ASUS ROG Strix RX Vega 64/56 Release Date???

[panzlock]

As per thread title, has anyone heard anything other than "September" for release of the RX Vega Strix? With GPU prices increasing and corporate greed taking advantage of the current cryptomining infatuation I want to know how much this new variant will cost. The reference RX Vega 56 has already seen a price increase which makes it around $50 CAD more expensive than existing GTX 1070 stock. In the past reference card models have typically been more expensive than AIB partner cards, however I'm still concerned that the 56 in particular will be sold at GTX 1080 prices given that the 64 is already priced where the 1080 Ti sits.

I picked the wrong time to assemble a new build. Although prices on other components have been reasonable with rebates, the price of new VGA's really nullifies the savings I've been able to achieve thus far. Again, I've heard little about the ROG Strix Vega other than September release date, some pictures and a respectable performance review which i can no longer find. And guru3d also took down their review due to an ASUS error which found the wrong card in guru3d's hands:

http://www.guru3d.com/news-story/review-asus-radeon-rog-rx-vega-64-strix-8gb.html

Lets get it together, people.

[panzlock]

I find it strange, too. How can they release a product without one of the major specifications?

Porsche: "We're releasing a new version of the 911."

Public: "How much horsepower does it have?"

Porsche: "......We can't tell you, but it's expensive."

[Asmoranomar]

These publishers...by that same logic....
oh wait hold on...

Guys, looks like ASUS officially released over a month ago...remember when they removed the specs from the webpage? That was their official launch, a totally quiet card that consumes 0 watts of power and puts out 0 TDW, even when overclocked. I mean 0mhz is pretty fast, and a 10x overlclock is just, wow....right?

No wait don't...somehow publishers will spin it into some crazy divide by zero and now we have a video card that's more powerful than a quantum computer...ugh.

Edit: "Theoretically, we have a card that can outperform a 50-qbit processor by a factor of 400, but test samples by reviewers show that the cards are under performing reference designs."

Edit2: I want my card ; ;

[panzlock]

More info:

While ASUS has not yet revealed clockspeeds on the RX Vega 56 card, eTeknix has gotten their hands on the ROG Strix RX Vega 64 graphics card and figured out the clocks for that card. Specifically, the Vega 64 card clocks its 4096 GPU cores at 1298 MHz base and 1590 MHz boost. The site further lists the memory clockspeed at 945 MHz which doesn't appear to be overclocked as it matches the referece Vega 64 HBM2 clocks of 1890 MHz. Users can use the GPU Tweak II software to push the card further on their own though.

https://www.pcper.com/news/Graphics-Cards/ASUS-Launches-ROG-Strix-RX-Vega-64-and-56-OC-Edition-Graph...

I reckon these clock speeds are the same as the ones posted on the ASUS products page before it was taken down. You guys planning on releasing a card that performs worse than reference? Or were there other tweaks to take care of before release? Like AMD drivers that seemed to boost performance significantly.

[Asmoranomar]

panzlock wrote:
More info:



https://www.pcper.com/news/Graphics-Cards/ASUS-Launches-ROG-Strix-RX-Vega-64-and-56-OC-Edition-Graph...

I reckon these clock speeds are the same as the ones posted on the ASUS products page before it was taken down. You guys planning on releasing a card that performs worse than reference? Or were there other tweaks to take care of before release? Like AMD drivers that seemed to boost performance significantly.

I believe there was a more recent German(?) review not too long ago that showed that the ASUS cards were actually performing...slightly better reference.

Apparently, according to a (another) German source (I'll just paste it):

" Asus told us, ... Also with the clock rates, there is now more information - according to Asus product page clocks the Strix RX Vega 64 with 1,590 MHz (GPU) and 945 MHz (RAM)..."

So either ASUS told them, or they got it from the product page. This was an update to the original post where the information was not available on the product page initially, so ...yeah...

There was also a recent Powercolor review that was posted within the past week, I've lost the link, but it performed benchmarks on both the normal and OC settings. Oddly enough, it also had problems under OC, and performed poorly. This turned out to be driver incompatibility issues, but Powercolor wanted the review published anyways to get as many reviews out there as possible. The reviewer complied, but put a heavy emphasis that the OC portion was due to an improper test setup and only the normal clock settings should be considered.

The reviewer actually did a sidebar on the reasons why performance was slow - I can't quite remember the details point for point, but without the supported drivers the card is forcibly clocked to reference. It is actually more detailed than this but the result is it was running reference clock.

SO...is it possible ASUS cards have had similar issues all along (forced reference clock)? Who knows...

ANYWAYS...yeah, you most likely hit it spot on. But ASUS isn't saying anything... If it is a driver incompatibility issue, would be nice to know which driver to use. Especially if it's not out yet and OC is DOA until a compatible driver is released.

Edit: I found the (powercolor) video, It's pretty long so I'll just shorten it up. Driver 17.10.1 worked, 17.10.2+ did not. The issue was that the driver would misread voltages and would cause the card to believe that it was about to overheat or working in an improper configuration. The drivers would force reference settings to correct the issue. They were able to override the driver briefly with utilities and the card would perform better, but the driver would still reset the card and prevent an actual test from being able to be performed accurately. THIS ISSUE WAS VERIFIED BY POWERCOLOR AND AMD. They were able to run a few tests before the review was pushed to publish and the 17.10.1 drivers worked but not enough testing was done to make it to final. I don't know if this affects ASUS cards, but if it does it would explain some of the mess we've seen.

[panzlock]

ASUS asked Guru3d to remove the review article of the ROG Strix RX Vega 64 because they sent them the card with the wrong BIOS. So...they sent JayzTwoCents a card with the wrong BIOS, as well??? I think this was driver, and not BIOS related from the start.

This new AMD update should allow all AIB cards to gain: https://www.youtube.com/watch?v=lJe2yOfoQb0

I was truly hoping AIB's were looking to find a stable undervolt whilst increasing performance but I suppose that was a little too sanguine.

[Asmoranomar]

https://www.gamersnexus.net/guides/3164-testing-amd-partner-card-driver-problems-strix-powercolor-ve...

https://youtu.be/Zc_XgsX2P40

Today, gamersnexus just published an article about similar details, but this is also about ASUS testing.

Bad news: The ASUS card is still having problems even on the 17.11.4 beta drivers, while Powercolor seems to have a 7% improvement.

[Guyvergamingtv]

at this rate i might just go nvidia for the first time since the 90s lol had one near when they started and was so bad i never went back.

[panzlock]

ASUS ROG Strix RX Vega 56 – The card for serious 4K gamers

https://mygaming.co.za/news/industry-news/122581-asus-rog-strix-rx-vega-56-the-card-for-serious-4k-g...

......

[Heini]

panzlock wrote:
[url]https://mygaming.co.za/news/industry-news/122581-asus-rog-strix-rx-vega-56-the-card-for-serious-4k-gamers.html

Interesting, yes. Having a Vega FE with a Raijintek Morpheus II Core and two Noctua 120's I can assure you that the temps are ok but not great with the exception that the Raijintek Morpheus II Core makes for a 3.5 slot card. As for performance I have a Titan Xp that laps the field in fps with a 1080Ti close in tow and that leaves the AMD's offering a distant third. Consider the prices before choosing your next GPU! PLEASE, this is NOT a slight at ASUS as AMD leaves their aib's little room to move other than adding better cooling and/or lights in comparison to the AMD reference cards.

[panzlock]

Heini wrote:
Interesting, yes. Having a Vega FE with a Raijintek Morpheus II Core and two Noctua 120's I can assure you that the temps are ok but not great with the exception that the Raijintek Morpheus II Core makes for a 3.5 slot card. As for performance I have a Titan Xp that laps the field in fps with a 1080Ti close in tow and that leaves the AMD's offering a distant third. Consider the prices before choosing your next GPU! PLEASE, this is NOT a slight at ASUS as AMD leaves their aib's little room to move other than adding better cooling and/or lights in comparison to the AMD reference cards.

It's not what they're saying that I find interesting. More so where this "information" is coming from.

I'll be a little more comprehensive on my point of view. ASUS, in August, announced the ROG Strix RX Vega and if I'm not mistaken it was the first AIB Vega card to be tested. The results, although not disastrous in my opinion were poor given the Strix's inability to outperform its reference counterpart. Guru3d and JayzTwoCents provided some feedback on performance and I remember guru being asked to take down their review as a result. ASUS simultaneously dispatched the clock speeds from their site. Clearly there was a problem. ASUS indicated it was the BIOS.

Now, present day, PowerColor released their Red Devil and suddenly the Strix was making headlines again, with hints of imminent release but no dates OR clock speeds on their site? When tested by GamersNexus the PC card performed better than the Strix, and better still following AMD's latest driver update. The Strix had no such performance escalation. Why? AMD's fault? ASUS's fault? No one knows. No one willing to admit.

Now the 56 getting praise from another site despite no thorough specification data......Interesting!!! If you do a search the 56 is a much better option compared to the 64. It's quite capable and the power draw is significantly reduced when pitted against its own sibling. But I'm extremely skeptical about news pertaining to the Strix because since August there was nothing but drivel from ASUS regarding release and performance. It would be nice to get an update regarding the source of the problem. But as it stands, ASUS is slated to release a custom Vega card that is sub par to AMD's reference card. PowerColor doesn't have that problem. Not with the 64, anyway.

I know retailers in Canada will have the Strix, XFX and Sapphire versions. But not sure about PowerColor. A custom Vega 56 is sure to be better than the 1070 and possibly the 1070 Ti, hence this is the card I want to purchase for my daughter's build. This was also her wish, AMD, Vega...and Strix. But I had to show her the Red Devil because thus far it's the better of the two custom options. Luckily she likes the subdued, modest styling of the PC card. However, my problem is that it may not be available in Canada and I'll have to pay yet ANOTHER premium (HBM 2, anyone?) to have it shipped here.

So, in conclusion, I just want some closure from ASUS regarding their progress with Vega. I understand it's not that simple but AMD already made a mockery of the Vega release. Now ASUS seems to be building upon that failure. PowerColor did it right. It was late, but they took what they had, made it a little better and then released it once it was tweaked as well as could be. ASUS did the opposite by announcing a product they couldn't tune well enough, inducing us with false hopes. AMD is p[partially to blame, yes...but so is ASUS.

And speaking of ASUS and the ROG Strix brand, they better get that tragedy called the B350-F sorted quick, fast and in a hurry.