem.de/news/ryzen-7-1800x-im-test-amd-ist-endlich-zurueck-1703-125996-4.html&prev=search
cpubenchmark.net/compare-test.php?cmp[]=2970&cmp[]=2919
twitter.com/SFWRedditImages

>that's not even overclocked
A 4.5 GHz R5 is going to destroy intel

I'm going to buy one of those but come on, it ain't gonna OC to 4.5. It's gonna go to 4.2 like the 1800X did

It's inconsistent as fuck, like most AMD launches.

youtube.com/watch?v=nsDjx-tW_WQ
>new results from same reviewer
>retard was GPU bottlenecked in old results
>ryzen btfo

They're very similar but Ryzen's load is way lower than Intel's, that's mean Intel chips are already working at maximum capacity.

that's pretty shit

seems like a very bad launch, but looking at the figures, ryzen is clearly way ahead

I don't understand this argument. On the one hand people are saying Ryzen is shit because it sucks at 720 and 1080. On the other hand it performs well and actually competes with Intel, when relying on the GPU. Wouldn't this be considered an advantage?

The fact that they're not CPU bottlenecked is good either way you look at it, faglet.

In computing one component that's utilized 100% means everything else in the system ends up waiting on that component, so they dont get stressed and it doesn't measure how they perform under stress. That's my best eli5

so you really created a thread to say that an 8 core processor beats a 4 core processor in multi threaded usage. christ you're a fucking genius aren't you.

okay, I see more drops on 7700 side
but average is almost the same
why are you so hype user, nothing changed

They in the same price bracket, that is why they are being compared.

it's on par with it in single threaded gaming applications too....genius.....

Different upgrade cycles, basically. If there are many software issues as some people have speculated then there is a chance that down the line more capability will be unlocked down the line. But as it stands you'll end up with a CPU bottleneck with Ryzen before you'd end up with one from Kaby Lake, but exclusively in games. It's still competitive with processing workloads against the 6900k for half the price.

In an ideal setup you have 1:1 scaling. 1440p/4k games are currently GPU bottlenecked, which means your scaling is gonna look like 1:4 CPU/GPU.

So when GPUs finally catch up to 4K scaling, and that ratio gets closer to 1:1, you will see the 7700k race ahead of the 1700/1800x in high resolutions.

it's not though is it, when it's overclocked to it max still behind the 4.9 - 5.1 7700k max. most benchmarks have it behind.

No it's really quite horrible

same price bracket different use case. you're comparing a golf ball to a baseball because they both cost the same.

>but exclusively in games.
But, I thought there have been numerous reviews in which Ryzen did well in many games, even shitty un-optimized ones. This is all very confusing.

>still getting 100fps
and you can't go over 105 fps in Fallout 4 before the shitty engine starts shitting itself.

> (OP)
>No it's really quite horrible
Depends how much cores game uses.
they just wrote gayming didn't even wrote which game.
fake news.
also it's cheaper

Oh, I see. Thanks, this seems to make sense.

>it performs really well when you're not using it

if it's not under load and you're relying on gpu then you're not testing the processor a better test would be removing the gpu all together.

>but my ryzen doesn't have igpu.

quoted wrong person

np mate.

Advantage or not, GPU bottleneck invalidates benchmarks.

>also it's cheaper

No it's way more expensive

A $340 Intel processor is beating the top of the line $500 Ryzen

buy 1700
overclock to 1800x
save money not having to buy intels jewish overpriced motherboards

>turns out the 1700 still performs pretty closely to the 7700k at 720p
You have an interesting definition of 'btfo'.

Even if you do that it still loses, 1700 is $330

intels motherboards cost more

It's pretty bad considering the 7700k still leads in GTA V, which does take advantage of 8c/16t.

So ryzen's 8c/16t in FULLY optimized games are still slower than i7 4c/8t. See performance in Tomb Raider and Sniper Elite which aren't 8c/16t optimized and the 7700k has a 40% to 100% boost in fps.

720p isn't a reliable benchmark so I have no idea why it's used so much. Scaling back rez is far from a 1:1 ipc cpu test as different architectures behave differently at different rez. 1080p is most common resolution and a good benchmark. higher rez's are good if gpu isn't limiting factor. majority of game benchmarks don't do justice to ryzen so this is all trivial right now. there will quickly be a round of patches pushed through windows and bios that rectify many issues, some already have begun. common sense dictates ryzen will be in the ballpark of 90% ipc compared to kaby lake. i believe computerbase benchmarks indicate this. you can see computerbase with update bench and 1800x within a few percent of 7700k on average at 1080p.

>AMDlet is still triggered by reviewers shitting on Ryzen's gaming performance so he keeps posting the same single benchmark posted by some Mexican Sup Forumsermin that has no idea wtf he is doing.

Cute.

Does anyone do frametimes comparing the CPUs?

You know the 7700K is overclocked to 5ghz there right.

But that makes the assumption that games five years from now when GPUs are commonly able to push >100 fps comfy on max settings will still be single threaded perfomrance limited.

A few did, and they clearly showed the core parking issue that's fixed with the BIOS update. To my knowledge none have been done on an updated board.

OK let's say Intel motherboards cost $50 more

i5-7600K is $240

Ryzen 1700 is $330

You still lose $40 buying Ryzen and you get even worse performance.

>When you eventually have a 1280 ti from nvidia in 3 years these cherry picked 480p benchmarks will matter
>When games use better API's in the future that can use more threads, the extra cores won't matter

ITT:

>people shit on AMD because it isn't top of the line
>comparing mid tier to top of the line

Well of course it's going to be shit in such a comparison you fucking retards. When compared to Intels mid range it does compete - albeit a mixed bag. This is what anyone with realistic expectations should expect. I for one am content with this, and will happily build my next rig using a Ryzen (because fuck Intel and fuck their scalping).

>When compared to Intels mid range it does compete

No it doesn't, see

And the 1700 is overclocked too (3.9 GHZ). So what's your point? The big takeaway is that in games FULLY OPTIMIZED for 8c/16t, 1700/1800x is still losing to the 7700k.

and furthermore a 1700 is right up there with 1800x at 1080p so it becomes clear 1700 at $300 is a clear winner. pic related is against a much more expensive part. as seen in previous post an 1800x is within a few percent of 7700k at 1080p.

why not get the 7700k :^)

Because most 7700Ks won't hit 5ghz, most cap out at 4.8ghz. the 1700 he has is low, others are getting it to 4.1ghz

>g-gimp your cpu because amd can't catch up
Really makes you think

Well you could and absolutely destroy the 1700

But you might spend more money if Intel motherboards cost more

what smoother frametime line tells us?
that CPU can't handle a GPU driver?
or that CPU can't handle part of game code?

my bet that GPU driver calls are multi threaded already it can handle driver better but gets stuck on game code a bit which will be fixed/adapted in newer games anyway
meaning it has higher longevity than 7700K which has much bigger frame time jumps and would bottleneck new GPU faster

explain to me why i'm wrong

Here we go. What's next? We don't need more than 60fps? The human eye can't see more than 15fps? AMD apologists should be dragged out and shot with their shit-eating consolecasual brethren.

And we're not looking at a CPU bottleneck in this video either. The CPU usage on a per core basis isn't 100% in most cases, yet the GPU isn't being pegged at all. So what we're seeing is how well these games are actually coded. Because even in most cases the 7700k isn't seeing max usage. It paints a kinda fucked up picture of game benchmarks when it comes to CPUs, if both systems were pegged at 100% on the proper cores (4 cores if the game is coded for it) it'd make sense. But neither CPU is really being stressed even at the single core level in most of these benchmarks.

It just kinda backs the idea that most PC games are just unoptimized piles of shit.

and before anyone spouts "gpu is limiting it!!111!!!1!" here is 720p, again, within a few percent of 7700k, keeping in mind lower rez =/= 1:1 ipc performance as there are other architectural considerations.

Who the fuck plays games at 720p????????????
If you're rich enough to afford a 7700K or Ryzen 1700 I think you can also afford a fucking monitor that is higher res than 1080p.

>others are getting it to 4.1ghz
"No!"

Are you retarded? Frame times look literally the same for both Ryzen and last 3 intel's gens. There is literally no difference.

That's actually a CPU bottleneck. Games don't tax 100% of the CPU resources.

most reviewers commented that in spite of a lower fps, ryzen plays smoother

Then why did some games see 99-100% usage and others didn't?

>6900k has lower clocks and only 2-3% higher IPC than Ryzen
>outperforms Ryzen by 20%

>FULLY optimized

fully optimized to deal with the various fine grain locks and critical sessions that are probably a-plenty in the code base?

i highly doubt the game is 100% linear scalable, corewise.

If your CPU is REALLY SHIT and the game is properly threaded it can sometimes utilize all resources. This only happens with weak CPUs though.

German friend, can you summarize the *updated* part of the article?

Ryzen proves what society is a bunch of spoiled white kids that think its all about them and getting what they want NOW..ryzen will get better over time. So return it if your so angry about not getting 300 fps+ in stupid ass games. I will be happy with mine and enjoy it.

No... just no

cpubenchmark.net/compare-test.php?cmp[]=2970&cmp[]=1190
arstechnica.com/gadgets/2017/03/amd-ryzen-review/
anandtech.com/show/11170/the-amd-zen-and-ryzen-7-review-a-deep-dive-on-1800x-1700x-and-1700/23

Every benchmark so far has been mixed. Yeah Intel has stuff that blows Ryzen out, but it competes fairly well for the price point AMD offers, which is what they are aiming for.

People who want top of the line performance with no expense spared will ALWAYS go with Intel, but for those of us who like to keep a smaller budget AMD is worth looking at. For now.

I should add that *weak CPU with high single core performance.
Because normally a game is bottlenecked by either the render thread or GPU. If your render thread isn't bottlenecked the CPU might bottleneck on all cores. Which isn't a good thing, because it can result in stutter.

golem.de is the first that has redone the benches with the new bios on a msi
translate.google.co.uk/translate?hl=en&sl=de&u=https://www.golem.de/news/ryzen-7-1800x-im-test-amd-ist-endlich-zurueck-1703-125996-4.html&prev=search

he saw increase from +17 to +22%...

So you're telling me the 7700k is shit and weak because it was seeing 90%+ on all core/thread usage in the Division and Tomb Raider while the 1700 was seeing about 50-60% per core? Sure, I can totally see the 7700k being weak and underpowered man.

>No it's really quite horrible
>7fps difference

Shut the fuck up.

can someone link to benchmark with updated bios?

Sorry, wrong CPU Benchmark link. Here is the correct one.

cpubenchmark.net/compare-test.php?cmp[]=2970&cmp[]=2919

no that is showing that its a ubisoft game it was stuttering on both sides while having 50%+ utilisation and that shouldnt have happened

I keep having to repeat this, but look at the prices.

A fucking $240 Intel 7600K is beating a $500 Ryzen 1800X

Both are on water mate.

Yea yea all the ryzen bugs haven't been worked out yet, but GTA V does take advantage of more than 4c/8t, and you see it stressing all the 1700's cores.

It's doubtful any game will reach 100% linear scalability with 8c/16t, which should just give you more doubt about the future "performance" gains of the 1700/1800x.

GTA V, CIV VI, Ashes of singularity are about as optimized as you're gonna get for 8c/16t atm. And if future games scale similarly, that means 7700k>1700/1800x and 6900k/6950x >/1800x for games.

why not go 360p retard?

No it doesn't. Look at the real world performance benchmarks, Ryzen is a very bad value.

So what exactly is stopping Intel from just bribing AAA developers to not "optimize" for Ryzen and then just paying a few billion dollars down the road in fines business as usual?

>People comparing the price of Ryzen to various Intel chips
>For some reason barely anyone even the cost of motherboards
Hmm...

SSE, welcome to ICC

>A fucking $240 Intel 7600K is beating a $500 Ryzen 1800X

Now put that 7600k up against Ryzen in a heavily multithreaded test and not a fucking childs game. Oh look, Ryzen fucking destroys it.

the fuck did you saw gtav stressing it?

the division stressed them rotr did gtav was having even 0% utilization at some points..

you have no clue what you're talking about.

this is not ryzen bugs. this is facts about core scaling.

that "utilization" may be due to their use of spinlocks instead of relying on signal vectoring to handle thread sleep/waking.

>"Destroys"

Considering during a worst case task a 7700k falls behind a 6950x by 21% in premiere

Then youd have to be a fool for sacrifing overall performance for shaving a few minutes off export.

Nothing

Yes if you do video encoding & ray tracing all day, Ryzen is the better choice.

If you do literally anything else, do not get Ryzen.

>Tomb Raider
>Ubisoft

Try again. And a lack of CPU utilisation just shows poorly how poorly coded some of the games were.

Yea the 7700K only has 4 cores. It's pretty shit.
It has high single core perf, so it won't bottleneck the render thread at high frame rates but it's too weak to render rest of the game because it only has 4 cores.

It's not really an issue, though. Most of the time you're bottlenecked by graphics card long before you are bottlenecked by either render thread or full cores.
Unless you're literally playing games at 720p in which case you can kill yourself because your existence is legitimately disgusting to me.
I don't think the Ryzen performance is interesting at all. It is about what I expected based on performance comparison and clock rates. I'd never buy it, because I have X99 platform and I have better CPU but the price on it is pretty good.

You'd have to pay me to get me to use trash like the 7700K.

>I meant watchdogs 2, Imma retard.
>THEY LOOK THE FUCKING SAME AT GLANCE

its shown that for the ryzen line, performance remains the same, its just clock speed and core count that changes. So you can also just buy an r5 1600x for $259 to get the same gaming performance as an 1800x, two more cores than a 7700k, while saving close to $100. You'll just have to "wait"

I just linked you about 20 real world performance benchmarks retard.

the division moron is a ubisoft title

and no having such high of cpu util and still manage to have stutter is even worse than having less scaling but decent fps

>A fucking $240 Intel 7600K is beating a $500 Ryzen 1800X

In vidya games. The 7700k also beats the 6900k which costs $1000. Are you going to buy a 6900k for gaming? No? Perhaps because it's a workstation CPU?

Really makes you think.

yes a $1,000 dollar usd part outpaces one costing half as much only on poorly performing games. the correct comparison is 1700 to 7700k which shows about 5% at 1080p in games but then the 7700k gets utterly destroyed in any multithread productivity task. 1700 wins ez over 7700k.

sorry don't translate

amazing...

Ryzen is a terrible gaming cpu because it can't play AAA games at 720p at 1000fps. I'm so glad I spent a stupid amount of money on my 4c/8t Intel processor that is only good for gaming.

it's 50-60 cheaper to go r7 1700 no x than 7700k even if these two were exact same price, but 1700 is cheaper in most places

Clearly you didn't even look at them

>Perhaps because it's a workstation CPU?
Then where are the PCIe lanes?
Then why does it only run dual-channel memory?
Then where is the PCIe passthrough?
Then why are they marketing it for gaming?

nobody uses premier with CPU as far as I know

I did, and it showed mixed results. Just because you don't like it doesn't mean it's invalid.

Even with all this I'm not going to make a final call on it (unlike retards such as you) as tests are still ongoing, so it might get better or worse. In particular the passmark tests have only had 2 runs with the ryzen I linked.