All good, thanks for the correction Steve! Videos are still amazing ^^

All good, nice knowing you guys aren't the verge when responding to mistakes

We appreciate that you don't skimp on accurate content. Keep the great work (y)

Blaim Age XD

Oh no! Nvidia got to him, they've threatened him with 3.5 bullets. No problem mate, everyone makes mistakes. I will watch it again gladly.

Never get's old :D Thanks mate!

3.5 geis incoming

You couldn't be bothered writing "mate"? you had to shorten it down 😂😂

@OddFuture nvidia did 9/11

I actully looked it up, after the first video. R9 290 was a GTX 770 competitor, R9 290X was the 780 equal. Nvidia then released the 780Ti a month after and got the performence crown back. After that AMD was never the same.

No, 290 was 780 competitor and 290x was 780Ti competitor, the reason why 970 became that was because AMD didnt released 390 sooner than 9 months after maxwell, get yo facts right.

@@MaggotCZ Your the one, who need to get his facts straight, or just use your brain. The 290X can't be a response to a 780Ti, given the latter was released after! The Maxwell architecture was released so soon, because TSMC 22nm never went anywhere, so Nvidia used 28nm instead of the 22nm they had planned, which in turn led to Pascal, which itself is the OG Maxwell, just on 16nm, instead of the 22nm.

i never said 290x was response to 780Ti i said they comepted which is true, there was 780 and TItan kepler, 290X rekted titan so nvidia released 780Ti.

@@MaggotCZ Again, not true. The Titan card was orignally meant to be the one to compete with the 7970 all the way back in 2012, but Nvidia realized they dind't need to release it, as the mid range chip(GK104 was more than capeable of beating the much more powerfull 7970). The Titan card was then released later, as a way for Nvidia to expand their lineup(Enter the ultra high end market as we know it today).

The CPU is a solid "huh?" For me dog

R9 390 wins if were only looking at stock speeds, but if you introduce OC into the mix it's a clear 970 win. R9 Fury is beaten in most games by 980, and suffers from low OC headroom, but will without a doubt age better over the years, than the 980. Same story with the R9 Fury X and 980Ti, not to mention the aio cooler employed by AMD is pretty noisy.

​@@Orcawhale1 R9 390 OCes almost the same as a 970. It also has more VRAM which helps maxing textures in games. It can literally enable settings 970 cannot. You can keep all other settings equal and crank textures to max with a390. lol the 980 is not that great. To be fair, Fury also suffers from its low VRAM amount, but lol. Fury X will lose to 980 Ti. Here we agree. No Nvidia AIB can match Sapphire though.

​@@CharcharoExplorer Yes, the best ones do(Sapphire Nitro cards for the most part). But the average R9 390, most certainly do not(nor are most aib coolers capeable to cool the Hawaii chip properly, looking at you XFX). And in regards to the settings? While textures is nice, the chip itself isn't fast enough to get proper fps, so while your correct, the real world it don't really matter. 980 is by no means a great card, but it's a much better one than the Fury card. Also, plenty of boardpartners has matched Sapphire's brilliance. Just not on AMD's side of thing, since the 7900 days. But on the Nvidia side, you can find plenty of great board partner cards(The RTX 2060 Strix for instance).

@@Orcawhale1 Yes, the chip is capable of driving Textures on Ultra. It will struggle with geometry and shading, but textures, provided you have bandwidth (390 Has it) and VRAM (390 has that) are not that demanding. So that part of your comment is not true. The reason certain people use that argument is because they use presets. I do not use presets. I am more than happy to use mixed settings. And the 3900 will match the 970 in *all* of them, except textures where it will win by a lot. The 980 is slower than a Fury. It doesnt even really win tessellation due to vertice reuse in Tonga and Fiji. If it wins something over the Fury, it is a slight geometry win or a driver bug. Otherwise its slower. OC to OC as well. And *no* 980 is equal to the R9 Fury Nitro/Nitro OC or Tri-X. Powercolor 390s are also good BTW. ASUS Strix is overpriced, but it is generally very good. It is no Nitro+ though.

I had the 6gb and did the same thing so I had to upgrade my gpu, 8gb to 16gb ram and put in a new HDD just to sort out the problem 😂

yep gotta lower textures to medium with that 3gb

Yeyeye, fuck off bro, my gtx 970 could literally heat both my and my neighbours place at auto RPM

Yeyeye, fuck off bro, my R9 290 has never gotten past 60 degrees while gaming with my custom fan curve

@@kire929 Chill both of you guys. Saying rude things doesn't solve a problem. Acting violent by cursing doesn't help with anything basically.

@@AdamBror the earth is a triangle, gravity doesn't exist

@@kire929 ok man.

Yes. The R9 290 was the direct competitor to the GTX 780 not the 970.

Becouse 290 is faster than 780 and 390 killed 970

Yes. The sad part was that 290 came 1 year before 970 and it was effect by mining back then. You could get 290s for $200 in the summer of 2014 before 970 came out in September. People did not care. They did love 780 performance for $330 though.

Was thinking the same thing, Maybe HUB didn't have a 390 on hand? Not sure exactly, maybe trying to stick to similar VRAM limits to keep some sort of parity in modern titles. Hard to say. Would have thought the 390 was the natural competitor though.

@@techforge-Nate Also 290 lost hard to 970 at 1080p. 1440p was where AMD had the advantage.

same

how bout now

im still using r9

A N I M E N I M E

What temps are you getting with it?

Kind of, the 390 is just a refresh of the 290 like the 580 is to the 480. They are just the same chip but clock higher. But yes it will edge out the 970.

The 390 had 8gb ram not that it would make much difference, it is marginally faster which would change the result.

290 has 1250mhz ddr5, 390 has 1500mhz ddr5 ....Noticably faster.

no and in his original review on techspot he found the 390 and 290 had no performance difference between them. even look at his benchmarks for 2016 games and more recent ones there is at maximum 2 fps difference between the 290 and 390 and 290x and 390x. the 300 series cards should never have happened and amd charged $100+ for just 4gb more vram. not worth it and he said that too in his review.

@@supermadkid Sure 390 with higher core clock and 20% faster vram must have same results with 290. Well no id doesnt have now fuck off. And lets forget 290x and 390x cos they probably didnt exist ( I did mining Eth with 2 390x but its my imagination probably)

lol

"I'm shitpascal I get worst in time so they'll make you buy rtx" -1000 series

@@ryusei9941 This myth is as bad as the AMD cards have better colours one. I've got an old 760 and it runs games exactly the same as it did before.

@@letni9506 they make 1000 gen worst every week, I DO feel it . And amd keeps improving :(

@@ryusei9941 Yes, look for example at the RX 580 defeating the 1060 in DX11 games. It just does not make sense being that they are comparable cards. Definitely Nvidia is doing downgrade to pascal.

more power consumption,the refrence design was trash running at 94c and sounding like an jet engine(had it myself) while aftermarket desings were not easy to come by and the 970 overclocked like a champ.I had both cards both refrence and the 970 was much better

I had got 2 290x sapphire 8GB cards on release and I have to say there were a lot of issues with them compared to the 970's. I loved them a lot and they looked awesome, but the power draw was no joke. Needed a 1200w PSU to get crossfire to work. The heat they kick out is insane as well. Only build I ever did that had higher power draw was my 1080 ti SLI rig. Even then the temps were a lot lower. They were good cards, but AMD was pushing their limits in a bad way. Overall I can see why people liked the 970. It ran far cooler and was a lot easier to work with in small builds. I always get the best cards from both teams and compare them when I can. AMD has good stuff, but Nvidia has the performance crown for a reason. Not a fan boy though, they are both evil evil companies.

@@Skwisgar2000 How is AMD an "evil evil company"?

@@DirtyPoul Nvidia may get a lot more press for the sketchy stuff they do, but AMD is no saint either. They have released poorly designed GPU's and obviously misleading products in the past just like Nvidia. Their bottom line is money. Just like any major tech company. In my book that means you, the customer, are nothing more than dollar signs to them and as I see it that makes them just another greedy money obsessed company. You of course are welcome to disagree.

@@Skwisgar2000 Releasing a poor product does not make them evil. Neither does making misleading products. Sure, it's bad, but I wouldn't call it downright evil on the scale of the monopolistic policies that Nvidia, and especially Intel, have done in the past. Those are the truly evil and greedy stuff. Not a bit of misleading advertisements, and certainly not bad products.

rahahaha :D Yeah they really rag on me there!

@@Hardwareunboxed which server?

@@mirshia5248 probs HUB's Patron server.

I doubt.

That goes to the 9700 Pro. But yes.

JohnnyNismo aren’t they relatively new? I’m considering getting V56 this week. :D

@@MalfunctioningAndroid they came out in 2017.

JohnnyNismo time does fly. It’s still pretty new in my mind. Should have googled it beforehand. At least it made some old fashioned conversation. :D

@@MalfunctioningAndroid I don't blame you. Vega is still AMD's most up-to-date generation of GPU technology. They really don't feel old at all.

This fine wine tradition never made sense to me. Sorry, but what's the point of buying a product that is inferior at launch to other options and inferior later on due to performance gains from new products that are now significantly cheaper.

used 1080ti with warranty.

The question is, 2060?

You can use something like GOOGLE to check this, as you can see in all the reviews of the GTX970 when released, the R9 390 is NOT present only a 290, strongly suggesting the R9 390 was NOT released when the GTX970 came out. www.anandtech.com/show/8568/the-geforce-gtx-970-review-feat-evga/8

It definitely should. Or, it should be the 290(X) vs 780(Ti).

390 is similar to 290. Maybe 5-10% better.

More power to him.

What is your PC specs? I'm guessing it's faster.

@@Armytoast Me? Why?

Wait for navi after computex. 2060 is better value than 2070 at least in america, but if u can wait for AMDs announcement it could be worth it to see what they offer.

Is it overclocked? 1050Mhz here with blower fan!

@@86FCS Yeah. I usually have the core set to 1075 and memory at 1300. I've had the clocks higher, but this is 100% rock solid and good enough for me. When it wasn't being water cooled it would heat throttle on stock settings. :s

My Fury Nano lands right above the 590 and about even with the 1660 (non-ti) as long as 4GB doesn't come in to play (wolfenstein). Fury and Fury X should be equal to a touch faster.

@@LohTec it REALLY depends on game and resolution...as they are AMAZING at 1440p, but just ok at 1080p, as the memory bandwidth doesn't really get utilized at 1080

@@Tallnerdyguy true, but in the end they are just downclocked Vega.

Omg I was stuck using gt9800 now using 50$ r9 290

Try 2013 with the 290. I bought one at release and chucked a waterblock on it, and it was still cheaper than the 780. A few years later, a 4K display went on sale and I upgraded to it from a 120 Hz 1080p display. One of the best upgrades I ever made. Sure, it didn't run it at max settings, but everything looked much better on medium-high without AA than things did on ultra with AA at 1080p. It was worse for competitive games, but I'm not a huge fan of those anyway. It was an amazing combo for years to come until I upgraded to a used 980Ti back in 2017 when Vega was released. I got that one with a waterblock for 80% of the price I gave for just the reference 290. An absolute bargain! I would've preferred AMD again, but for that price, I can't complain.

What does that mean? Do you want to know how good the different cards do at 3440x1440? It's somewhere between what the cards do at 1440p and 2160p.

@@DirtyPoul what you mean what does that mean? I'm using 970 and got ultrawide monitor capable of 120Hz in 3440x1440. I'm currently on the market looking for new graphics card and looking for recommendations (would be nice to see an "upgrade guide"). Do you understand now?

@@d4r1us2 No, not really. Why do you need a specific upgrade guide? Isn't it enough to see how the different cards perform in different games under different resolutions? You can divide the average numbers (or the numbers from the games you play the most) with the price to get the performance/price in your region. Isn't that a guide enough in itself? What else do you need?

@@DirtyPoul no, not really. Following your logic you could calculate all resolutions as far as you would have results for 1080p, you could just multiply that, right? who cares about the benchmark when everyone can just calculate that. Steve could do then more videos because tests based on 1080p resolution would take several hours less. :) now seriously, I asked Steve because these tests are not widely available and I know Hardware Unboxed got equipment, parts and monitors to thoroughly test it, as usually they do. That kind of test would be something different to what everyone else is doing, it could be good for the channel as well. I bet I'm not the only one that struggles with most benchmarks as I don't see my resolution and have to, as you say, calculate it based on averages and would be happy to see proper test. Answering your questions, this what I would like to see and no, this what you said isn't enough

@@d4r1us2 Okay, so you're asking for tests being done in 3440x1440 resolutions? I see your point, but given how close that resolution is to 1440p (just 34% larger), I don't think it's necessary. Multiplying the 1080p score wouldn't make sense since a doubling of pixel count does not result in a halving of performance. Far from it. I see your point though. It would be useful to have, but I think you could just decrease the performance numbers from the 1440p benchmarks by about 10% and it would be close enough to the numbers you should expect from 3440x1440.

PimptatoPCs what power supply do you have? If it’s an 80+ gold over 650 Watts; get whatever’s cheaper; if it’s 80+ bronze/white and under 550 Watts, get a 1070. 600 Watt PSU depends on the quality of the power supply.

I have a Thermaltake semi-modular 850W 80+ Bronze. I'm not too concerned about power draw, more concerned with FreeSync compatibility. I'll be upgrading my monitor to a 1080p 144Hz FreeSync and don't want to hunt down a monitor that works 100% with Nvidia hardware. That's the reason I'm leaning toward Vega.

Choose one with good warranty.

obligatory wait for navi and hope it drives all other prices down. or, wait until the upcoming AMD investor's meeting (15th, i think?) and see if gpu roadmap is laid out to see if waiting makes sense. most rumors say navi could be announced at computex. but no telling when the release. q3, tentatively? i guess 3 months of waiting doesn't meet your "very soon" criteria, though. /shruggie

I forgot that Navi is coming soon, that should help prices. I'm looking at a mini 1070 for about $230 and a XFX reference 56 for about $250. I just don't know how much cheaper they can get. Both of the cards are kinda limited because of the cooling (one reference and one single fan), but I would rather water cool the 56 down the road rather than the 1070 mini. I should probably wait and see what happens with pricing when Navi hits. Maybe I'll be able to pick up a 1070ti for closer to $259 then.

Oh no. Don't mention the FX series when Steve is here!

​@@DirtyPoul Its fine, I actually bought some and have several years of real world experience with them and actually know what they're good at what they aren't. its hard to be objective when you get 2 things for free and one is more expensive than the other and you spend 1 day with them. A normal reviewer will grab a motherboard (with 4+1 vrm) overclock with both fsb and multi like 225x20 to 4.5Ghz dumping in 1.4v and end up with a system that throttles and performs nowhere near where it should with wierd speeds like 2250NB and 2250HT. Any real user gets a half decent board with a 6+phase (and most people don't even know how many phases they have), only puts in enough voltage for what they need so more like 1.325 (on average) with 200x22.5 and a 2400NB and 2600HT clocks, it doesn't throttle and would kick the other systems ass.

@@tomstech4390 Interesting. I didn't know that some of the issues could be fixed. Interestingly, the FX series has aged rather well too, and have closed the gap in some areas because of greater reliance on more cores today.

​@@DirtyPoul Its not that the hardware his improved (obviously) its just people have changed. Reviewers have maybe 6 hours to test a whole new architecture and do thier testing, so they'll overvolt to the point of slowdown, look at ryzen1 reviews and you can see how many did 1.4v for 4Ghz (180w?) when on the same chips you could probably get 3.9Ghz on only 1.35v (155w) which is a miniscule difference in performance. actually if the 4ghz 1.4v chip is pushing the board too much the 3.9Ghz chip might perform better. So most reviews need to be taken with a pinch of salt and compared against each other (see the multi core enhancement launch problem where reviewers were getting diffent results). Hardware testers like myself and others (who are pretty much invisable on youtube because who wants to subscribe to and watch a video of me rebooting 50 times tuning uncore and nb settings?) will spend a year on an architecture finding out what things make a difference and what things dont, What are the requirements to run certain hardware a certain way, Can you run a 9900k at 4.7Ghz on a gigabyte z370P D3 (its on the supported list but its also £86) and so on. When pentiumD launched and athlonx2 launched people didn't get it then either (okay pentiumD WAS shit) "but games only use 1 core" sure at the time they did but windows xp still had multi core support so the game might run on 1 core still but windows will run itself on the other, when an athlonX2 was 2.2Ghz and an fx61 was 2.6Ghz yeah the game would run faster on the single core....but what if you ran the athlon x2 at 2.6Ghz? noone tested it so people kept buying single cores. Now comes vista/7 and the core2 duo and phenomX4 and core2 quad "but the e8400 can go to 4Ghz while a core2 quad is what 3Ghz? games only use 1-2 cores anyway", sound familier? Completely ignoring the fact that the same architecture core2quad and core2 duo could obviously overclock the same and get the same game performance and games would actually use all 4 cores, but because they performed the same people though games weren't using them. No one has ever heard of task manager it seems. As if having 32gb of ram would make you faster than 16gb of ram? now with windows 7 we see 6 core phenoms and 6c12t i7 980x's etc "games only use 2-4 cores" 98% of the time said from people who only have 2-4 cores, thats a bit like taking sex advice form the pope.. Go and look and you can see screenshots from 3930k users and such with 6c12t in a game and you can see its using more than 4 threads, as far as an OS is concerned those 12 virtual cores are all real so if its using 7......its got 7 threads of instructions to assign to cores, end of. When ryzen1 first came out and then intel replied with the 8700k people still said "games only use 4 cores" but they caved in and bought them anyway. funnily how intel released 6 and 8 cores and only now people are starting to say 4 is borderline. As if before it wasn't allowed until intel and gabe newell meet on a hillside and sacrifice a disc copy of half life 3 and say "let it be done". See many people with a 9900k or 2700x running far cry3 or hitman absolution with with OSD open? No, because no one plays those old games so how are we to know how many they *actually* used back then? ......Well I'm testing this stuff but I cant produce videos because of stagefright (im fuck ugly anyway) so no one will ever know. But theres always been people saying "games only use x cores" but no one, even youtubers actually knowing how to test to show what they actually use. 1) get a second screen, 2) check current core usage in task manager set to update fast as possible AND OR msi afterburner showing cpu thread usage, 3) make sure you have nothing else running which could give fales reading, 4) launch a game and actually look at usage over a 30-60 second period. But apparently linus and HUB etc are not 1337 enough for those mad hax. TL:DR FX isn't getting better, games are just getting more demanding which crushes weaker i3 systems FX was priced against. Because manycore is "mainstream" (£550 is mainstream?) peoples minds are a bit more accepting of it now. Its always performed well enough in capable hands on capable boards.