• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

New cards from AMD in two weeks ?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
There have been some people flashing their 290x with a 390x for some ungodly reason (same EXACT card sans 8GB vRAM). But outside of that, I haven't run across any.
 
Tyerker said:
I wonder if the 390X BIOS would offer any benefit for the 8GB 290X...
Click to expand...
...

EarthDog said:
There have been some people flashing their 290x with a 390x for some ungodly reason (same EXACT card sans 8GB vRAM). But outside of that, I haven't run across any.
Click to expand...
The 'ungodly reason' was my way of saying, there isn't a difference between the cards, period. No efficiency increases. It is what it is last I understood.
 
I thought some reviews came out showing a performance increase clock-for-clock over the 290x, but it wasn't huge. Granted, I would have to double check they weren't comparing the 390x with current gen drivers versus 290x at launch (which has significantly improved since then)
 
That has been the problem from what I have seen, is comparing apples to apples. Something was always different, be it drivers/clocks, something. But nothing I saw lead me to believe there are any performances differences in the bios.
 
Every review I just looked through showed about a 10 or so % improvement over the 290x. I couldn't find one that had both of them at the same clock speed though, as the 290x shipped at around 1000mhz and the 390x at 1100mhz (assuming no factory OCs) and a ram overclock too as well I do believe. That right there is a 10% clock speed increase.
 
Yep. While scaling isn't exactly linear, that lends a bit of credence to what I was saying. This TPU review is the best thing I have seen for apples to apples. They use the same driver, but the clockspeeds were different. The reference 390x runs at 1050/1500 while the reference 290x is 1000/1250. So there is a 50/250Mhz difference there.

.. it shows the average ~10% gain.
 
The 390x has higher core and memory clocks at a lower TDP than the 290x. While it may technically be the same physical GPU core, they are fairly different cards. They didn't just slap a 3 on it.
 
They didn't just slap a 3 on it.
Click to expand...
Yes, they did. Outside of some microcode and bios changes (which makes GPUz tell you what kind of card it is), its the same thing as the 290x 8GB with faster clocks.

Its also the same TDP... 250W. See my TPU link above for details. ;)

Since all R9 300 Series cards are rebrands of existing R9 200 Series products, we don't expect any significant power consumption improvements with the R9 390X; these will hopefully come with the Fiji GPU on the Radeon Fury.
Click to expand...

The Radeon R9 390X is based on the "Hawaii" silicon (now referred to as "Grenada" without any silicon changes)
Click to expand...

The MSI R9 390X Gaming we are reviewing today is based on a re-brand of the R9 290X. The clock speeds have been increased a little bit, but the GPU itself is the same, offering the same features, such as shader count, ROPs, and texture units. Only memory size has been doubled to 8 GB, a capacity that has been available on some R9 290X cards before.
Click to expand...

My recommendation, especially when on a budget, is to look for a R9 290X with 4 GB because it can be had for around $300. Overclock it some and boom, you have R9 390X 8 GB performance levels at a fraction of the price.
Click to expand...
 
Last edited:
EarthDog, I love you... But why must you always shatter my dreams?

Sites like GPUBoss and some reviews I was looking at had the 290x @ 275W... Freaking internet.
 
Well, in fairness, it is REPORTED at 250W, however a lot of testing has pushed that closer to 300W. But since the card is the same, the power use is also the same (remarkably similar).
 
290X 8GB = 390X 8GB but I don't think that BIOS from 390X will work on 290X 4GB. I already saw comments it's failing in 3D and it's probably true as 8GB will probably have different timing set. Performance clock to clock should be exactly the same but in most reviews around the web you see comparison of 290X on old drivers ( from older reviews on the same website ) and 390X on new drivers. As we know 290X on new drivers is maybe 15-20% faster than it was ~2 years ago. If you see any significant differences in performance between 290X and 390X then reviewer tested only 390X and other results took from old database.
 
Is it me or did AMD forget to release a R9 380X?????? If they chop the memory interface from 384 to 256, I will stay with the old stuff :popcorn:

*AMD HD-7950 = 1792 Stream Processors & 3GB 384-Bit DDR5 (Tahiti Pro)
*AMD HD-7970 = 2048 Stream Processors & 3GB 384-Bit DDR5 (Tahiti XT)
*R9 280 = 1792 Stream Processors & 3GB 384-Bit DDR5 (Tahiti Pro2)
*R9 280x = 2048 Stream Processors & 3GB 384-Bit DDR5 (Tahiti XTL)

Them Bast** Castrating ^^Good Cards ^^

-)R9 285 = 1792 Stream Processors & 2GB 256-Bit DDR5 (Tonga)
-)R9 380 = 1792 Stream Processors & 2/4GB 256-Bit DDR5 (Antigua Pro)
?)R9 380x = 2048 Stream Processors & ???????

And then they JUMP straight to the R9 390 & R9 390x
 
They didn't forget. They said that won't be another refresh of 7970/280X. R9 380 = R9 285 so it's a chip with 256 bit bus. 380X is not in plans for now but maybe they will make something like 380 but with more SP. I doubt they will make anything with wider bus in this price point. On the other hand 380X would be bad idea as price would be between 380 and 390 so a spot which has kinda bad performance/price ratio. Gamers will stick to cheaper series like 370 or go for 390+. 380 is also bad idea but AMD really likes to release products which shouldn't even hit the stores.
Right now best option is R9 290X as price is not so much higher than new R9 380 while performance is much better.

From new series I see it like R7 370 -> Fury , all between is waste of money and better is to look for older series which are collecting dust in large stores. Even R7 370 is not so good but price/performance is not that bad.
 
390(x) are not rebrands, though the changes aren't that significant. There are some slight performance upgrades( a few percent), but decent power efficiency improvements. Most sites showed the 390/390x at the same or slightly higher power usage than their 200 counterparts, despite having 2x the memory, much faster memory busses, and overclocked cores. Most sites used MSI/Sapphire cards which have an overclocking bios and will draw more power than a stock bios. If you look at another model (HiS iceQ) which doesn't do this, the cards actually draw less than the 200 models despite the higher clocks and memory.

http://www.techspot.com/review/1019-radeon-r9-390x-390-380/page7.html
Power_03.png


Again, not a huge difference, but not the same exact die either. Even @ techpowerup, the average difference is 12% and in some tests gets above 15%

crysis3_1920_1080.gif

In the end, a 390x pulls to about equality to a 980 at stock, but loses once overclocked and uses more power.
 
The power consumption difference of 3w isn't remotely compelling. I can get 10 reference cards and show 10w difference between them or one card a few watts just on different fan speeds.. I would expect it to be about the same or more considering the ram. It could just be binned slightly better as ram doesn't use much power anyway (a few watts).

To confirm, as i stated earlier here, we would need to match clocks and drivers. The summary shows a 10% difference over all their games tested...which can easily be clockspeed and driver differences...or just clocks.

http://www.techpowerup.com/mobile/reviews/MSI/R9_390X_Gaming/30.html

, while the rest of the lineup saw a cascading re-badging from the previous generation. AMD's previous generation flagship, the HD 7900 series, went on to become the performance-segment R9 280 series, and so on, and the performance-segment "Tonga" silicon was added afterward.The story is predictable even today.

With this generation, there is essentially one new silicon, the HBM-equipped "Fiji," which will be launched later this month and will eventually drive up to five products from AMD. The previous-generation flagship silicon "Hawaii" now drives AMD's performance-segment products, the Radeon R9 390 and R9 390X we're reviewing today.

The Radeon R9 390X is based on the "Hawaii" silicon (now referred to as "Grenada" without any silicon changes) and features the same core-configuration as the R9 290X.
Click to expand...

From tech spot (you should read the intro to that article too):
...The Radeon R9 390X delivered exactly what you'd expect for a slightly beefed up R9 290X with 20% faster GDDR5 memory.
Power consumption figures were comparable to previous generation cards, with the R9 390X being in line with the R9 290X for example.
Click to expand...

So again, it is tbe same exact die. Same shader count, rops, tmus. If there are any differences it came through the microcode and the bios... but I haven't seen anything compelling otherwise. Outside of the fury, fury x, and nano, these are rebrands with more memory and different clocks. Outside of that, they are the same thing.
 
Last edited:
EarthDog said:
The power consumption difference of 3w isn't remotely compelling. I can get 10 reference cards and show 10w difference between them or one card a few watts just on different fan speeds.. I would expect it to be about the same or more considering the ram. It could just be binned slightly better as ram doesn't use much power anyway (a few watts).

To confirm, as i stated earlier here, we would need to match clocks and drivers. The summary shows a 10% difference over all their games tested...which can easily be clockspeed and driver differences...or just clocks.

http://www.techpowerup.com/mobile/reviews/MSI/R9_390X_Gaming/30.html



From tech spot (you should read the intro to that article too):

So again, it is tbe same exact die. Same shader count, rops, tmus. If there are any differences it came through the microcode and the bios... but I haven't seen anything compelling otherwise. Outside of the fury, fury x, and nano, these are rebrands with more memory and different clocks. Outside of that, they are the same thing.
Click to expand...

tpu is mistaken and the evidence bears this out. Here is an interview with Richard Huddy of AMD:


Richard Huddy said:
There are changes to the silicon
Click to expand...

Pcper has a good paragraph on it from their review:

http://www.pcper.com/reviews/Graphics-Cards/Sapphire-Nitro-Radeon-R9-390-8GB-Review

Pcper said:
So what changes were made in these new spins of GPUs? AMD was quick to comment on the term "rebrand" that will no doubt be associated by many with the Radeon R9 300-series. They insist that engineers have been working on these GPU re-spins for over year and simply calling them "rebrands" takes away from the work the teams did. These GPUs (the 390 and 390X at least) have a "ground up" redesign of the software microcontroller that handles the clocks and gating to improve GPU power efficiency. As you would expect for a GPU built on the same 28nm process technology that has been around for many years, AMD has tweaked the design somewhat to better take advantage of evolutions in TSMC's 28nm process. And, thanks to higher clocks on both the GPU and the memory, performance increases will be seen over the existing R9 200-series as well. Being able to run around 50 MHz higher on the GPU and 250 MHz (1.0 GHz effective) on the memory inside the same power envelope shows that AMD has done SOMETHING, though how much that means for consumers is up in the air.
Click to expand...

Acknowledging that the new series doesn't use more power is acknowledging power efficiency tweaks. All else being equal, power increases linearly with frequency. So, you take the HiS card running at 7% higher core clocks, with 20% higher memory clocks and double the amount of memory, the card without tweaks should consume approximately:

200 W (shaders) * 1.07 = 14 W over baseline
30-40 W (memory controller) * 1.2 = 6-8 W over baseline
5 W (memory) over baseline

Total = 25-27 W over baseline or about 10% increase. Instead, you get the same power for more performance. In Crysis 3 the HiS card was 14.7% faster than the 290x despite only a 7% core frequency increase. I like how pcper called it a re-spin as that is pretty much what it is. It is still Hawaii architecture with the same specs, possibly even the same RTL, but work has been done at the transistor/microcontroller level to get better performance per watt from the same architecture. It's not a large change, but the difference is there and it's up to each consumer to decide how compelling the changes are. To me, not enough to get me to buy a 3xx card over a cheaper 2xx card but as that inventory runs dry, that won't be an option any more.

P.s. Also from pcper review:

pcper said:
UPDATE: I used a modified version of the Catalyst 15.15 driver to re-test the Radeon R9 290 and saw nothing that points to AMD purposefully altering performance with the driver swap.
Click to expand...
 
You must log in or register to reply here.
Back