• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Radeon RX 480 "Polaris" Launched at $199

Ah ok, thanks. I was hoping they would be challenging the gtx 1070. :(

Well, rumors are calling it a great overclocker, but even then it probably won't challenge the 1070 directly. The 1070 is a much pricier card. It may be a viable alternative if you only want 1080p or 1440p at 60fps and the 1070 may seem to be overkill. If that doesn't work out, then it will at least grab the lower end market share.
 
If true then price to performance:

Stock 480 (Ref Blower fan) - £230(Maybe) - 2096 = 12.6 points per £
Stock 1070 (KFA Blower fan) - £365 - 4290 = 11.7 points per £

Would 0.9 points be a big enough win in price to performance to warrant crossfire over 1 more powerful card, maybe not but both good price to performances for single card purchases depending on how much power you want.

EDIT: change to correct score and recal.

There is not a single 1070 below 550 EUR where I live. Most custom ones are over 600 EUR. Not sure where you getting £365 from.
 
I guess it depends on what games you want to play.

That's not really helpful as most gamers play a wide variety of games, and what about games that haven't been released yet?

And that's kind what I was asking, it's very hard to decide on what the price/performance is. There are always driver improvements etc. Things are always changing. Look at how many rows breakout about cherry picking benchmarks to suit an argument. So the price/performance ratio is going to depend on which site's benchmarks you use.

But, My point to Scramz was that you can't just pick one benchmark and use that to work it out. I do think you need to compare a good few games and benchmarks to get any real indication of the price/performance.
 
Well, rumors are calling it a great overclocker, but even then it probably won't challenge the 1070 directly. The 1070 is a much pricier card. It may be a viable alternative if you only want 1080p or 1440p at 60fps and the 1070 may seem to be overkill. If that doesn't work out, then it will at least grab the lower end market share.

I was hoping it would be an upgrade over my gtx 970 mainly in 1440p, but seems like it might be more of a sidegrade for me personally.
 
How do you work out performance? It's no problem when comparing Nvidia cards to other Nvidia cards, or AMD cards to AMD cards. But, benchmarks and games can vary quite a bit depending on which brand card you use.

Is there a benchmark that doesn't favour any brand of card? And, we hear time and time again that none of the benchmarks are any sort of indication of game performance.

And then what resolution or graphic settings do we use? There is always games etc that one brand card is faster at lower resolutions than the other. And some cards have better performance at some at higher resolutions.

Sorry, those are just my some of the thoughts I have when I see people talk about price/performance.


Anyway, back to your post, You did a calculation on one benchmark and claimed that the price/performance of the 1070 and 480 were the same. But one benchmark does not even begin to tell the story of the price/performance ratio. You are also working it out without knowing what the actual performance of the 480 is in said benchmark.

So you might end up been right, but, you can't make any predictions just yet.

Just want to add, that I don't think price/performance ratios work at all across different price tiers. Because the people that buy £400 cards will be different from the people buying £200 cards.

Price to performance = bang for buck = performance to cost ratio. Whether it's a £600 card or £200, you can work out which will give you more for your money. And as I said previously some people will looked to buy 2 to save money but match performance of a bigger card and it helps you work it if the implication of multi GPU is worth the saving over 1 big card.

I didn't say it is the same, I said nearly. I also said "if was to be true" implying it is not real and we don't know for sure yet.
 
Definitely going to grab one to tide me over until Vega, I'm running a 7950 and the 480 looks like an amazing buy to me, even if it's only on par with a 390 out of the box. Honestly, this card isn't really made for most OCUK posters, it's for the guys on a budget and frankly it looks like a grade A winner.
 
Last edited:
970 g1 1418mhz 1753 mem = 2831 g score
970 g1 1600mhz 2000 mem = 3089 g score

http://www.3dmark.com/3dm/12630609

http://www.3dmark.com/3dm/12630726

Any chance you or somebody else can give the score percentage increases from my stock 7950 to its overclocked score?

I just ran my 7950 at stock 800 core clock with a 1625 memory clock (to match the overclocked memory clock) to see the difference but im kinda rubbish with percentages :D

Stock 7950 core (800/1625) > http://www.3dmark.com/3dm/12631020 > GPU Score 1132 (Driver - 16.4.2 Crimson)
OC 7950 core (1200/1625) > http://www.3dmark.com/3dm/5257360? > GPU Score 2112 (Driver - 14.501.1003.0)

Only asking about this cuz people are saying the 480 maybe another good clocker like 79xx series..

..then im sure we could get a rough estimate of what a 50% overclock on a 480 could bring, as obviously people around here have nothing else to do until the 29th :p

I know i know, different arch's etc etc.. but its all in a bit of fun :)
 
Last edited:
Ah ok, thanks. I was hoping they would be challenging the gtx 1070. :(

Just an honest question, were you really expecting a £200 card to challenge a £400 card?

This card is a replacement for the 380/380x. Those cards are a tier below the 390x/390. So it will bring the performance of last generations mid range to the mainstream market.

What is muddying the water is the price of the 390 and 390x. These cards are available for around £210/£260, which seems to be the price range for the new 8GB 480 cards.

What people tend to forget is that the 390x and 390 have been reduced in price a lot since launch. And considering that these cards were basically just rebadged 290x/290's that were released in 2013. So the price comparison isn't really a fair one.

People in this thread are saying why would anyone buy this level of performance now, when it's been available for the past year? Well there are lots of reasons. People who have/had 200 series cards might not have bothered changing to the 300 series cards as they were mainly rebrands. Others might have been looking at 390/390x and realised they might have needed an upgraded power supply and didn't bother. Others might have just been waiting for the die shrink before upgrading.

The 480, by everything I have read so far, is going to use about 150w less power than the 390 cards. That's a huge improvement.

And if the 480 cards are at 390x performance levels they won't be a fail at all. I fact I am pretty sure they will sell really well.
 
Some peoples 7950s wouldnt even hit 1000 mhz though :) a few of us got lucky, one of mine did 1200 on stock volts and cooler, and 1315/1975 watercooled and increased volts. The 480 could be just as bad/good a clocker depending on your luck with the silicone lottery.
 
Any chance you or somebody else can give the score percentage increases from my stock 7950 to its overclocked score?

I just ran my 7950 at stock 800 core clock with a 1625 memory clock (to match the overclocked memory clock) to see the difference but im kinda rubbish with percentages :D

Stock 7950 core (800/1625) > http://www.3dmark.com/3dm/12631020 > GPU Score 1132 (Driver - 16.4.2 Crimson)
OC 7950 core (1200/1625) > http://www.3dmark.com/3dm/5257360? > GPU Score 2112 (Driver - 14.501.1003.0)

Only asking about this cuz people are saying the 480 maybe another good clocker like 79xx series..

..then im sure we could get a rough estimate of what a 50% overclock on a 480 could bring, as obviously people around here have nothing else to do until the 29th :p

I know i know, different arch's etc etc.. but its all in a bit of fun :)

1200/800 = 1.5x base clock speed with 2112/1132= 1.8657x performance.

That sounds too good to be true. A 50% clock speed increase results in a 87% increase to performance?? Maybe the drivers are a factor there.
 
Price to performance = bang for buck = performance to cost ratio. Whether it's a £600 card or £200, you can work out which will give you more for your money. And as I said previously some people will looked to buy 2 to save money but match performance of a bigger card and it helps you work it if the implication of multi GPU is worth the saving over 1 big card.

I didn't say it is the same, I said nearly. I also said "if was to be true" implying it is not real and we don't know for sure yet.

I know what bang for buck is, I know what you are talking about when you say price performance. I am just telling you that you can't base it on one benchmark. I suggest you read my post again.
 
1200/800 = 1.5x base clock speed with 2112/1132= 1.8657x performance.

That sounds too good to be true. A 50% clock speed increase results in a 87% increase to performance?? Maybe the drivers are a factor there.

The stock run i just did are on way newer drivers than the OC result (Dec 2014?) and the rest of my system is the same except for the stock run on Windows 10.. and 10 has always given me better results in any benchmark than Winodws 7 which i was on before.
 
yeah summat is wrong there, the score is normally effected by a smaller percentage than the actual overclock. Must be summat else on ya system messing it up.
 
Back
Top Bottom