• 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.

AMD Radeon 300 Series Won’t Be A Rebrand, New GPUs Coming in June

Wrinkly

Wrinkly

Wrinkly

Permabanned
Joined
31 Aug 2013
Posts
3,364
Location
Scotland
The AMD Radeon R9 300 series will reportedly contain no re-branded graphics cards from the R9 200 series. An entirely new lineup is coming at Computex. The report in question alleges that AMD’s original plan was centered around introducing a new flagship Radeon product this month and rebranding a few other R9 200 series products to R9 300 series products. Essentially making the R9 290X the new R9 380X and the R9 285 the new R9 370, so on and so forth.

AMD Radeon R9 390X To be completely honest, that’s what we had believed would happen as well. However according to the report there has been a change of plans and a massive one at that. Relatively recently AMD decided to take a different high-risk high-reward route with the R 300 series instead of the boring rebranding plan. Senior management and senior engineering heads at AMD have decided to introduce an entirely new line-up of GPUs, with no re-branding of older chips.
AMD Radeon 300 Series Won’t Be A Rebrand, Brand New GPUs Coming in June

I’m going to side-track for just a bit so bare with me. AMD’s Gaming Scientist Richard Huddy had stated on several occasions that “all of AMD’s future GPUs will support FreeSync“. The astute among you will deduce that for that to happen AMD simply cannot re-brand older Radeon 200 series GPUs to 300 series products. Purely because a significant portion of the R9/7 200 series does not support FreeSync. According to Huddy adding FreeSync support requires updating the display controller integrated into the GPU die itself, which would warrant a re-design of the die, something that’s extraordinarily expensive.

Currently AMD’s Hawaii (290 series), Tonga (285) and Bonaire (260 series) GPUs support FreeSync. This meant that AMD would have three chips which it can choose to re-brand without breaking FreeSync compatibility. There are however a few pitfalls to this strategy which may have led AMD to forgo it. One of which is the considerable gap between the Tonga chip, which is essentially an updated Tahiti (280X/280/7970/7950), and the Bonaire chip. This meant that AMD would have to design a new GPU to fill this gap anyway because it can’t use Pitcarin (270X/270/265/7870/7850) since it doesn’t support FreeSync. Another pitfall is that AMD cannot significantly improve the performance i.e. competitiveness of rebranded GPUs.

Now that we have successfully discussed what did not happen in the past and why, let’s talk about what will happen in the future. Just a couple of days ago AMD teased both its fans and the press with a yet unreleased, unannounced, ultra-enthusiast class Radeon graphics card. Which, AMD told us, was actually used to power several systems to demo Oculus VR at the GDC show floor.
Unfortunately that’s all that we’ve managed to get from the folks at AMD. Fortunately however we were told , in addition to several other journalists, that “much more” will be revealed about this mystery graphics card “soon”.
System housing AMD's unannounced card at GDC. Image credit @PCPer.com

System housing AMD’s unannounced flagship R9 Radeon graphics card at GDC.
Image credit @PCPer.com

Back to the report by Kitguru. According to their information AMD’s preparing to introduce the R 300 series lineup at Computex in June. We don’t know for sure if this also includes the fabled Fiji XT AKA R9 390X. Because what we have discovered, repeatedly, is that this card’s actually ready. First when we found out that the board had passed the certification phase. Later AMD themselves said that they were “putting the finishing touches” on the new series. Once again when we had exclusively told you that the cooling design for the board had been decided upon. And finally we were reminded once more that the new chip is ready when it was used to run several public virtual reality demos at GDC. AMD can’t simply run several, performance intensive, demos publicly on a brand new chip unless it knows that it’s in tip-top stable from.
AMD Radeon R9 390X

Leaked cooling shroud allegedly for the AMD Radeon R9 390X

I could certainly imagine that AMD still has got some work to do on the rest of the lineup since their development has allegedly begun later, but there’s evidence that Fiji XT / R9 390X is ready for prime-time. So is it true that AMD’s holding this card close to its figurative chest until inventory levels of R 200 series cards in the channel are in a more healthy level ? That’s a tough question to answer without delving into the semantics of business practice. At this point we’re merely hoping that when AMD says “soon” it doesn’t mean three months later. This space needs more aggressive competition, the sooner the better.
Click to expand...

Source - http://wccftech.com/amd-r9-300-series-not-rebrands/
 
I find this article wording a little odd... like, AMD "suddenly decided" to release a new GPU architecture? No, something like this was clearly planned for a long time, not a snap decision.
 
AMD R9 390 Series To Launch Alongside Computex 2015

AMD is preparing to time the launch of its next-generation Radeon R9 300 series with that of Computex 2015, in early June. The company had earlier planned to launch some products that are essentially price-adjusted rebrands of existing ones, such as the R9 380 series (being rebrands of R9 290 series on a slightly improved silicon), and the R9 370 series (being based on the "Tonga" silicon); but has decided to launch the two along with its flagship R9 390 series, based on a brand new silicon, around the same time. AMD's answer to the GTX TITAN-X from NVIDIA, the R9 390X will feature around 4,096 stream processors based on the Graphics CoreNext 1.3 architecture, and will implement an HBM (high-bandwidth memory) interface, with bandwidths in excess of 600 GB/s.
Click to expand...

http://www.techpowerup.com/210453/amd-r9-390-series-to-launch-alongside-computex-2015.html
 
Great, it'll fly :cool:

Although tbf I don't do a lot of gaming in the Summer, so will pick up a pair once they've been out a few months :)

Hopefully they've got a trick or two up their sleeve to counter the memory limitation...
 
"Another pitfall is that AMD cannot significantly improve the performance i.e. competitiveness of rebranded GPUs."

I'm sure that has been known for some time but we've still had to suffer cards like the 285 and 960 which offer no real significant gains on previous generations. I guess this stuff is quite typical of a rumour site like this
 
Dunno why everyone's expecting a hotter card than the 290??? If you actually bother to look into the tech you will realise HBM requires significantly less voltage than GDDR5 which will result in a lot lower temps.

I'm not convinced on the AIO rumor either; and the above article basically only confirms the release time and also no rebrands.

Would be nice to see some actual stats on the cards, I think with the TitanX reveal AMD are just keeping quiet now to wow everyone, sometimes silence is better than flooding the tech sites with tons of info, better to keep quiet and just drop the bomb when they are ready, however this is AMD so it could just be silence due to yet another inferior product :(
 
If they wait till June, and the 380/370 are still just rebrands, I will buy an nV card out of spite :p

They better be all new, and they better be freaking unbelievable. Otherwise love is over.
 
At least for those that can't afford/justify/want the Titan X, they will have something else to choose from. The HBM memory is what interests me the most and how well it adds to performance.
 
SiDeards73 said:
however this is AMD so it could just be silence due to yet another inferior product :(
Click to expand...

I think AMD lost when NVIDIA made Maxwell. Some people are ignorant to the significance of less power usage but this is why NVIDIA clearly have the upper hand. Less power = less heat waste = faster GPUs while not burning the house down. NVIDIA can release GM200 cards with ease but AMD have a hard time taming a 28nm beast with 4096sp, even if the ram voltage is lower.
 
FoxEye said:
If they wait till June, and the 380/370 are still just rebrands, I will buy an nV card out of spite :p

They better be all new, and they better be freaking unbelievable. Otherwise love is over.
Click to expand...

Why are you expecting the 370/380 cards to be anything other than rebrands with tweaks?
 
SiDeards73 said:
Dunno why everyone's expecting a hotter card than the 290??? If you actually bother to look into the tech you will realise HBM requires significantly less voltage than GDDR5 which will result in a lot lower temps.

I'm not convinced on the AIO rumor either; and the above article basically only confirms the release time and also no rebrands.

Would be nice to see some actual stats on the cards, I think with the TitanX reveal AMD are just keeping quiet now to wow everyone, sometimes silence is better than flooding the tech sites with tons of info, better to keep quiet and just drop the bomb when they are ready, however this is AMD so it could just be silence due to yet another inferior product :(
Click to expand...

It's silly to expect either really. For eg, due to proximity HBM requires the cooling system for the die to handle its heat output in contrast to gddr5 vram on the PCB. If you save power somewhere such as will be the case with HBM vs current gddr5 you can just keep your power use low or use it to drive more SP's etc. Not knowing that much about the card I wouldn't be surprised if they did come with an AIO and equally unsurprised if it didn't.

Silence can work but in the modern age AMD could really do with learning how to uncommittedly tease the marketplace - the way Nvidia do better and keep people looking forward to what they do next.

RanxZy said:
I think AMD lost when NVIDIA made Maxwell. Some people are ignorant to the significance of less power usage but this is why NVIDIA clearly have the upper hand. Less power = less heat waste = faster GPUs while not burning the house down. NVIDIA can release GM200 cards with ease but AMD have a hard time taming a 28nm beast with 4096sp, even if the ram voltage is lower.
Click to expand...


Advances in the industry pretty much guarantee greater perf with lower power usage as time goes. Hence any recent mid end card achieving the same performance level of previous gen higher cards. I don't think (or at least I hope) people are too ignorant on the fact that lower power (however it comes) = potential to spend that power headroom on more performance. It's certainly not a recent revelation to the industry. Nvidia had great timing, intentional or not, and made a good choice in developing a reduced feature gaming orientated arch and then followed up with implementing in it fine grain power management. Ball out the park combination with clock boost tech, especially in the current situation being parked on 28nm, for producing strong gaming performance with smaller and cheaper to produce dies.

I see it sometimes said that AMD/Nvidia must collude because of how closely their performance increases generationally. I think people are ignorant (if you like) to the fact that perhaps both companies have good competent engineers and design teams, and that what you can get out of a set amount of silicon area and when subjected to similar external economic/manufacturing constraints is going to get you similar performance levels year by year by year.

Now this is a point or perspective I feel that is somewhat missed but perhaps understandably on an enthusiast forum (although in some respects you could argue OCUK is not). If you are not on the bleeding edge upgrade cycle and instead upgrade every X years simply because you need more grunt eventually as games become more graphically demanding; regardless of vendor and even when switching between vendors the performance increases are the same. The fact that different companies release products out of sync and different perf points (leapfrog performance and releases) only gives the perception that this isn't so. They are all data points closely aligned around the 'upward performance' or 'power/perf' trendline.

TL;DR There is no real victory, neither side has lost or won. The consumer should be aware regardless which they end up purchasing. I'm off to eat lunch.
 
Last edited:
Though in a Capitalist economy it's all about overheads and ever increasing profits. If the profits go down or not up enough even if the company is making healthy returns it can go out of business.
 
You must log in or register to reply here.
Back
Top Bottom