RX 480 / 1060
Images are sometimes not shown due to bandwidth/network limitations. Refreshing the page usually helps.
You are currently reading a thread in /g/ - Technology
You are currently reading a thread in /g/ - Technology
Why to choose RX 480 over 1060?
* NVidia gSync costs more
* No SLI for the 1060 3GB version, RX 480 4GB/8GB both support crossfire
* 1060 has less memory, it has only the 3GB and 6GB versions
* 1060 has a 192 bit bus interface vs the AMD RX 480 256 bit bus interface (AMD probably doing better in higher resolutions)
* AMD has 3 display ports, perfect for triple monitor setup
* RX 480 does better in DX 12, near 980 ti performance in Hitman, near 980 performance in AoS, over 980 performance in Total War: Warhammer
* RX 480 already out, NVidia seems to have trouble producing cards and the 1060 will probably also have production issues
* NVidia has problems with HTC VIVE
* AMD cards mature better, they are more future safe than NVidia cards, they also deliver driver updates for older cards to improve performance
* RX 480 is cheaper ($299 for NVidia FE vs $199/$239 for AMD stock)
Why to choose 1060 over RX 480?
* RX 480 (stock) has no DVI, 1060 probably has
* Lower TDP
* Slightly better DX 11 performance
* Better OpenGL performance
Also
* Both stock versions have cheap coolers
* AMD supports open source (for example Vulkan)
* AMD has drivers now, they seem to do well on the software side (for example the RX 480 power draw fixes came fast)
* 1060 and RX 480 aftermarket cards might have similar price while the RX 480 being a bit cheaper on average (maybe around $20) however this is a prediction only
* There will be aftermarket RX 480-s with 8pin power connector and this will fix all power draw problems, probably have better OC capabilities also
The RX 480 seems to be a great card, even the stock one seems to be actually a good buy. I don't see many reasons to wait for the 1060. The FE model seems to be overpriced, need to wait even more for cheaper aftermarket cards while the RX 480 is already out.
Picture related: Sapphire RX 480 Nitro (8 pin, 1 x DVI, 2 x HDMI, 2 x DP)
Probably going to buy this baby, what do you think? Will this card run stable 1400 mhz?
>>55472192
Who gives a fucking fuck
Go take your /v/irgin card shit to a general you oxygen thieving cuntwhistle
NOT
TECHNOLOGY
OP IS A DUMB FAGGOT AND IT SHOWS
>>55472208
>>55472231
Reasons 4chan is cancer for 200 please.
I bought a rx 480 to upgrade my gtx 750tioc and im pretty happy with it so far did I do good /g/entoomen?
>>55472231
how?
>>55472260
yes
>>55472192
There's no 1060 3GB version you fucking mongoloid
>>55472260
I'm doing the same in a couple of months. What's your CPU?
you can tell op is underage AND new just by the way he wrote the post
>>55472260
Here is the overclock guide.
https://www.youtube.com/watch?v=oss1Ziys5Fc
>*
is this nigga serious? lurk more, faggot
>>55472260
maybe
>>55472320
Do you really wanna set it on fire?
Please can the 1060/480 threads just be deleted or moved to /v. Mods. Please.
>>55472260
>buying reference cards
No, you didn't, kys.
>RX 480
>8gb memory
>dx12
>GTX 1060
>3gb """"memory"""
>no dx12
Sounds like an easy decision to me
>>55472192
Who the fuck do you think you are telling me what to buy? I'll buy whatever the fuck i want you fucking small dick shill. Stop being a faggot worrying about what people buy if i want to spend 400+ on a gimped card that's on me
>>55472371
There's no 3GB version of the 1060, and why are you saying it has no dx12?
>>55472371
AYYMD doesn't even support DX12, GPUs that are not Feature Level 12_1 are not DX12 GPUs
>>55472409
Cause hes trying to push DX12 but doing it incorrectly. It will support DX12 but it will have have full hardware compatibility like the AMD cards. That means AMD card will be more efficient doing things like hardware based Async computing where Nvidia will have to do it using software/drivers.
>>55472437
>Feature Level 12_1
Includes Software Async updates for Nvidia cards. AMD has ignored this as their cards don't need it.
>>55472192
The RX 480 nitro is fucking beautiful, and they came up with a clever way to get similar ventilation as the r9 380 nitro without making a massive over sized cooler. However, this system uses vents in the pcb itself, so it blows the hot air straight up into your cpu cooler. Do not suggest if you're using a stock cooler. But you don't use a stock cooler, do you anon?
Why don't they make a hardware general on /vg/ ?
>>55472350
Are you having a conversation or is it meme time?
>>55472454
Wrong, Feature Level 12_1 is not async
DX12 is not async, get that through your head
No Conservative Rasterization
No Rasterizer Ordered Views
No Tiled Resources Tier 3
AYYMD GPUs are not future proof and won't even run future games, just like AYYMD SM2.0 GPUs couldn't run SM3.0 games
>>55472465
Isn't it obvious? We're on /g/ after all.
>>55472373
>Here we see an angry, violent cuck spotted in the wilds of /g/, venting his frustrations at the world en route to making another poor life decision by purchasing an overpriced Nvidia card.
>>55472475
I like the way I presented you an argument with out insult but you have to insult some one to get your point across. I will look it up today or can you provide source?
>>55472192
poo in the loo pajeet. Everyone is now hype by 1060.
>>55472192
>* NVidia gSync costs more
Fair point. JewSync is ridiculously expensive.
>* No SLI for the 1060 3GB version, RX 480 4GB/8GB both support crossfire
CF/SLI are pretty much dead because so many new games are completely incompatible with them.
>* 1060 has less memory, it has only the 3GB and 6GB versions
Is the 3GB version confirmed? Anyways benchmarks show 8GB has no benefits over 4GB in RX480.
>* 1060 has a 192 bit bus interface vs the AMD RX 480 256 bit bus interface (AMD probably doing better in higher resolutions)
RX 480 is already too slow for 1440p so it doesn't really matter.
>* AMD has 3 display ports, perfect for triple monitor setup
So does NVIDIA.
>* RX 480 does better in DX 12, near 980 ti performance in Hitman, near 980 performance in AoS, over 980 performance in Total War: Warhammer
All three games are sponsored by AMD, literally nobody plays AotS and Hitmans engine is complete mess, Episode 1 runs better on AMD and Episode 2 on NVIDIA and DX12 doesn't draw as much detail as DX11 renderer:
http://www.pcgameshardware.de/Hitman-Spiel-6333/Specials/Episode-2-Test-Benchmarks-DirectX-12-1193618/
>* RX 480 already out, NVidia seems to have trouble producing cards and the 1060 will probably also have production issues
This requires some proofs.
>* NVidia has problems with HTC VIVE
True but I doubt that people throwing $600 on VR HMD buy $250 GPUs.
>* AMD cards mature better, they are more future safe than NVidia cards, they also deliver driver updates for older cards to improve performance
Fermi does much better than it's old competing AMD VLIW architecture cards.
>* RX 480 is cheaper ($299 for NVidia FE vs $199/$239 for AMD stock)
AMD stock card is garbage. Lets see where custom model prices settle.
>>55472192
Except despite all this, the RX 480 is a literal embodiment of the AMD Housefire meme and will destroy your motherboard.
And for that reason alone the 1060 wins.
>>55472475
>Conservative Rasterization
I'm pretty sure this is a proprietary Nvidia "feature" that works within the DX12 framework. Saying AMD doesn't haven't it is kind of fucking obvious since Nvidia developed it.
>>55472535
Are you stupid or what? Intel fully supports Conservative Rasterization in Skylake GPU, it is a part of the D3D12 spec
You AYYMD uneducated morons are truly ignorant and clueless
>>55472475
https://en.wikipedia.org/wiki/Feature_levels_in_Direct3D#Direct3D_12
AMD has
* tier 2 resource heap
* UAV slots for all stages (full heap)
* Stencil reference value from Pixel Shader
* Resource binding tier 3
BUY OUR THIRD REHASH OF GCN PLS PLS SAVE US
>>55472554
>>55472521
>Anyways benchmarks show 8GB has no benefits over 4GB in RX480.
Actually it's half true.Some games actually runs better with the 8GB model (like 3 - 5 fps), but a lot of them don't.
I think it's more a question of future proof, since games will use more and more ram.
>>55472192
Hey guys, first post, but need opinions. Best way to block the systems administrators at the corporation I work for from monitoring me constantly, it's annoying and I use 2 personal laptops on top of my work pc. I believe they are using meraki network edition because he said my pc was infected when I booted my laptop and steam updated and he saw this (google it, from 2012) ReVuln_Steam_Browser_Protocol_Insecurity.pdf. so overall, best ways to block then from seeing anything, they also use other tools and no a vpn does not work. just need ideas.
>>55472192
If you have a pci-e motherboard, I suggest a pci-e compliant card.
Non standard cards may violate your motherboard warranty.
>>55472521
> CF/SLI are pretty much dead because so many new games are completely incompatible with them.
Somewhat agree, it isn't dead yet. Probably will be replaced by multi-gpu software solutions built using the features that new API-s offer.
> Is the 3GB version confirmed? Anyways benchmarks show 8GB has no benefits over 4GB in RX480.
Aftermarket versions will have 3GB, the FE will probably be 6GB only at $299.
> 1060 has a 192 bit bus interface vs the AMD RX 480 256 bit bus interface (AMD probably doing better in higher resolutions)
Seems to be just enough performance for 1440p gaming. A bit performance tuning using game settings and it will run well. I'm not talking here turning everything low, more like medium/high. Most of the benchmarks seem to be on max settings. However about the 256 bit bus, it will perform better for vram heavy workloads.
> So does NVIDIA.
Actually yes, this point should be thrown out, agree.
> All three games are sponsored by AMD, literally nobody plays AotS and Hitmans engine is complete mess
Seems like most of the DX 12 have AMD icons on them. Well, good for AMD, there are tons of Gameworks titles that AMD has to compete with. Cannot tell nothing about the Hitman engine, haven't seen the source code.
> This requires some proofs.
The proof is the 1070/1080 launch and the final proof will come when the 1060 is finally launched.
> Fermi does much better than it's old competing AMD VLIW architecture cards.
I had GCN in mind. Agreed with the VLIW part.
> AMD stock card is garbage. Lets see where custom model prices settle.
Agreed, that's why I' waiting for the Sapphire Nitro. Also I have an async monitor so nvidia isn't really an option here for me. Nvidia should fix this, it is a big turnoff. However nvidia has better recording features. Anyways I won't do any recording.
>>55472554
Conservative Rasterization was pioneered by Nvidia after GCN was already created so of course it just so happens to be a technology that is incompatible with GCN. That doesn't mean that intel couldn't license the IP from Nvidia for inclusion on Skylake iGPUs.
I'm just pointing out how hollow it is to say that AMD doesn't support Conservative Rasterization when it basically exists as a bullet point for Nvidia to say "look at this thing we have that they other guys can't have."
>>55472864
I meant *vsync in the final part and also copied the wrong line for the third statement. I'm completely aware of that.
>>55472765
vpn does work if its encrypted.
>>55472864
My 380X plays new games at high/ultra at >60 FPS at WQHD. 480 is perfect for the resolution, if you're not some dumb faggot who needs everything on ultra even if it brings no visible benefit.
>>55472875
Yo, DX12_1 and all the goodies are what bring new rendering modes to DX12/DX11.3, DX12_0 supported by AMD lacks any of it, meaning all they have is better CPU usage.
Of course saying that it can't be done by AMD(or older Nvidia GPUs) is silly because it can be done on software mode using shaders, not that you would want to unless you want games running at 10FPS.
Also Intel has supported some features of DX12_1 even before Nvidia or even before it was called "DX12_1" or even before DX12 was even announced.
The "DX12_1 is just Nvidia" has been throw around by AMD people trying to downplay it for some time now and people are falling for it.
>>55472483
Cuck/Cuckold :A demasculated Millenial liberal male who often speaks of white privilege and modern feminism. Blames white men for all of the world's problems.. 480 cost 399+ where i live fool
>>55472260
good and if anyone says otherwise they are memeing, shitposting or shills
>even joking about buying obsolete Nvidia gpus
if they dont sort their bullshit out in Q4 with some HBM2 cards geared for DX12+, AMD ill be pushing them out of the market.
the rx4-- series is built on obsolete components but at least its geared for DX12+ standards.
>>55472192
>>55472192
l
Why are the atidrones so buthurt this gen? The constant threads trying to do damage control is really making them look pathetic.
>>55474563
Not good, why the fuck did he buy a reference card?
>>55474785
>all gaymeworks games
wew lad
DX12 is a marketing scheme, if a card has better fps in a game than another card, it's the better card.
Doesn't matter what API they're using.
>>55474759
>>55474771
>>55474785
I wonder what world you live in...
http://www.pcgamer.com/radeon-rx-480-review/
Also look at this:
http://www.legitreviews.com/amd-radeon-rx-480-undervolting-performance_183699
3-4% more performance when optimized. This means it would score around 75 average in the pcgamer test and this means it would be 11-12% faster than the gtx 970. Also AMD updated the drivers and gained about 3% performance when optimizing the card plus less stuttering.
>>55475651
Only it does.If you look at dx11 performance the 480 is currently on par with a 970 meanwhile with dx12 it easily matches or is better than 980.
>>55472192
Primary concerns with the 1060:
Incompatibility with then HTC Vive (or its successor?)
The MSRP is a lie because partner boards for the 1070 and 1080 all seem to be the same cost as the founder edition and i expect the same to be true for the 1060.
Otherwise i'm quite tempted... always bought AMD though, if the 490 ends up being competitive when Kaby Lake is released i might go for that over a 480 or 1060.
>>55476356
>bothering with kaby lake when zen is coming in like 3 months
wew lad
>>55476396
But Zen is probably going to be just as disappointing as the 480
why should he wait?
>>55476463
>implying the 480 was disappointing
NVIDIOT SHILL SPOTTED
>>>/v/
>>55476706
1 rupee has been deposited into your AMD toilet building fund.
>>55476726
>>55476714
samefag shitposter
>>>/v/
Still running a XFX Radeon HD 7870. And I'm still getting good FPS at 1080p. Really for now that's all I care about. Was thinking about upgrading to either the 1060 for the 480.
>>55476837
Lel just let it go. You AMD fags can't be even funny. Probably cuz your school had to waste time trying to teach you how to use loo instead of introducing you to nuances of English language.
>>55472875
>Conservative Rasterization
you really dont know what amd did with their new geometry processor eh
hint they already have it but for marketing BS aka vega they havent really said anything why you think they directly intergrated the HWS into the geometry processor instead of replacing all the ace engines with HWS?
>AMD general
>Promises, speculations and accusations edition
>>55472352
No it is tech and everyone who cries about it either has some serious autism or is just a little bitch
Is there a chance that i can pre-order Radeon RX 480 Nitro+ OC 8192MB GDDR5 PCI-Express Graphics Card for cheaper than 275gbp?
Preferably in euro currency.Any site that would sell do this and would ship to Hungary too?
>>55476356
- VR is DoA same way 3D TVs were the big thing dee years ago and now nobody gives a fuck about them. It's simply too expensive to hit mainstream market which means no games. It has nice uses in education and porn but games are expensive to make.
- 2016 overall is just a massive paper launch both for Nvidia and AMD. Believe it or not but those cards WILL drop in price once people stop pre-ordering them. It's funny to see you complaining about day 1 prices when you're an AMD waitfag.
>>55472192
when's this coming out Stateside
>>55477114
are you actually comparing 3d with vr?
vr is a totally different monster
>>55477093
Anybody?Please
>>55472192
I would never recommend AMD graphics. They've fucked me one too many times with promises of functionality that didn't work and with shitty drivers.
Beyond that, AMDs practical support of Linux is terrible. If you want good 3D performance on Linux, the choice is clear.
>>55477213
see on overclockers if they have any
>>55477223
That's where it's currently available for pre-order.Hence my question if there is other site where i could pre-order it for 275 dollar or euro.
>>55477222
>amds practical support of linux is terrible
meanwhile amdgpu driver has already 335% better perf than the previous on linux..like 3 months now
paid shills just dont know where to go in order to find a flaw
>>55477277
if you have enough brain cells to follow the maze of french and german sites you might actually find one i saw some days ago on 282 on a random french
>>55477213
nitro is £249.99, and the powercolour devil one is £254.99 on overclockers.
>>55477281
>compares AMD to AMD
>crows about AMD beating AMD.
>>55477281
335% increase doesn't really mean anything when your performance is close to non-existent.
As someone who's stuck with a 7970GHz for now, it's pretty fucking annoying to be limited to using IG+VM passthrough or just not do shit like play muh gaemz
>>55477335
and you will never see any gain since amd already dropped support for gcn 1.0 ....
problem is the 7xxxx didnt had any programmable ace engines while from gcn 1.1 and after they do have
sooner or later you will have to buy a new card its been 5 years already and while its a great card i dont deny it sadly if you choose to go to botnet 10 you wont be able to play any of the new games that support dx12 (warhammer/rotr and most probably every single one of them from now on)
So, seems like undervaluing the 480 actually results in a performance increase, maybe the initial batch of cards were just lemons.
>>55477213
http://www.mindfactory.de/Highlights/MindStar
>>55477504
>retail
>>55477393
That's the problem though, if I just plop in my old 580 I can actually play games because the drivers function better on GNU/Linux.
I like AMDs hardware, but I've never ever had a card where I didn't have any issues (crashes in windows, simply not working in linux, flickering, etc.)
It's laughable and has left a tainted feeling.
>>55477523
its with import fees but without vat
so its EU retail
>>55477555
Also it's not nitro.
>>55477548
Same experience here, in fact I'm running a 570 in my primary machine.
>>55477555
Without VAT? I can read german and it it with 19% tax. It is a special offering, they have only 1000 cards available and 800+ already sold.
http://www.mindfactory.de/product_info.php/8192MB-Sapphire-Radeon-RX-480-Aktiv-PCIe-3-0--Retail-_1109425.html
Only 249 wow, most of them are 269.
>>55477880
my bad anon my german skills are good only to say hail hydra
lol
tho in my country they started 350 and they have dropped on 346
and they expect to sell them
The problem with 480 is the power consumption.
I want to upgrade my pc I built years ago. I currently have a a phenom x4 and a GTX 770. Running on a 450 watt PSU.
If I buy a Vishera the PSU won't be able to carry the 480, but will carry the 1060 without any problem.
The problem is more common than you think, most people use a 400 to 500 watt PSU and buying a card that draws 250 watt under load is a huge issue.
>>55478008
the RX 480 would be a good upgrade for you
>>55478008
the card draws 164 watts on gaming loop....................................
unless that you think torture is normal mode..
>>55472192
>1060 has a 192 bit bus interface vs the AMD RX 480 256 bit bus interface
And yet both have a reported 8gb/s.
>>55478171
they also reported that 1060 is 15% faster than 480
will literally less hardware than its capable
but ok nvidia never lies
>>55472864
>The proof is the 1070/1080 launch
Fun fact - Overclockers sold out their 1000 cards received in the first batch of 480s. Just like they did 1080s. The proof seems to be in the opposite.
>>55474785
>project cars
filtered
>>55478190
You sound like the fanboys hyping up the 480 claiming "Nvidia always lies about its TDP." So far Nvidia's been more true to their TDP than AMD.
>>55478008
GTX 770 eats 200W - 250W under load. Phenom 965 eats ~200W. Your PSU should be almost dead. Also the backbone of your PC seems to be from the stone age, haven't you considering upgrading it? I mean throwing a $300 card into it, wtf...
>>55478231
>So far Nvidia's been more true to their TDP than AMD.
[citation needed]
>>55478231
since when? please tell me one nvidia card that was actually less or on par with the tdp they said
protip you will find zero niet nada every single card always ALWAYS draws more power
>>55478008
why would you upgrade to a vishera (am3+) from (am2+) if am4 mainboards will come out this year?
Also with that tight of a power budget you shouldn't get a 1060 either. You have 0 overclocking headroom. A 8350 system with an RX 480 takes about 370watt. With a 1060 it would be around 340watt. Better but still just scraping along.
>>55478261
>[citation needed]
>>55478262
480 regularly went over its TDP on launch.
480 (150w): https://www.techpowerup.com/reviews/AMD/RX_480/22.html
http://www.pcper.com/reviews/Graphics-Cards/Power-Consumption-Concerns-Radeon-RX-480
1070 (150w)/1080( regularly on target with their TDP
1070: https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1070/22.html
1080: https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_1080/24.html
>>55477312
Was it a nitro?Also how does one find such sites?
>>55478008
>most people have 400 or 500 watt psu
i just wanna know what 3rd world country u live in
>>55478330
>TPU
Uh, no. Try a reviewer who has proper testing equipment.
>>55478330
its nice cherry picking sites that doesnt have any serious equipment to measure
meanwhile back on reality
http://www.guru3d.com/articles_pages/amd_radeon_r9_rx_480_8gb_review,5.html
1070 161
480 164
and this is before the 15% reduction toms said it got
so add that -15% and you actually get a better number than 1070.......
>>55478378
Such as?
>>55478380
So if we go by your review AMD released their product above the TDP just like Nvidia, ie. they were both lying about the TDP?
>>55478253
I have a gimped 6 pin 770 version and 95w CPU (945). The card itself complies to what it says - 150w draw under load.
Now if I get a 125w CPU and the RAM will run at higher clocks I'm afraid there won't be much room left.
>>55478295
I bought a good am3+ mainboard back then, so I only have to swap the CPUs and the cards.
>>55478390
G3D and Tom's
>>55478360
you better create alerts
https://www.google.ch/alerts
>>55472192
gtx 1060 4 tflops = rx 470
>>55478380
Oh and you should also note:
>and not the average gaming power consumption.
![3086066-5455250892-polar[1].jpg](https://i.imgur.com/xIcjlos.jpg "3086066-5455250892-polar[1].jpg")
Why would I put budget hardware in my computer that I use everyday?
AMD is shitty budget indian tech.
>>55478407
they always do that wasnt my point lol if they were to meet the tdp the cards would have cost like x5 more given they will need industrial level regulators and capacitors and a lot more phases..
>>55478441
NVIDIA is shitty overpriced chinese knockoff tech
>>55478412
>G3D
>We have a device constantly monitoring the power draw from the PC. We simply stress the GPU, not the processor. The before and after wattage will tell us roughly how much power a graphics card is consuming under load.
Uses "proper testing equipment."
Versus TPU:
>we measure the power consumption of only the graphics card via the PCI-Express power connector(s) and PCI-Express bus slot.
Are you sure you aren't just picking benchmarks by their results and then trying to later justify the results?
>>55478475
>ignoring Tom's
Convenient
>>55478451
>they always do that wasnt my point
Thank you for supporting MY point:
>So far Nvidia's been more true to their TDP than AMD
Even in your own citation the 1070 is closer to its 150w and the 1080 is closer to its 180w than the 480 and its 150w.
>>55478475
i dont know since guru is pretty much bang on with the rest of the reviewers while tpu just flies in the moon and back pulling numbers of their ass
>>55478500
before amd drivers guru had 164 same as toms
1070 has 161
if you really believe that 3 fucking watts is something to brag about then by all means..
>>55478490
It isn't enough to show your bias in cherry-picking your reviews with one clear lie?
After a quick look I see that Tom's did provided power consumption in the 480 review but not the 1070?
>>55478502
>i dont know since guru is pretty much bang on with the rest of the reviewers
[citation needed]
TPU and Anandtech have been two of the most reliable review sites.
>>55478536
>if you really believe that 3 fucking watts is something to brag about
I'm not bragging about anything. I'd suggest you stop trying to project your fanboyism on others.
>>55478558
Tom's are an NVIDIA shill site. If there was a problem with the 1070 then they avoided showing it.
>>55478425
Thanks for this.I didn't even know it existed.
>>55478567
anand still is
but tpu a site that uses bf3 wow wotanks as a relevant games only to increase the average fps count isnt really trustworthy anymore...in fact they arent even serious anymore
>>55472192
Unless you have huge memory requirements, the 1060 is probably the better card.
>freesync requires another hardware investment, so add a couple hundred dollars, but so does gsync
>less memory matters less on lower resolutions
>very few, mostly niche games use dx12, and improved dx11 performance will impact more games in a current library
>after watching AMD's 6-pin disaster, NVIDIA would have to be legitimately retarded to have a similar problem
480 seems much better for multimonitor/high resolution setups, but the 1060 is likely the better contender for someone who wants to play a current library without buying a new monitor
That being said, I imagine the aftermarket 480s will curb stomp nvidia since it can be assumed they won't have power draw issues
>>55478595
>Tom's are an NVIDIA shill site.
I agree with you that Tom's is not reliable but someone argued that Tom's is a better choice than TPU. See >>55478412
>>55478586
you are the one saying its a huge deal while infact its not
>>55478609
I'm looking for a citation but it seems you've got nothing but your opinion to support your opinion.
>>55478631
>you are the one saying its a huge deal
Where did I say it was a "huge deal?" I believe I only pointed out how AMD loyalists were defending AMD as "always truthful" and Nvidia as "always lying." So far, the launches of both companies this generation has proven this belief to be a lie.
>>55478610
>very few dx12 games
>2016 comes and not a single relevant dx11 game
>almost all the games will be dx12
>dx12 naturally favors amd
>dx12 is niche
those paid shills are finding new excuses every day
For me this card is literally a R9 390/390x but with lower tdp.
The only thing I wished for is the same bit size instead of half because AMD is not that good in image comprassion.
>>55478650
well in terms of lying there isnt really any competition
http://www.legitreviews.com/nvidia-highlights-directx-12-strengths-amd_138178
>>55478635
obviously because we all know games from 2013 are so relevant today
>>55472317
is he wrong though
>>55478669
>well in terms of lying there isnt really any competition
Moving the goalposts?
>obviously because we all know games from 2013 are so relevant today
Really? Twice in one post? If you don't want to discuss the point raised why do you bother even responding to the post? Are you paid to defend AMD or something?
>>55478653
>very few dx12 games
That's correct and won't change any time soon.
>>55478684
not really just pointing out that the history is just full of shady shit especially from nvidia and its natural nvidia is basicly 3dfx i loved my 5500 but their tactics with glide could make nvidia gameworks to blush
>>55478709
kek that picture
>>55478708
dx11 had its first game 2 years after it got released
dx12 had its first game 8 months after it got released..
currently there isnt any big AAA game on 2016 isnt isnt or will be on dx12/mantle...literally every single one of them will be on it
and we already have 16
but i guess your fact contradicts the reality
>>55478709
>the history
Is this history relevant to the points raised or are you just soapboxing? I ask because I am trying to treat you like a sensible adult rather than a blind brand loyalist/paid shill but so far you've given me two reasons to conclude you are the latter.
>>55478740
I don't know what reality you're on, but there's not a single major title boasting DX12 support this year.
>>55478709
brilliant.
>>55478780
obviously lol i guess total warhammer isnt one
neither bf1
or deus ex
or star citizen
or catalyst(mantle)
honestly i could go on the list is BIG denying the truth is futile really
>>55478709
Photoshop level: master
>>55478825
If Nvidia doesn't win in a game benchmark it doesn't count. Thus why according to /g/ the only DX12 game in existence is tomb raider (I would cite age of wushu but /g/ doesn't know that exists at all).
In fact off the top of my head tomb raider is the only DX12 game where AMD doesn't murder their equivalent Nvidia card.
>>55478709
Is there a difference between DDR5 and GDDR5?
>>55479101
they released the async update today
and literally if you have nvidia dof on amd gets negativa scaling with async on
and they keep saying its a good metric
>>55479111
yea, DDR5 doesn't exist.
Open Source BSD Support
>>55479151
The issue with tomb raider is people tend to be looking at it from a gpu perspective which isn't where that particular game benefits. Where tomb raider benefits most from DX12 is the cpu side (hell the main point of DX12 in general is to unfuck the cpu side of things to stop them stalling gpus). Its why if you look at AoTS with high end gpus you can physically see your cpu holding a gpu back because Intel doesn't make a cpu fast enough to keep up with the latest and greatest gpus.
In tomb raiders case this mostly means a boost for minimum fps on low and mid range machines but no real increase to max fps (sometimes a negative) so the average doesn't change, but you are getting a smaller delta between minimum and maximum which leads to smoother gameplay. For all this said though something is broken with tomb raider given its basically the only DX12 to exhibit such poor gains from the api - even with gamesworks disabled performance is poor.
>>55479160
DDR5 is better in every way, that's why NVidia uses it. The G stands for gimped.
>>55479255
in rotr that holds the game back is literally all the shitty gameworks filters that they are activaly trying to pass through dx12...
while its clear that doesnt work they are still trying
unless someone is to believe a 960 is faster than a nano
Pic related.
And on a reference card with its 6pin and connector
never AMD/ATI again in my life, it literally catched fire while i was idle
>>55479295
>unless someone is to believe a 960 is faster than a nano
In an overall sense, no it clearly isn't. Without knowing exactly how the game hits the hardware we can't say for certain but it is possible under certain situations a 960 is faster than a nano.
Makes me wonder what sort of dynamic clocking maxwell uses - Nvidia uses different clock rates for different parts of the gpu so overclocking only ups the clockspeed for the shaders. AMD uses a single clockspeed for everything which is why you see much closer to linear gains (and equally such a huge jump in power draw).
>>55479351
hey thats ok mine caught fire while it was on the box
>>55479372
Mine caught fire while being delivered.
>>55479371
well nixxess already have told us
>we took the popular hbao+ and created our own based on it
then they say
>we have vxao the evolution of hbao+
they just cant seem to understand that things that are developed for a certain api cant work on another
>>55479385
mine was forged inside mt doom...
one 480 to rule them all
>>55476396
zen is coming sometime in q4, not just 3 months from now, and it's already confirmed to be a paper launch.
>>55478567
TPU isn't reliable nor is it credible in the slightest. All of their benchmarks have inflated numbers for AMD cards that you won't find any other credible review from anand, tom's, guru3d, etc.
>>55472260
Yes.
>>55479493
if its coming on q4 pretty sure its not a paper launch......
unless they somehow made the mobo manufactures to create them months before so that they release them months later...
>>55479529
AMD has never released an architecture that went from tapeout to availability in less than 15 months. Zen taped out in january of this year. Therefore zen isn't going to be available before q1 2017, it may even be delayed further to allow AMD to build up stock before launching.
>>55479523
You are fucking right my good man.
>>55476993
Dis nigga knows whats up.
>>55479529
>unless they somehow made the mobo manufactures to create them months before so that they release them months later...
AM4 chipsets are being used for more than zen, it's also being used for excavator apus.
>>55472260
could have waited a bit for this but good buy
>>55477142
Middle of July from what I've seen anon. I'll be getting one too.
>>55477281
At the end of June AMD bought a software company, HiAlgo, and if rumors are to be believed they will have the sole job of making better drivers. I genuinely think AMD didn't have the resources to pour into driver dev until the RX480 took off.
It's time to be cautiously optimistic.
>>55478668
If I could just add another 4Gb and lower the TDP on my R9 290 I would. Sigh...
>>55479785
I have to admit, if you look at the known issues for AMD driver release notes today, they're basically nothing compared to known issues 18 months ago.
Though you'd never know anything improved, since AMD fanbois have said they were good all along - lol.
>>55479815
Run the silicon lottery and undervolt your 290 - GCN respondsn icely to undervolting because 1ghz is quite far outside the clocks the architecture was designed to run at. Its why polaris is so heavily redesigned because old GCN is a fucking furnace when trying to go over 1ghz.
>>55472286
There may be.
>>55478780
lol Rise of the Tomb Raider is DX12 and just had an update to fix CF/SLI in DX12 as well as add async (Maybe it will even be able to run both teams at the same time. Who knows!). Also Doom is due a Vulkan update at some point although they are dragging their feet over that one (maybe Nvidia already bribed them).
>>55479859
my 290x can get to 1200 mhz and its a reference board so stop spreading memes that amd cant overclock
>>55479898
You apparently can't read at all. I never said GCN cannot overclock - I said those high clocks are far beyond the architectures ideal range for sane voltage (and thus, power draw).
Fun fact: my own 290x does 1200mhz core but the voltage required is insane.
Lets not forget the fpunders edition for the 1060 costs 300 US, but as we have seen the aftermarket cards aare not going to tough msrp, and be 50+ more than the FE versions.
>>55479653
Literally going to just return my 480 after aibs come out so I can see how well it'll even work with muhgames
>>55479889
If Doom gets a Vulkan update and a linux client I'll buy it.
But then again, if DOOM got a Vulkan game engine I'd buy it (again) too.
Who is actually worthwhile on AMDs side when it comes to cards? Is it still XFX?
>>55479859
If I overclock I have to pump the +50 power, add +75Mv and only get 1100/1500 before things start to get hairy (and have to make the fan spin like a blower too). Above that and I get those little flashing squares and game crashes.
Can you recommend a safe undervolt to try? I'm curious to see if I can get the TDP down a bit without losing stability or default performance..
>>55479859
Not that guy but i have a very loud and hot PCS+ 290, can you tell me if it's possible to underclock and undervolt it so it won't get so hot and loud? And if yes, how can i do that?
>>55480070
>Can you recommend a safe undervolt to try?
Not particularly as its not my area of expertise - I know OCN had a thread on undervolting hawaii ages ago but fuck knows if its still there. Ultimately the simplest way is to dial back the voltage 10mv or so a time and run a benchmark (unigine valley is a nice one) and see how much it'll take before crashing. You might do well, you might crash if you even look at the voltage funny. Naturally dropping clocks will also mean less voltage so if you backed down to (say) 900mhz you might be able to get a hefty drop in voltage - though naturally you lose some performance.
>Above that and I get those little flashing squares and game crashes
You might be pushing the memory too hard - the 290/290x memory (depending on brand) generally doesn't like going above 1375 (sauce: the stilt). Its why (iirc) all 390/x cards use the same memory because thats a lot of strain on the controller.
>>55472231
Computer technology is NOT technology? The fuck are you smoking dumbass?
>>55472463
Why doesn't/g/ want to discuss tech?
>>55480042
Sapphire.
>>55480122
Two main approaches
1) Use the overclocking tool in the radeon settings to select a lower core clock target - the card should undervolt as required to sustain the clocks, but this is only works (iirc) in relation to its power states as defined by the bios.
2) Do it manually via something like msi afterburner or sapphire trixx. Manually set the voltage, fan curve (and if you need to core clocks, ram clocks and power limit) to limit exactly what the gpu can do.
BONUS OPTION:
3) Bios editing. There are a bunch of custom bios images out there for hawaii cards (The Stilt has made a load) but thats a bit more invovled and if you are stupid you can brick your bios - though a lot of 290/x cards have a dual bios switch so you will (in theory) only brick the first bios. I know the tri-x cards by sapphire have a bios switch that does nothing as both bios are identical.
>>55479785
Ever since Crimson drivers became a thing, AMD drivers have been pretty good.
Just check out HiAlgo's website if you wanna know what they're going to do. Its pretty interesting, doubly so for those who like to supersample.
>>55479954
how long is the return window? isnt 14 days the average?
>>55480250
the first option seems good for someone that doesn't know what he is doing like me, is there any chance i might fuck my shit up with that?
>>55477022
is there any other edition?
>>55479111
The G actually represents /g/
why should I buy a AIB 480 if the card can't needs 1,35v and liquid cooling to hit 1500mhz?
I'm almost buying a reference card...
>>55480768
because the reference cards don't meet the pci-standards at stock. If you're looking to overclock and fry your mobo, go ahead.
>inb4 it's fixed
Pcper already ran testing on the 480 post new drivers, they're all still drawing more than 6A from the allowed 5.5A
>>55480815
"""professional overclockers""" said that wasn't a problem...
>>55480840
Numbers don't lie and when you exceed the specified standard, you're opening yourself up for lawsuits and compensation claims.
It might be okay for all you know, but the raw numbers say that if you get a fried PCI-E slot you can sue AMD for damages.
>>55480815
it's a non-problem
if you examined every nvidia card with the same amount of scrutiny the amd cards are suddenly receiving, you'd find as many issues and more.
And what is the effective result? The 480 doesn't actually fry motherboards, the draw is over spec, but the spec should be increased from what it is to begin with and every gpu card in recent memory is going to have some amount of overdraw like this.
Mountains out of molehills being made by nvidia shills desu
>>55481001
Power spikes =/= overcurrenting. The issue is thermal. Things melt because they get hot. They get hot because they draw more current than they should.
>Nvdia shills
Please understand that your strawmanning doesn't change the fact that AMD cards draw 6-9A from 5.5A maximum allowed slot.
If you think about overclocking please don't buy a reference 6 pin RX 480, you're literally asking for a burnt pci-e slot.
>>55481035
please understand that your shilling doesn't change the fact it's a non-issue and both amd and nvidia cards do this specifically because it's a non-issue
>>55479523
>TPU isn't reliable nor is it credible in the slightest.
They are huge Nvidia shills too. A new Geforce card is out? Make a new item for every AIB, making the entire front page talk about "Geforce 960 was released". AMD releases anything new? They drop a quick note or a short article.
And if you go there often you'll notice that the article they leave on top on every Friday (making it the top headline for the entire weekend) is either something good about Nvidia, or something bad about AMD.
then they bitch about not being sent free samples of the Fury X.
>>55481081
>it's a non-issue
Can't you just admit that AMD fucked up and they should have just included a fucking 8 pin ATX connector?
Jesus fuck, it's already factually evident. You don't even have to protect your shitty pajeet tech. They even released a driver to fix the supposed non-issue.
Please kill yourself.
Will the 6gb and 192 bus limit the 1060 down the road?
>>55479930
>368W
>gpu max 68C
>vrm max 80C
Nice cooling you have on that. What is it, watercooling?
>>55472562
Thanks for posting this I been really interested in this stuff.
>>55474785
Project cars...fucking shit are you even trying not to look retarded!
>>55472192
Why would you buy either of those pieces of shit when you could pick up a 970 or 980 used on the cheap? Just got a 970 that was practically new for 130.
>>55472231
Your mother should have used the technology available for abortions.
>>55481147
I don't think so, 1070 and the 970 have the same 256 bus limit, I guess the new compressing technology can handle it better. Or nvidia doesn't care if the card is shit and is only putting on the market to take amd sells.
So will aftermarket fix that power problem in the 480?
>>55481519
Yeah, they just add a 8-pin connector and presto no more problem.
>>55480133
Well I tried. Valley ran fine but slowed down FPS. But RotTR just kept shitting itself even at the smallest -Mv. I've reset to defaults and will live with it for now.
8 pin will not help the power situation. See this for why.
http://youtu.be/Jq47qmwcus8?t=89
>>55481124
Hindsight is 20/20; they should've had an 8-pin yes, but nvidia would've had their shills nitpick some other minor problem as part of their PR campaign against AMD.
Yes, it's over spec
No, it doesn't fry motherboards
Nvidia does the same thing on their cards
The spec needs to be changed because it's outdated
You're only making an issue out of it because you're an nvidia shill, it really is that simple.
>>55481738
What does AMD's fuckup have to do with Nvidia shills?
Goddamn you're retarded.
>Spec needs to be changed because it's outdated
No it doesn't Literally the only card to break spec is retarded AMDpooshit 480
>>55481738
See
>>55481723
8 pin does not do what you think.
Misinformation sucks.
>>55481214
DP 1.3 and HDMI 2.0 and full DX12 support
>>55481723
The 'power situation' is over.
>>55472562
Why do GPU manufacturers other than Intel even exist?
>>55482363
>still over pci-e spec at stock after 'driver fix'
no
>>55482591
>pci-e spec meme
yeah nice one, 6pin can draw up to 300W with no problem you uneducated cunt
>>55472754
>>55472521
when you need ram, you need ram, there's no wiggle room there unless you want to turn settings down and fuck me if game artists where worth a damn anymore they would make their textures work at lower resolutions than as big as possible.
>>55477143
>>55477114
3d died because of post conversion content... basically shit that looked like hell, but was called 3d, when most people saw that they wondered what the fuck was the point.
however you watch real 3d content, i remember a rollercoaster vhs with blue and red glasses, that shit make me feel like i was there.
I am a full believer in 3d, but i accept that it died due to short sighted profits.
what made me a believe in 3d for games though was vr that i got to experience damn near 20 years ago.
when that shit works, its fucking amazing.
even for a racing game, just having the monitor on your head, and your head used as a new control method... that alone makes it worth having, even if you don't use it for 3d.
>>55479785
look at what that company has made... if anything they will make the amd drivers worse.
i hope to fucking christ that if they really do get charge of drivers i'm able to disable anything the implement.
>>55481897
>>55481214
new drivers, continued support... you know, all that shit that comes with amd cards and only the newest nvidia ones.
>>55481035
>don't buy reference cards
No fucking shit
>>55481403
It's the latter. The 1060 is a gimped 1070, that's why the BUS is screwed
>>55481519
It should. AMD already issued a driver update, a 3rd party 8-pin card should have no problems handling power
>>55472192
I'm curious as well, considering getting either the rx 480 or 1060 as upgrading from a 2gb gtx 670. However i have a i5 2500k (so pcie 2.0) and z68 motherboard. Would i bottleneck either of those gpus on pcie 2.0, should i grab an ivy bridge cpu? 3570k/3770k (older brother is upgrading soon and can buy his 3570k off of him), or will pcie 2.0 still manage to run them alright.
when will 470 be out?
>>55483705
Mid july, 460 early august, that is what I read during 480 reviews.
>>55483823
Thanks
>>55476913
I run near enough the same card. R9 270X and it is fine on high/ultra most games at 1080p.
>>55472192
How to attempt to prove to myself why I am buying a bag of wank gpu.
Just buy it and live with the bucket of stale cum
So Nvidia partners are planning on putting out their own 1060 designs by the 19th, right? Any word on when the 480 will be receiving similar treatment? I would assume before the 19th but I would certainly hope so given their head-start.
>>55472192
I'm waiting for the RX460. I currently have a GTX460 that has served me for all these years. I have had no reason to upgrade because all the games I was hyped for turn out to be shit.
I hope the RX460 manages to offer the same level of performance as a GTX 680.
>>55481154
Rajintek Morpehus.
>>55481001
>whatabautism
Guys, when are the non-reference rx480 cards coming out?
>>55475651
>if a card has better fps in a game than another card, it's the better card
not if the that FPS looks like 5 FPS because of stuttering. I will take smooth 40 FPS over 60 FPS that hands and jitters constantly.
>>55472192
Fucking knew it
All those fucking nvidia shill threads, but the 480 is still better
>>55472192
>* No SLI for the 1060 3GB version, RX 480 4GB/8GB both support crossfire
>wanting to crossfire a mid range gpu
please don't do that, especially on amd where they have terrible mutli gpu support.
>* NVidia has problems with HTC VIVE
the only problem is when you don't use the vive over hdmi, or are too cheap to buy a $40 adapter to go with your $800 HMD
also the 1060 is supposed to be about $300 which would mean that it should be a decent chunk ahead of the 480 that's 2/3 it's price. They're almost in a different price bracket entirely.
>>55487375
22. July
Where do I get a 480 now? Newegg has yet to fulfill my backorder since it released and I want to play gams
>>55487488
15% more performance claimed nvidia. So probably overall <10% better while aftermarket cards will cost about 80 shekels moar. Not worth for 5fps.
>>55487577
but worth to have not terrible drives, and not fucking retarded cpu overhead.
I don't even want to know what the cpu overhead is on the 480, because on my oced 4670k with a 7970 it would peg my cpu at 100% at times.
>>55487593
Drivers got better since the crimson drivers on my r9 390. Even better than whatever NVidia did the last 6 months
>>55487760
>Drivers got better
Yeah they did, if by better you mean started crashing a fuckload more.
any1 talking about cards they dont own should STFU
