RYZEN LOL

LOL

Other urls found in this thread:

cpu-world.com/Compare/303/Intel_Core_i3_i3-4170_vs_Intel_Pentium_Dual-Core_G4560.html
anandtech.com/show/11170/the-amd-zen-and-ryzen-7-review-a-deep-dive-on-1800x-1700x-and-1700/18
reddit.com/r/hardware/comments/5xcnye/720p_r7_1700_vs_7700k/
twitter.com/NSFWRedditGif

RYPOO, NOT EVEN ONCE

>broadwell 3.3ghz = kabylake 4.2ghz

BTFO by based TODD

Why are they even testing things at ultra even at 1080p. Shadow distance kills performance in that game and I thought they didn't want to be GPU bound? Also old benches lol get out Intel shill.

>DirectX 11
2009 has called

>outperformed by multiple dual cores

So this... is the power of Raisin.

Yeah but nobody on g cann afford x99 so it doesn't count.

For once it sort of makes sense - shadows are traditionally rendered by the cpu. I still don't trust the results given the hilarious variance we've had with ryzen due to motherboard BIOS being in various states of maturity (gigabyte has theb est in the here and now).

when you need to increase the clock to get more performance in a NEW product, you're doing it wrong.

the 5775C isn't on x99, its on the same plaform as the haswell 4790k

fallout 4 is a terribly optimized game running on an engine older than a decade

shadows aren't rendered on the cpu, sure shadow maps and the info is batched by the processor just like the rest of the data, but you need access to vertex normals and the depth buffer to actually render the shadows correctly

RYZEN? MORE LIKE... FALLEN

...HAHAHAHAHAHAHA

>8 fps faster
>40% higher ipc
broken and soon to be deprectated benchmarks general

>reee why are you testing low resolutions where the game is CPU bound and Ryzen shits the bed
>reeeeee why aren't you benchmarking one of the

>no 6900k on list
Also
>benchmarking CPUs with gaymez

>beat by a pentium

AHBAHAHAHAHAHAHAHAHAHAHHHHAHAHHA

AHAHAHAHAHHAHAHHAHHAHAHHA
AHAHHAHAHHAAHA
HA

ryzen on its own might be a bit on the weak side but the ryzen + vega combo will be fucking bazinga

Nah, Fallzen

>beat out by a $60 pentium

KEK

>HSA made manifest

At least, thats my theory.

Oh, Sup Forums is leaking again.

>$500 CPU that performs like an i3 in gaymes
>$500 GPU that delivers GTX 1080 performance a year and half later

The match made in the loo.

>The match made in the loo.

Can someone explain to me the

>b-but that game has an old engine!
>its poorly optimized! it'll get better!

memes?

What if I want to play those games right now? What if I don't want to put my faith in things "getting better" and cross my fucking fingers?

I'm using an AMD cpu right now but I am due for an upgrade. Ryzen hasn't impressed me in gaming. Don't give a shit about work station stuff. I'm jumping ship and getting a 7600k because it **already** works.

Fucking shills everywhere.

...

...

R7 1800X isn't a gaming CPU clearly. R7 1700 is though, considering it clocks with better temps and voltages, and will get a Windows patch soon to boost performance, and non-shitty BIOSs will make things even better.

If you want gaming performance right now and don't have any other CPU-heavy use for your machine then yes, go ahead an buy a 7600K or 7700K. If you're not in a huge hurry I'd still suggest you wait for a few months to see if things shape up for Ryzen in terms of Windows patches and firmware updates.

you can tell when a game is shit when non of those cpus will get you past 60 fps

It's just AMDelusion. Games will not get better, by their very nature they must be programmed serially in the majority of situations.

Consoles have had 8 cores for years and this is still how they look.

That's all pure speculation, you should not present it as fact. As it stands right now the 1700 is a worse choice than an Intel CPU strictly for gaming, especially if you're not going to OC it to ~4GHz.

>muh drivers will fix it

What is it with AMD customers and the complete lack of any ability to see recurrent patterns?

lol yeah ryzen is like such a failure lol

...

BIOS updates fixed the horrific performance of X99 memory.

because they are idiots and don't know what they are talking about. Yes there is some day 0 fuckups but EVEN with those issues addressed it just means the 1800X approaches the performance of the 7700k, its not going to surpass it for gaming and even the techs at AMD said this talking to the reviewers they sent copies to.

They spent something like 6 years on DX12 and the reality is that shit does not scale past 4 cores. Those are the kind of time frames we are actually dealing with. Ryzen is not going to magically perform better by the fall or the next year or the year after that.

wtf is going on here?

my fx8320 & gtx1060 get about 40-50 fps on all ultra with graphic mods in FO4...

>$65 cpu @ 41fps
or
>$500 cpu @ 40fps

>2017
>intel and nvidia have better bang for buck AND performance
>amd literally has nothing

Is this the end for AMD?

AMD is like a religion for these morons.

They just love to get AMD's shit shoveled into their mouths and pay for it.

6 threads is the sweet spot for DX12.

...

its a certain section of the game and they have max AA on. Benchmarkers have to use presets typically for their own sake but on your own system you should pretty much never use them.

>Fallout 4
invalid

Remember to hide and sage shill threads!

For actual games its effectively 4. It can theoretically scale to 6 decently but on actual games that sort of optimization realization isn't occurring. And keep in mind this is just for the section of your program that is feeding and receiving frames from the GPU.

>Speed test between ferrari and a ford pinto
>Set the ferrari on 30mph cruise control
>Wow, just look at our pinto go! It's as fast as a ferrari in our tests!

You know, with all the fudging AMD has been doing with their benchmarks, they may as well put an fps_max hardcap on their intel tests, it's basically the same thing.

Perhaps, perhaps not. Right now the only game engine that can into DX12 (and preumably vulkan) is the nitrous engine and that does - for all its flaws - stop at 6 threads for effecient scaling. It can address many, many more but the gains are marginal percentage wise.

It is also worth considering that while DX11 exists for legacy purposes DX12 will be crippled by the need to not entirely break DX11. It is a factor into why the community prefers Vulkan (heralded by Doom) but the reality is vulkan's documentation is far behind DX12. All in all its a complicated situation but when the big players make the jump and tell DX11 and older to GTFO things will leap forward cpu wise.

You can see this in the console space with not-DX12 and sony's custom API leveraging cpu respurces in a way rarely seen in the PC space. People complain about consoles holding PC's back but its untrue - PC's cripple themselvews due to legacy support and the stranglehold MS has.

>LONG LIVE GLIDE

Is x99 expensive now? I've been through 2 different x99 platforms in the last 2 years alone. 5820k is only like 40$ more than an i7 and still has plenty of power to spare.

The good boards are expensive and frankly if you are going to leverage the platform advantages that means shitloads of ram and/or multiple gpus.

Yeah I noticed that the boards were wildly expensive. Average of 180-300 versus 80-170.

But I just did the same layout I always did in the past. Highe4 end CPU like that 5820k at the time. I had a 980SC. 16gb RAM. 480gb SSD and 2tb drives

But I even had SSD, 16GB RAM back on my very first PC back when the 660 just dropped. Overkill with utility sometimes I guess. I'd rather never have to stress about shit loading and processing than having good graphics.

In the here and now (for less dosh) thats what ryzen is. My gut tells me AMD is aiming squarely for those critical voids between Intel's pricing tiers which in practical terms means AMD will offer more for less or less options for less money - effectively neutering Intel's division between overclocked mobile chips and gimped xeons.

>I'd rather never have to stress about shit loading and processing than having good graphics.

Given the slow progression of cpus relative to gpus if one has the budget to allow for it I would rather go balls out on a cpu than gpu this day and age. I'm building a 4k machine off the back of my existing platform and I will be spending considerably more on the cpu front than gpu as I expect to the cpu to last for 4-5 years compared to the 2-3 for the gpu to deliver performance I desire at the settings I want at 4k.

cpu-world.com/Compare/303/Intel_Core_i3_i3-4170_vs_Intel_Pentium_Dual-Core_G4560.html
look at the "Instruction set extensions" section...
my bet is that the game is optimized for intel CPUs *only*

Now have him enable High Performance Mode on Win10 and run it again.
W10 cripples performance to save power, it's a well known "feature"

add more physics and make 4cores suck ass

It's a Memory sensitive game. Ryzen motherboards and Ryzen it's self has memory timing issues so they're super slow times, and it only gets worse the more clockspeed you give RAM.

Fucking Sup Forums told me there was no improvement from Sky and Kabylake yet my 4690K is falling behind at the same clock.

You fucking cunts spreading desinfo.

>1.8 fps apart from competition
LOL

Yea, it's amazingly funny how shit inlet is at 1k$

anandtech.com/show/11170/the-amd-zen-and-ryzen-7-review-a-deep-dive-on-1800x-1700x-and-1700/18
I'm in treats right now.Suck it, lard-asses!

A fucking $60 Pentium is beating a $500 Ryzen

Kaby and Sky have higher memory bandwith allowance, put shit memory in them they will perform as your CPU

a fucking 60$ Pentium is beating a 1000$ Intel

>my bet is that the game is optimized for intel CPUs *only*

As far as most software development goes there are only two platforms.

>Intel Mainsteam
>Intel HEDT/server

Ryzen isn't on that list.

Which is why no one here is suggesting people buy the $1000 Intel.
But everyone is shilling the $350-$500 Ryzen 8-cores despite the fact they perform in a very similar way.

Nobody is telling you to buy it if you don't need it.
1700/1800 is filling price gap of 8c/16c, and does so very efficiently.

actually, nevermind, this is a rather good argument

for this specific game...
for real world tasks, if you really need computing power, just go and buy a ryzen cpu

why are these 8 core 16 thread CPU's being compared to quadcores which can obviously reach vastly higher core clocks and thus perform better in single-threaded garbage like most games?

the 8 core cpu's must be compared to the intel workstation line-up such as the 6900K, not the fucking 7700K you retards. of course it won't do as well in single threaded. the 1800X would destroy any intel quadcore in rendering or calculation work (video, 2d, 3d, compiling) which is what it was made for (content creators).

fucking gaymen faggots

These were marketed as gaming and enthusiast chips, they fail horribly at both

>AMD launches a ryzen ad campaign for gaming
>people point out that they still suck shit in gaming
>AMD drones throw a tantrum and defend AMD "they were never meant to be for gaymen even though they lied and advertised it as the case!" "test it on these multithreaded apps that only 5% of Sup Forums uses instead!"

Every fucking time

dis

Not exactly true, R7 was marketed as competing with 6900x in workloads and gaming

1700 was shown to compete with 7700k and showed to be competing with it but biggest talking point was that you can stream, play games without any FPS inpact on the game compared to 7700k

There are more of these retarded benchmarks than there are stars in the sky. Doesn't make any of them relevant or useful.

...

Pretty sure they installed the wrong ram for this one.
Because every other test is a lot more positive.

You never got my permission for using my shitpost in your image, but I'm giving it to you now.

Guess what, it fails at streaming too

5775C has a 4th level cache which is actually 128gb of eDRAM and I'm pretty sure it largely benefits some workloads, but doesn't favor others. I think giving the top end quad core Intel CPUs more cache would actually significantly increase performance in some aspects.

mb* not gb

Cpus are not for gaming anyways. Stop being stupid.

Only Intel CPUs are for gaming.

lmfao

Nice benchmark. Here's some more

like clockwork

Ryzen is competitive with Broadwell-E on compute performance and power usage. 8C/16T for $500 is not to be underestimated.

For gaming, Zen is generally better than Ivy Bridge but worse than Kaby Lake. A couple of games have shown markedly better frame times and minimum FPS than Intel.

Software has not been optimized with Ryzen in mind, only with Intel because Bulldozer was totally obsolete on arrival. Most software will not be patched for Zen, nor will future software necessarily be tested for performance on platforms other than Intel and consoles.

Basically, AMD is competitive again. They didn't BTFO Intel like Athlon 64 did, but it's good enough all around with enough extra juice for multithreaded work (especially if you're programming your own stuff) to have its place in the market.

You realize that is a fake benchmark that was already redone:

reddit.com/r/hardware/comments/5xcnye/720p_r7_1700_vs_7700k/

The retest shows a 20% performance difference between the 7700k and 1700

>AMD is competitive again

>Costs twice as much
>Has worse performance

How the fuck is that competitive

The R7s are competitive with Intel's heavy multithreaded processors like the 6900K for heavily multithreaded workloads. R5 will probably offer competitive pricing with Intel's quad cores, but will probably be about the same as R7 for games unless they cough up a lot of extra clockspeed.

I hate Sup Forums so much

It's not. What you're witnessing is an AMD fanboy in the bargaining stage of grief.

I didn't know my 7600k was this swag

With Intel gimping supply of baby broadwell nobody ever really tested it - but it remains (to this day) the king of Intel IPC. The huge slab of L4 helps immensely in both general workloads and igpu usage (the latter is very obviously - its the only way iris pro beats AMD's offerings) but reviewers ofcus on the relatively speed demon kaby lake rather than the more expensive broadwell/broadwell-E which are IPC over clocks.

To wit the likes of the 6900k and 6950x absolutely murder across the board when clocked up but not by massive amounts in vidya and thus are boring to reviewers. When you start testing HEDT/ever workloads (like tom's hardware german arm tends to do so) you see the difference in performance.

With zen generally matching broadwell-E its a sign AMD is playing not for keeps, but to WIN in the HEDT/ server market. The 1700 is especially dangerous given its implications for lower clocked (but higher core count) chips.

Consider this: in the server space the actual cost of the chips is irrelevant- its the power draw that matters and ryzen appears (asper The Stilt's testing) has Intel beat on that front until you leverage AVX2 heavily. Watch this space - Naples is a game changer and Intel knows it.

>game is optimized the worst i've fucking ever seen, probably not even optimized at all

>SEE DIS PROOF

ok this particular game is using an outdated engine but really, even with the best game engine, you're not going to get anywhere near the same multi-threaded usage as with a pure video rendering benchmark like cinebench, except maybe for some contrived AOTS-style game

>Broadwell-E beaten by a quad-core
I smell bullshit

>Costs half of Broadwell-E
>Performs on par, if not slightly better
Yes, this is competitive

OP here, it was all in good fun. Ryzen is meh but has a niche, IMO.

The way it's meant to be played™

> on par with a broadwell/i7-5960x and costs a fraction of the price
Everything checks out. What's the problem?

>he didn't get the memo

You can shift the Intel product stack around as you see fit, as long as ryzen loses.

>ryzen beats a 6900k
JUST WAIT FOR ST BENCHMARKS
>ryzen loses to a 5ghz 7700k
AHAHA AMD IS DOOMED

Pirice is literally a non-factor. A 6900k losing to a 7700k is fine, as long as it shows AMD doing worse.

Okay, I just googled the source myself before you posted that
>While I wasn't able to capture the number of dropped frames from the stream, I was able to capture the FPS numbers from Dota 2. All the eight-core CPUs did well, but the 7700K took a significant hit to performance by as much as 18FPS with OBS active. Ryzen did well, only dropping 3FPS while streaming, but was ultimately slower overall than the Intel systems.

The only that part said anything about the OBS setting is
>That's particularly true in the OBS/Dota 2 benchmark, where the game was software-encoded at 3500Kbps and streamed
It doesn't said anything about the preset setting, So I guess they are at default "very fast", even a shitty i5 can do this.
But Very Fast is shit for high pace game like FPS, it will create a lot artifacts, Dota 2 too if the camera moved around too fast.

Its not I'm trying to diss intel here, but I do works as video mixer in live streaming Esport event.
If you want to get better quality you need to changed your preset, but it will used more CPU power for better quality.

So if anyone can find me an OBS benchmark of ryzen running at better preset It would be helpful instead of some random reviewer who known nothing about encoding in streaming.