That may sound unimpressive but don't forget the maximum power consumption of <150W for FirePro W7100. I think AMD just didn't bother doing any power optimization for R9 cards.
-
And nvidia may not even give us the full chip and only release a heavily underclocked 980M.
Just like the 680M. And then 680MX, 780M & 880M. -
So disappointed with that, i was expecting something like 150WCloudfire likes this.
-
What the hell is the point on releasing a product that match an existing crappy 880M half year after?
I`m actually thinking that this card will match or barely beat GTX 880M because surely it must be a deeply downclocked R9 285 to get TDP down from 190W to 100W. Yeah I think they will put it at 125W just like Nvidia did with GTX 880M. Hot!
I had great hope for this card, because I was planning to go AMD for the first time in many years. But I wont touch GTX 880M or R9 M295X. They are not even trying to put out something to counter Maxwell and GTX 860M/850M.
GTX 980M is my only hope now, and why should Nvidia even bother to release it when they got no competition?
This is not good
Cammac66 likes this. -
By the looks of it
R9 285: 1792 cores @ 918MHz @ 190'W
W7100: 1792 cores @ Way lower than 918MHz @ 150W
I hope you are right that W7100 is more efficient but tons of rumors said R9 285 and W7100 and M295X is using the same Tonga chip
Horrible live show, bad news and no mention of R9 M295X.
Ugh -
The 7970M will stick around with me for another year it seems. No reason to "upgrade"
-
"Good news" as it seems. One very last push of my hope reserve till 2nd September after which I guess I would just give-up and become a princess or something.
Cloudfire and Marecki_clf like this. -
King of Interns Simply a laptop enthusiast
Or just keep the 7970M to yourself instead of giving it to your brother
triturbo likes this. -
Now it makes sense why they're keeping it in the R9 200 series as opposed to gracing them with the R9 300 nomenclature.
-
What is the point of this?!? Still slower than my now 3-year-old 7950. How pathetic.
-
Totally knew it was going to be a fail.
And now that AMD once again couldn't bring anything to the table, nVidia has no reason to do so either.
Glad i chose to sleep today.
Sent from my HTC One_M8 using Tapatalkocticeps likes this. -
Same. Why the hell would I get up at 7 AM on a beautiful Saturday morning?
-
Robbo99999 Notebook Prophet
I've not been following much of the rumours on the AMD side of things for the next generation GPU tech. Does this mean that AMD's next mobile flagship GPU is going to be just a little faster than the 880M then as a guesstimate, and any ideas when the successor to this likely disappointing GPU will appear - ie, how long would we have to live with this disappointing GPU?
-
Or build a desktop
Now that I realized I don't need the mobility, I'm seriously entertaining the thought of building an X79 setup. We shall see... -
It isn't, judging by Firestrike scores it's as fast as XFX R9 280X (1000MHz core clock) with older drivers. Not really sure how much newer drivers improve synthetic scores though.
-
Its a chip thst competes against GTX 880M. Thats all there is to it.
Nothing is new about it. Just as power efficient as their 3 year old architecture.Robbo99999 likes this. -
My 7950 has same 1792:112:32 core config @ 1150 MHz and faster 6.4 GHz GDDR5 on a wider 384-bit memory bus. Tonga is still GCN 1.1 and the R9 285 is no more efficient (perf/watt) than the first GCN 1.0 cards that came out 3 years ago. A stock 7950 is 200W TDP, this is 190W with a cut down memory interface. I got my 7950 for $180 last year, this is $250.
This is just epic fail in every possible way. -
R9 285 1792 cores 190W: Firestrike Extreme: X3513
R9 280 1792 cores 220W (ish) Firestrike Extreme: X3400
R9 280X 2048 cores 250W: Firestrike Extreme: X3600
R9 285 Firestrike Performance: P7066
R9 280 Firestrike Performance: P7000
R9 280X Firestrike Performance: P7200
There are some improvements but hardly any amazing gains imo. Could be worse than this too because AMD used a pretty beefy 4960X with the R9 286 and the above is total score benefiting from a faster CPU -
This would've been a nice publicity stunt if AMD had released this on April 1st, but godammit AMD, it's the end of August ok? Get your feces together man.
-
That is fine for me, I still expect M295X to be 20-30% faster than 780M at 100W TDP, which is honestly more than enough for anything at 1080P. If people really want more than 1080P resolution they are better off building a desktop. It should even be fine with higher resolutions with no AA because who needs that at those crazy pixel densities on small high res displays.
-
We shall see.
My guess:
GTX 880M performance +\-10%
TDP around 125W like 880M like rumored couple of weeks ago Click here)
Remember the 723MHz core clock leak? Kinda makes sense if they want to remove 70-90W from R9 285 doesnt it?
GPU, OpenGL and OpenCL database -
What a disappointing turn of events. Since its the same architecture just cut down and rearranged, the new mobility chip will be comparable to 880m just as you say Cloud.
While at least it means another alternative to current high end GPUs, this also means no pressure to nvidia and we will most likely get a bleh 980m. -
Same... I'm looking at a cramming everything into a raven rvz01 which I'll sit in my entertainment unit...
Ruining a 5m display port cable cable to the monitor on my kitchen table (it's where my laptop is now and only gets moved when guest come over) and using a wireless keyboard and mouse. -
Meaker@Sager Company Representative
Lol you guys are WAY too fast to jump to the conclusions you have. Insufficient data so far.
from anandtech:
The R9 285 will have a rated typical board power (AMDs analogue for TDP) of 190W. Notably this is only 10W higher than the Pitcairn based R9 270X despite the 40% larger SP count, or alternatively is 60W lower than the Tahiti based R9 280. While we dont have a ton of details on the GPU at this time, taking into consideration the R9 270X comparison in particular, its clear that AMD has done some work on efficiency to squeeze out more compared to the GCN 1.0 based Pitcairn and Tahiti parts that R9 285 is going to be placed between.
Also we don't know if this is a gcn 1.2 type revision or not since even 1.1 is an unofficial name.GuniGuGu likes this. -
TPU says R9 280 is 200W?
AMD Radeon R9 280 | techPowerUp GPU Database
Also how did same GPU go from 190W (7950) to 225W (7950 Boost) to 250W (R9 280)?!? 7970, 7970 GHz, and R9 280X are all 250W yet have higher clocks and 4 more CU's.
-
Meaker@Sager Company Representative
Boost clocks and turbo level adjustments. Each version was clocked higher so power went up.
-
So 250W for 7970 GHz and R9 280X is a lowball number? Cause they also have the PowerTune Boost stuff.
And +8 MHz on core took Tahiti Pro from 225W (7950B) to 250W (R9 280)? ROFL
I'm sorry, I just fail to see where AMD made any significant efficiency gains with Tonga Pro (R9 285). -
I'm probably going full tower if I build this desktop. Actually still need to settle on X79, X99, or Z97 lol but I have a sketch of each build, I mean really the only variable is the mobo+CPU (and DDR4 for X99) so it's not too hard. Doesn't hurt to wait a bit and see how X99 does, plus it might drive X79 prices down which is always a plus.
Of course all this is hinged upon nVidia releasing GTX 880 on schedule, it performing as rumored and priced <$500. Otherwise I'm not building jack.GuniGuGu likes this. -
The real question is: Cosmos II or 900D?
-
Meaker@Sager Company Representative
And i quote from anandtech again:
7950B required a very big step up in voltage to hit its boost state, causing it to consistently fall back to its base state and putting the card’s average clockspeed well below its top clockspeed.
Setting a tdp does not magically mean it will always boost at full at that power level. Ask 880m users
-
Well that's dumb. Glad I got a vanilla 7950 and not the Boost or 280. PowerTune Boost and GPU Boost should all be called Cripple Boost anyway since all they do is throttle performance and get in the way of overclocking.
-
Meaker@Sager Company Representative
They prevent things like furmark requiring the overall clockspeed to be lowered. Sure tdp can be annoying but for us we just raise up the tdp and carry on.
-
Also it's not helping things that 280/280X are faster than 285/285X. AMD already threw a curveball with the stupid R5/R7/R9 naming scheme this generation, but this is on a whole 'nother level.
-
Heard many good things about the 900D, but that bland look... Although it'll fit well with my Clevo LOL
Honestly right now I'm just playing the waiting game -- hope to be decently surprised by X99, but expecting an overpriced turd that won't perform at its full potential until 2016 when DDR4 becomes more mature and affordable. I pity the fool who buys that 8-core 5960X, I'm 99% certain it's a converted Xeon, hence the absurd 3.3GHz turbo boost. -
Meaker@Sager Company Representative
Where are the benchmarks to show that?
-
If FurMark causes a card to hit temp limits or trigger OCP, it's gonna throttle clock speed regardless of Boost or not. Boost just adds more complexity to gauging a card's performance and more variables when overclocking. It's stupid. That's why all the modded overclocking vBIOSes have it disabled.
Comparing specs on paper, no benchmarks. Unless the Tonga core is a significant revision of GCN and it can overcome the cut down memory bus, it's gonna perform slower. -
In case Meaker's question was directed towards me: this is what I went by when I said 1st gen DDR4 will likely be slower than DDR3 at the same price point.
-
All the top-end i7's on the server platforms come from Xeons of the same uarch, hence why they always lag one generation behind the mainstream Core platform.
Hard to say when he replied same minute as you. I think he was asking me. -
Yeah you're right, had a brain fart moment there.
-
Meaker@Sager Company Representative
It was the gpu issue, memory is memory and is one of the most boring components out there unless you run an igp.
-
King of Interns Simply a laptop enthusiast
September 2nd then we all know what this 285 can do.
Meaker is right the proof is in the pudding! I for one will say nothing more until I see concrete results in a week or so.
Still yesterday's show was terrible waste of time. -
R9 M290X:
1280 shaders * 850MHz =1088000
R9 M295X:
1792 shaders x 720MHz = 1290240
1290240/1088000 = 1,185 = 18.5% = 20%
I think the R9 M295X will be on par with the GTX 880M. Sometimes the 880M will win, sometimes the M295X will win, but across a wide range of games, I expect them to be equal.
The M295X would be far better than GTX 880M if it was a new architecture like Maxwell (+35% per core over Kepler) but since its the same architecture as current R9 200/7000 series, I cant see anything other than the rough estimation above. -
King of Interns Simply a laptop enthusiast
Unless we see it shipping with higher stock clocks!
-
True.
Between 720MHz and 850MHz the performance will be 18-40% above R9 M290X depending on what clock it ships with.
But I think its 720MHz.
7870 170W: 1280 cores @ 1000MHz
7970M: 1280 cores @ 850MHz
R9 285 190W: 1792 cores @ 920MHz
R9 M295X: 1792 cores @ ?
So there are two things going against the M295X:
1. The R9 285 which it is based on runs 80MHz lower than the 7870 which the 7970M was based on
2. R9 285 which the M295X is based on, runs 20W higher than the 7870 which the 7970M was based on
This is why I was so dissapointed last night. Although we didnt see the M295X, the specs of the R9 285 speaks volumes -
King of Interns Simply a laptop enthusiast
I think there is a chance it could ship at somewhere between 800-850mhz. They could very well bin the m295x chips far better. M290x is 900mhz vs 1000mhz and runs at lower voltage vs the 7970m. I know this because even the 8970m can run at 975mhz at just 1.025v. The 7970m needed 1.05v to achieve the same in most cases. The architecture and 28mm fab process are even more mature.
Lets hope we get 850mhz at least. Fingers crossed. I would laugh if it ships at 875-900mhz!! -
King of Interns Simply a laptop enthusiast
Do your sources say this ships on 2nd September too?
-
We shall see what happens.
He only told me it will be available for notebooks in September. No specific date I`m afraid.
So who knows when information about M295X arrives -
where did u get this info? my 7970m does 1040Mhz on stock volts...
-
TDP is not consumption
-
Hi guys maybe it will be not that bad i found nice review of old 7950 wchich on stock core clock had power draw 145W(HIS) so maybe ~800 mhz on mobile Tonga could be achieved...
page 9 peak power draw Power Consumption - Radeon HD 7950 3 GB: Six Cards, Benchmarked And Reviewed In this case even 10 tdp Tonga reduction can be helpful.
Radeon R9-M295X
Discussion in 'Gaming (Software and Graphics Cards)' started by Tsubasa, Mar 15, 2014.