Whats the point

[Eric1987]

Whats the point of people using 390x bios on their 290x? Just for slightly higher clocks?

 

[IEC]

Trying to get the portion of performance boost that is NOT obtainable via higher clocks.

Of course, I'd be inclined to believe a large part of it is memory timings but since the 390X cards tend to use different GDDR5 chips I don't think they'll get the improvement they are looking for.

 

[Shehriazad]

I feel like people are just trying to buy Snake Oil...or chasing ghosts...whatever you prefer.

I really doubt that it will do what they expect it to do.

 

[vissarix]

Whats the point of people using 390x bios on their 290x? Just for slightly higher clocks?

flash a 390x bios on a 290x, and then put a 390x sticker in the 290x and bam...your sub $300 gpu transformed into an ''enhanced'' refreshed or whatever you wanna call it hawaii worth $429...

http://www.hardocp.com/article/2015/06/18/msi_r9_390x_gaming_8g_video_card_review/9#.VYc0JlWqqko

 

[Shehriazad]

flash a 390x bios on a 290x, and then put a 390x sticker in the 290x and bam...your sub $300 gpu transformed into an ''enhanced'' refreshed or whatever you wanna call it hawaii worth $429...

http://www.hardocp.com/article/2015/06/18/msi_r9_390x_gaming_8g_video_card_review/9#.VYc0JlWqqko

Except that the rebrands seem to be able to do more... at lower power consumption and heat.

(Taken from Guru3D)

I feel like a lot of idiots that blindly want the "ermagerd driver enhancements" are going to fry their cards because they are receiving a rather improper OC on possibly inferior cooled cards (depending on brand). This jump in efficiency is actually quite significant...not only is the GDDR5 clock 1.1 Ghz higher and the core clock 100 mhz higher in boost compared to AMD stock R9 290X (Their test uses an already slightly OC'd 390X)... YET it seems to consume 30w less even though it has twice the ram compared to the tested R9 290 that is on the list. If it was 4G...the difference would be even bigger (possibly 50W?).

I can already see a lot of angry people wondering why their magic 390X bios flash suddenly throttles or misbehaves.

More happened there than "just" driver and clocks...it's obviously not a major hardware/arch upgrade they did, but SOMETHING was quite obviously improved OTHER than drivers and clocks.

 

[vissarix]

Except that the rebrands seem to be able to do more... at lower power consumption and heat.

(Taken from Guru3D)

I feel like a lot of idiots that blindly want the "ermagerd driver enhancements" are going to fry their cards because they are receiving a rather improper OC on possibly inferior cooled cards (depending on brand). This jump in efficiency is actually quite significant...not only is the GDDR5 clock 1.1 Ghz higher and the core clock 100 mhz higher in boost compared to AMD stock R9 290X (Their test uses an already slightly OC'd 390X)... YET it seems to consume 30w less even though it has twice the ram compared to the tested R9 290 that is on the list. If it was 4G...the difference would be even bigger (possibly 50W?).

I can already see a lot of angry people wondering why their magic 390X bios flash suddenly throttles or misbehaves.

More happened there than "just" driver and clocks...it's obviously not a major hardware/arch upgrade they did, but SOMETHING was quite obviously improved OTHER than drivers and clocks.

Except that the rebrands seem to be able to do more...like 54 watts more vs a 290x reference :|

 

[.vodka]

MSI gaming series cards are like that. For example, their 970/980s consume absurd amounts of power compared to other brands. MSI cards are representative of MSI cards in the power consumption department... This has been shown on one of the other many threads about the subject.

 

[PG]

Whats the point of people using 390x bios on their 290x? Just for slightly higher clocks?

Why didn't you just read the other thread about this same topic?

http://forums.anandtech.com/showthread.php?t=2436535

 

[Eric1987]

Because there is like 15 different responses why thats the case. Not the same topic. Sorry.

 

[Shmee]

Will the re badges support crossfire with the 290(x) series cards? Is there an updated crossfire support chart?

 

[Shehriazad]

I wonder why those values are SO much in conflict on the reviews... They both review the same exact card.

 

[MagickMan]

Whats the point of people using 390x bios on their 290x? Just for slightly higher clocks?

Better/lower latency timings, if your GPU memory can handle it.

Edit: Oh yeah, and the TDP thing, which... yeah, whatever. I couldn't care less about that.

 

[sandorski]

I feel like people are just trying to buy Snake Oil...or chasing ghosts...whatever you prefer.

I really doubt that it will do what they expect it to do.

Pretty much. It's based on the presupposition that the 390 is simply a rebadge of the 290. That said, if it actually results in the same Performance and other measurable aspects, then it would seem to indicate a Rebadge.

 

[Leyawiin]

Except that the rebrands seem to be able to do more... at lower power consumption and heat.

(Taken from Guru3D)

I feel like a lot of idiots that blindly want the "ermagerd driver enhancements" are going to fry their cards because they are receiving a rather improper OC on possibly inferior cooled cards (depending on brand). This jump in efficiency is actually quite significant...not only is the GDDR5 clock 1.1 Ghz higher and the core clock 100 mhz higher in boost compared to AMD stock R9 290X (Their test uses an already slightly OC'd 390X)... YET it seems to consume 30w less even though it has twice the ram compared to the tested R9 290 that is on the list. If it was 4G...the difference would be even bigger (possibly 50W?).

I can already see a lot of angry people wondering why their magic 390X bios flash suddenly throttles or misbehaves.

More happened there than "just" driver and clocks...it's obviously not a major hardware/arch upgrade they did, but SOMETHING was quite obviously improved OTHER than drivers and clocks.

Power consumption is off the scale (and this review is more along the lines of every other review I've read). Yes, yes, there's always the come back "Enthusiasts don't care about power consumption" but this is a little too much to ignore.

http://www.techpowerup.com/reviews/MSI/R9_390X_Gaming/28.html

 

[xthetenth]

For crying out loud how many times do we have to cover the MSI card being an outlier and not representative of the preponderance of 390(X)s?

 

[Stuka87]

Why are people/sites comparing the MSI 390X to non MSI 290X's? This has been covered to death, the MSI cards use more, regardless of it being AMD or nVidia.

As for people flashing their 290X with a 390X bios, there is no reason to. Sure you get an extra 50MHz clock jump, which you can easily do manually. But a big chunk of the performance boost on the 390X comes from memory changes, which a 290X wont support.

 

[LTC8K6]

Why are people/sites comparing the MSI 390X to non MSI 290X's? This has been covered to death, the MSI cards use more, regardless of it being AMD or nVidia.

As for people flashing their 290X with a 390X bios, there is no reason to. Sure you get an extra 50MHz clock jump, which you can easily do manually. But a big chunk of the performance boost on the 390X comes from memory changes, which a 290X wont support.

A couple threads on flashgate:

http://www.overclock.net/t/1561278/amd-r9-290-290x-flashed-to-r9-390-390x

:biggrin:

 

[krumme]

A couple threads on flashgate:

http://www.overclock.net/t/1561278/amd-r9-290-290x-flashed-to-r9-390-390x

A lot of crashing. Lol. But hey its fun.

 

[bryanW1995]

Power consumption is off the scale (and this review is more along the lines of every other review I've read). Yes, yes, there's always the come back "Enthusiasts don't care about power consumption" but this is a little too much to ignore.

http://www.techpowerup.com/reviews/MSI/R9_390X_Gaming/28.html

I'm, frankly, surprised that w1zzard missed that. Here's some information from SKYMTL that paints a different picture:

This is an interesting metric since we are comparing a hot-running reference-clocked R9 290X to a custom cooled yet higher clocked R9 390X. In addition, AMD’s newer card also has double the amount of memory that’s operating at substantially higher speeds.

The end result actually puts credibility behind AMD’s claims about their Grenada core’s efficiency improvements. As you will see in the result on the following pages, the R9 390X is a good 12% to 20% faster than the R9 290X and yet according to the numbers above, it consumes less than 5% more power. It certainly looks like something has been done behind the scenes to enhance Hawaii’s performance per watt ratio.

w1zzard thought that the card was 100% identical after taking it apart, then he tested the one brand that is known to be extremely power hungry. That's unfortunate for AMD, though, at the end of the day, I don't think that it really matters. Boring old tech is still boring old tech, whether you get 2% more performance or 9% more is moot. Fury is where all the fun is for this launch.

 

[werepossum]

Except that the rebrands seem to be able to do more...like 54 watts more vs a 290x reference :|

Yep, LTC8K6 was absolutely right - this is a rebrand with more and slightly faster memory, not a true refresh. Claims of increased power efficiency seem to be vaporware once one really looks at hard numbers. A very, very slight difference at best, considering that better memory takes less power to force it to a higher speed.

EDIT: So, am I right in thinking that unless one games at above 1080p the 300 series cards really offer nothing desirable and the GTX970 is still the best card?

And if so, does that apply also to ultra-high resolution textures such as we're likely to see from modders for Fallout 4, or does 8 GB VRAM deserve a caveat for that instance? Sadly, all the Fury cards will just be more than I can spend.

 

[digitaldurandal]

Yep, LTC8K6 was absolutely right - this is a rebrand with more and slightly faster memory, not a true refresh. Claims of increased power efficiency seem to be vaporware once one really looks at hard numbers. A very, very slight difference at best, considering that better memory takes less power to force it to a higher speed.

EDIT: So, am I right in thinking that unless one games at above 1080p the 300 series cards really offer nothing desirable and the GTX970 is still the best card?

And if so, does that apply also to ultra-high resolution textures such as we're likely to see from modders for Fallout 4, or does 8 GB VRAM deserve a caveat for that instance? Sadly, all the Fury cards will just be more than I can spend.

There are already a lot of threads debating on the usefulness of 8 GB.

Basically even above 1080p it hasn't been shown to be much of an issue, at least not with current single GPU performance.

Some people will post showing that their card has reserved over 4 GB of VRAM and say it is proof. [H] Will post saying that this is the case as well, going so far as to say the 290x is held back by it while having a 295x2 in the same game showing higher FPS and no stuttering.. Oops.

If the 295x2 isn't held back by 4 Gb per GPU the 390x will not be either.

 

[3DVagabond]

Yep, LTC8K6 was absolutely right - this is a rebrand with more and slightly faster memory, not a true refresh. Claims of increased power efficiency seem to be vaporware once one really looks at hard numbers. A very, very slight difference at best, considering that better memory takes less power to force it to a higher speed.

EDIT: So, am I right in thinking that unless one games at above 1080p the 300 series cards really offer nothing desirable and the GTX970 is still the best card?

And if so, does that apply also to ultra-high resolution textures such as we're likely to see from modders for Fallout 4, or does 8 GB VRAM deserve a caveat for that instance? Sadly, all the Fury cards will just be more than I can spend.

You do realize MSI Gaming series cards have higher power usage than other cards? I know it's been reported over and over. Or, have you honestly missed all of the other posts about it?

 

[LTC8K6]

What it boils down to is that the replacement for the 290X, coming out over 1.5 years later, is not much faster. Call it whatever you want.

You need a Fury variant to get that 1.5+ years of improvement.

 

[flopper]

What it boils down to is that the replacement for the 290X, coming out over 1.5 years later, is not much faster. Call it whatever you want.

You need a Fury variant to get that 1.5+ years of improvement.

Thats called 14/16nm die shrink.
we are stuck in a transition phase.
The die shrink will change things up for mid range whenever its ready that is.

 

[werepossum]

There are already a lot of threads debating on the usefulness of 8 GB.

Basically even above 1080p it hasn't been shown to be much of an issue, at least not with current single GPU performance.

Some people will post showing that their card has reserved over 4 GB of VRAM and say it is proof. [H] Will post saying that this is the case as well, going so far as to say the 290x is held back by it while having a 295x2 in the same game showing higher FPS and no stuttering.. Oops.

If the 295x2 isn't held back by 4 Gb per GPU the 390x will not be either.

Thanks. That's a good point.

You do realize MSI Gaming series cards have higher power usage than other cards? I know it's been reported over and over. Or, have you honestly missed all of the other posts about it?

I've seen just a bit, but even so, if you drive pretty much anybody's 290X card to 390X speeds there isn't much in the way of power savings here, especially given that better grade memory requires less power for the same speeds. With a card that's already so much more power hungry than the equivalent nVidia 970/980, do you really feel that "more power - but not quite as much more as you'd think" is a big selling point? Anyone that cares about the power is buying the nVidia cards, so while it's nice, seems more smoke than fire to me. Also, I read several reviews couple days ago and that was the general consensus, though granted with a couple outliers who were more impressed (though generally still calling them rebrands rather than refreshes.)

EDIT: Here's TechPowerUp's review of the MSI R9 290X Gaming card for comparison.http://www.techpowerup.com/reviews/MSI/R9_290X_Gaming/22.html 231W versus 344W for average, 369W versus 424W peak. Doesn't seem like a refresh to me, although as LTC8K6 points out the term is irrelevant. It is what it is, and what it is is hungry.