Tuesday, July 28th 2015

AMD Readies Radeon R7 370X to Counter GeForce GTX 950

AMD is reportedly giving final touches to the Radeon R7 370X, to preempt launch of NVIDIA GeForce GTX 950. The card will be based on the "Trinidad XT" silicon, and will max out components physically present on the chip. This means that the card will feature 1,280 Graphics CoreNext stream processors, 80 TMUs, 32 ROPs, and a 256-bit wide GDDR5 memory interface, holding 2 GB or 4 GB of memory. Leaked screenshots that disclose these specs suggest that AMD will carry over clock speeds from the R9 270X, working out to 1180 MHz core, and 5.60 GHz (GDDR5-effective) memory, working out to a memory bandwidth of 179.2 GB/s.
Sources: VideoCardz, Expreview
Add your own comment

34 Comments on AMD Readies Radeon R7 370X to Counter GeForce GTX 950

#1
SNM
In GPU Z Result pic it shows AMD Radeon (TM) R9 370 Series and you guys have entitled it as R7 370, what is reality other than that every detail in GPU Z is exact copy of the R9 200 series....would it be the same R9 270X with just a new name...???

The original line up never shows a card naming R9 370 Series so it should be R7 370 Series I think...
Posted on Reply
#2
HumanSmoke
btarunr said:
. Leaked screenshots that disclose these specs suggest that AMD will carry over clock speeds from the 270X, working out to 1180 MHz core, and 5.60 GHz (GDDR5-effective) memory
The R9 270X has a frequency of 1000MHz (1050MHz boost). If those GPU-Z shots are legit then the 270X shown is the PowerColor Devil 270X.

In the end, it makes little difference I suppose, it's still a recycled three-and-a-half-year-old card with little marketing appeal - no FreeSync support, nor TrueAudio ( if TrueAudio is still a thing that is).
Posted on Reply
#3
the54thvoid
It's giggles and all to think the term 'readying' is applied to a recycled 7870. Don't get me wrong, great little card but yeah, old tech being smothered in lipstick and given a latex red dress with a plunging neckline. C'mon AMD, you should have delivered more options on Fiji.
Still, Nano is still to show so perhaps innovation (of sorts) there will he a bit more appealing.
Posted on Reply
#4
hojnikb
If this is over 150$, then AMD can go f*** themselfs.

270 retailed for 179$ back in 2013
Posted on Reply
#5
RCoon
Gaming Moderator
the54thvoid said:
It's giggles and all to think the term 'readying' is applied to a recycled 7870. Don't get me wrong, great little card but yeah, old tech being smothered in lipstick and given a latex red dress with a plunging neckline. C'mon AMD, you should have delivered more options on Fiji.
Still, Nano is still to show so perhaps innovation (of sorts) there will he a bit more appealing.
For all we know Fiji may scale down terribly. It may be that Tahiti/Pitcairn provides better performance at a much lower cost when scaled down in comparison to Fiji. Not every architecture has a full top to bottom presence due to some scaling better than others. Either that or Fiji is probably A) expensive, particularly to cut down, and B) in very short supply for cut downs to be a thing.
Posted on Reply
#6
Prima.Vera
Is this faster than my good ol' HD5870 ?? Dead curious.
Posted on Reply
#7
jabbadap
Heh, so nvidia can launch gtx950ti to counter with this. Hopefully we see proper price war this time. Although that pitcairn misses lot of things compared to nvidia or even amd's own newer cards(nvidia has hdmi2.0, hevc and newer amd have freesync, trueaudio). But in pure gaming performance it could be close call between the two...

Then again if this is more castrated tonga, it could be competitive vs nvidia's offerings.

Edit:
Prima.Vera said:
Is this faster than my good ol' HD5870 ?? Dead curious.
Yes, lower clocked same chip pitcairn hd7870 is all ready fair amount faster than hd5870:
http://www.techpowerup.com/reviews/AMD/HD_7850_HD_7870/26.html
Posted on Reply
#8
GhostRyder
Well while I would love a new chip this chip still competes quite well which is probably the reason they felt no reason to replace it. The lack of Freesync support on this lower card is a bit sad but to me its not really something even necessary because its performance would not be able to keep up with where most of these monitors are in terms of specs (IE 1440p 144hz). I mean while its nice to have the option, would you really invest in an R9 370X and a 1440p 144hz Freesync monitor? It would be like investing in a GTX 750ti (Or similar) and then buying a Rog Swift (1440p 144hz G-Sync monitor), kind of pointless as unless you drop the setting super low your not going to be able to drive anywhere near that.

I would like to see if the card has evolved at all, would be nice to see theses things in OC models being around 1300mhz.
Posted on Reply
#9
the54thvoid
GhostRyder said:
Well while I would love a new chip this chip still competes quite well which is probably the reason they felt no reason to replace it. The lack of Freesync support on this lower card is a bit sad but to me its not really something even necessary because its performance would not be able to keep up with where most of these monitors are in terms of specs (IE 1440p 144hz). I mean while its nice to have the option, would you really invest in an R9 370X and a 1440p 144hz Freesync monitor? It would be like investing in a GTX 750ti (Or similar) and then buying a Rog Swift (1440p 144hz G-Sync monitor), kind of pointless as unless you drop the setting super low your not going to be able to drive anywhere near that.

I would like to see if the card has evolved at all, would be nice to see theses things in OC models being around 1300mhz.
Forgive my ignorance. 1440p 144hz monitor for adaptive sync? Not many cards will give you anywhere near that fps (even dual set ups).
Isn't the entire point of adaptive sync to eliminate tear at less than optimal frame rates (40-60 on a 60hz?). On a 144hz monitor, what would your cards need to run at? Or is it simply that above 60fps will cause tearing regardless, so 144hz is the next step? So if your cards can output 90fps, you'd need the 144 refresh to drop down to match the 90fps, something a 60hz monitor can't attain?
Posted on Reply
#10
GhostRyder
the54thvoid said:
Forgive my ignorance. 1440p 144hz monitor for adaptive sync? Not many cards will give you anywhere near that fps (even dual set ups).
Isn't the entire point of adaptive sync to eliminate tear at less than optimal frame rates (40-60 on a 60hz?). On a 144hz monitor, what would your cards need to run at? Or is it simply that above 60fps will cause tearing regardless, so 144hz is the next step? So if your cards can output 90fps, you'd need the 144 refresh to drop down to match the 90fps, something a 60hz monitor can't attain?
What I am referring to is that most monitors that have G-Sync or Freesync are 144hz monitors (most being 1440p). When it comes to games and tearing it happens more when your FPS is constantly changing which in more cases than naught will occur at anything beyond 60hz since most game companies focus on 60hz as the goal of optimization. Even running 1080p 120/144hz is extremely difficult even though it should be similar in a way to running around the 1440p 60hz mark because games are not designed for that high speed of refresh. Therefor your going to experience a lot of FPS changes which results in things like tearing, stutter, etc. Hence why almost all the monitors available have the 144hz refresh (minus the 4k and a select few exceptions) because your not going to benefit enough from the techs enough at the low range.

If your constantly changing FPS dropping by significant amounts while shooting for 60hz, it would be helpful to a point as well but below a certain point the tech starts to lose its powers as well which then would show the need for more GPU power (besides poor game optimization which could also be a flaw). That is the way I view these techs at least.
Posted on Reply
#11
SonicZap
I don't think that the lack of FreeSync is a big problem here. Not because nearly all FreeSync monitors are 1440p, but because all FreeSync monitors are too expensive to be paired with mid-range graphics cards. Not many spend 150€ on a GPU and 400+€ on a monitor.
Posted on Reply
#12
ProTofik
If it won't have HEVC Hi10p decoder nor HDMI 2.0 (something that currently only GTX960, and soon GTX950 will have), then it's pretty much useless.
Posted on Reply
#13
Analog_Interface
I bet Tahiti was a "miraculous" great innovation by AMD. I think they're still surprised how they made such a revolutionary chip (and architecture) that time and perhaps, this is a very precious bit of silicon for AMD. That's maybe why AMD's not letting Tahiti go.

But as long as the end user gets to see performance, all these doesn't matter really. We just have to see how well 370X can do with 950 (and 950 Ti too perhaps). But do expect to see some raging fury (I don' mean perfo'mance) from these GPU - provided coming with such high clocks.
Posted on Reply
#14
Casecutter
Face it now suck at 28nm it’s a race to the bottom. Neither are looking to invest at this point, when next time around (generation) iGPU’s and APU’s will perchance off acceptable entry 1080p. That said, Nvidia would be odd-man-out, they’d play it smart to keep a cost effective entry/middle ground gaming card in the mix to cultivate brand loyalty at those early stages.

If you are looking at $150 middle/high 1080p gaming card, most aren't gravitating toward every wiz-bang feature, or obscure codedecs, all you want is BfB and for that… we'll wait and see.

A boost clock of 1080Mhz stock would be nice, still not the 1100Mhz of the likes of PowerColor 270X PCS+; that PCS+ had 1425MHz (5.7Gbps) memory. I sure hope AMD bumps the memory up to something like 1500Mhz (6.0Gbps), I'd think sourcing better memory might be cost effective at this point.
Posted on Reply
#15
GorbazTheDragon
SonicZap said:
I don't think that the lack of FreeSync is a big problem here. Not because nearly all FreeSync monitors are 1440p, but because all FreeSync monitors are too expensive to be paired with mid-range graphics cards. Not many spend 150€ on a GPU and 400+€ on a monitor.
That's not the point...

The problem that AMD is showing itself to be more and more incapable of releasing good and profitable products. The only reason they are re-releasing this chip is because they don't have anything better, all their effort was put in Fiji, which still is a half-assed product in many ways... It uses exactly the same base architecture as the HD 7000 series. You can say exactly the same for their CPU archs... Nothing really new since FX release, which was so bad that they had to FIX the architecture for the 8350, etc. And don't come telling me that anything since the FX-8350 is a new architecture because it might as well be saying Broadwell is a new architecture, it's not.

Sure, I might buy an AMD card in the near future because of the low cost, but I would not be at all willing to invest a single penny in them. Not for the forseeable future, and definitely not before Zen is released, and only then if Zen really shows progress beyond just better price/performance.
Posted on Reply
#16
SonicZap
My point was that the lack of FreeSync support isn't an issue here, but whatever. If you want to argue..

GorbazTheDragon said:
The problem that AMD is showing itself to be more and more incapable of releasing good and profitable products. The only reason they are re-releasing this chip is because they don't have anything better...
Or maybe they don't have a reason to invest more in 28nm GPUs at this time. Pitcairn is still performance-competitive with Maxwell (HD 7850 > 750 Ti, highly clocked HD 7870 can be fairly close to 960), while being reasonably efficient (a bit more so than Kepler GPUs). I'm also fairly certain that Pitcairn has been a profitable product and is still being one, they designed the chip in 2012 and it still does its job. Btw, not even Nvidia seems to be doing more on 28nm anymore, the 950 will be a cut-down 960.

GorbazTheDragon said:
...all their effort was put in Fiji, which still is a half-assed product in many ways... It uses exactly the same base architecture as the HD 7000 series. You can say exactly the same for their CPU archs... Nothing really new since FX release, which was so bad that they had to FIX the architecture for the 8350, etc. And don't come telling me that anything since the FX-8350 is a new architecture because it might as well be saying Broadwell is a new architecture, it's not.
I pretty much agree, but how exactly is this relevant to the competition between the R7 370X (aka HD 7870) and GTX 950? I'm afraid I can't see the connection. The R7 370X is going to be decent, just like the R9 270X and HD 7870 were. It's nothing new or exciting, but it's reliable old technology that still gives decent performance.
Posted on Reply
#17
arbiter
If info in the gpu-z is correct then a 370 is a straight rebranded Curacao which is a r7 270(x). Not really that shocking at this point.
Posted on Reply
#18
profoundWHALE
SonicZap said:
The R7 370X is going to be decent, just like the R9 270X and HD 7870 were. It's nothing new or exciting, but it's reliable old technology that still gives decent performance.
So the drivers will be solid?
Posted on Reply
#19
Dieinafire
Old tech passed off as new tech with a higher price tag. The AMD way!
Posted on Reply
#20
HumanSmoke
GorbazTheDragon said:
That's not the point...
The problem that AMD is showing itself to be more and more incapable of releasing good and profitable products.
Exactly. As I pointed out earlier in the thread...
HumanSmoke said:
In the end, it makes little difference I suppose, it's still a recycled three-and-a-half-year-old card with little marketing appeal - no FreeSync support, nor TrueAudio ( if TrueAudio is still a thing that is).
No marketing appeal over the exact same cards being sold for over three years - which should cannibalize new sales as they are discounted down to clear AIB inventory ( some SKUs are already below the 370X's expected price range)...so, regardless of the occasional person cheerleading the incoming cards viability, I doubt that even they would pay more money for the exact same card with a new sticker just to validate a point that won't be reflected in a wider market.
SonicZap said:
It's nothing new or exciting, but it's reliable old technology that still gives decent performance.
The AMD mantra. With a marketing perception such as that, is it any wonder the companies market share is nosediving.
Posted on Reply
#21
arbiter
HumanSmoke said:
The AMD mantra. With a marketing perception such as that, is it any wonder the companies market share is nosediving.
With them claiming its new gpu calling it "Trinidad" which was "Curacao" when it was a 270(x) and called "Pitcairn" when it was 7870(7850). When you look at how they reused a gpu from early 2012 as main end product. it should be more like r5 340 at this point for it to be acceptable IMO. People are mad when review sites slamming AMD last few years over things, this is reason why. It still wouldn't help with complaints but would least give them something, should used tahiti for 370 series so least they got a performance bump though that means 380 series would had to use hawaii, and fury could took over 390 series name. Still rebrand hell but least series got valid performance boost's instead of boosts from overclocks.
Posted on Reply
#22
damric
At 1180MHz this will actually be faster than the GTX 960 in many games. There isn't a whole lot more headroom left for any more overclocking though.
Posted on Reply
#24
GorbazTheDragon
damric said:
At 1180MHz this will actually be faster than the GTX 960 in many games. There isn't a whole lot more headroom left for any more overclocking though.
If you look at all of AMDs cards they are rubbish overclockers when compared to maxwell... NV's cards are all pushing steadily towards the 1.5GHz mark out of the box...

I'm quite curious as to what exactly is allowing those chips to be much faster
Posted on Reply
#25
HumanSmoke
arbiter said:
, should used tahiti for 370 series so least they got a performance bump though that means 380 series would had to use hawaii, and fury could took over 390 series name. Still rebrand hell but least series got valid performance boost's instead of boosts from overclocks.
Wouldn't make too much difference from a marketing point of view. Tahiti - like Pitcairn/Curacao/Trinidad, isn't overly DX12 compliant....and according to many AMD devotee's, DX12 is going to be the companies graphics performance saviour. Pretty difficult to make a case for DX12 being a prime reason to choose the Fury/Fury X, but dismiss DX12 for a card where the features would provide a solid performance leap.
GorbazTheDragon said:
I'm quite curious as to what exactly is allowing those chips to be much faster
28nm is a pretty mature process - much more so in the 3.5 years since its debut. Nvidia's own silicon has shown similar boosts ( A1 vs B1 silicon GTX 780's for example)
Posted on Reply
Add your own comment