Radeon 7 (Vega 2, 7nm, 16GB) - $699 available Feb 7th with 3 games

Where are you finding this information? The amd release notes read like they improved stability of the overclocking software, not of overclocking itself. A single user on the amd subreddit reported good OC results in the new driver discussion thread but there's no quantitative before/after info that I've been able to find.

Yar, and according to the just released [H] review, overclocking is still borked. Valid point on the fan noise, sounds like that issue can be avoided using Chill.
 
You are a another mindless Nvidia fan boy. Tell me how ray tracing is doing in all those dozens of games that support it. RTX failed from launch. The hardware can't handle the demands of ray tracing. The price is outrageous for very small performance gains if RTR id turned off. A b.s. technology called DLSS that degrades image quality with shimmering artifacts in order to create an AI faked-resolution and FPS scores. DLSS is only good for competitive gamers who care little about image quality and only about frame rates. Driver updates in the first week of availability of a video card that are significant are fairly common even for Nvidia. Are you afraid of a retest because perhaps your blessed RTX 2080 may be shown not to be so overpowering afterall? I also hope when the review is repeated with the new drivers that they also test Star Control Origins from Stardock. I would like to see how the RTX 2080 stacks up against Radeon VII with that game.

I have an nVidia card in my system, true. It's the first nVidia card I purchased in ten years. I could have purchased a Vega 56 if it were released 13 months sooner than it was. If I had the money I would have purchased a Radeon VII and a ThreadRipper system but since my wife and I are purchasing house plans to have a house built that ship has sailed.

With that said, I never ever expect slack to be cut for piss poor releases. The fact that you attack me for being a fanboi for stating a simple fact (don't expect reviewers to cover up for your fuck ups) places you in the realm of fantasy and fanboi-ism well beyond my apparent bias. I also did not and will not purchase an RTX card; I guess I should go out and get one since I'm a closet fanboi.

Grow up.
 
Yar, and according to the just released [H] review, overclocking is still borked. Valid point on the fan noise, sounds like that issue can be avoided using Chill.
Chill? your nuts chill lowers performance. You just hit auto undervolt in wattman and that lowers fan speed and temps. You are wrong . I have 2 sources wit thousands of followers on YouTube who have done the testing and are now able to do substantial manual overclocks that they were unable to do when they did their initial reviews.
 
Last edited:
I have an nVidia card in my system, true. It's the first nVidia card I purchased in ten years. I could have purchased a Vega 56 if it were released 13 months sooner than it was. If I had the money I would have purchased a Radeon VII and a ThreadRipper system but since my wife and I are purchasing house plans to have a house built that ship has sailed.

With that said, I never ever expect slack to be cut for piss poor releases. The fact that you attack me for being a fanboi for stating a simple fact (don't expect reviewers to cover up for your fuck ups) places you in the realm of fantasy and fanboi-ism well beyond my apparent bias. I also did not and will not purchase an RTX card; I guess I should go out and get one since I'm a closet fanboi.

Grow up.
"coverup for your fuck ups" Man you have lost it. I give only facts about the new Radeon 7 driver and what it can do and state it alters the terrain and suggest new reviews be done. Then you go off on a rampage. You are the one with anger management issues
 
"coverup for your fuck ups" Man you have lost it. I give only facts about the new Radeon 7 driver and what it can do and state it alters the terrain and suggest new reviews be done. Then you go off on a rampage. You are the one with anger management issues

And new reviews cost absolutely nothing to produce and edit?

AMD produced a Valentine's Day driver, not a Groundhog Day driver. You don't get a do-over for a launch no matter how badly it went.

I only have anger management issues around stupid people.
 
Where are you finding this information? The amd release notes read like they improved stability of the overclocking software, not of overclocking itself. A single user on the amd subreddit reported good OC results in the new driver discussion thread but there's no quantitative before/after info that I've been able to find.

"not an apple fan" pn You Tube did extensive testing and reported all results in his video. Jim at Adored TV also did test . Both of their tests confirm the improved overclocking of memory and gpu. I do NOT invent anything I report here. All is based on test by guys who know what they are doing and have solid reputations.
 
And new reviews cost absolutely nothing to produce and edit?

AMD produced a Valentine's Day driver, not a Groundhog Day driver. You don't get a do-over for a launch no matter how badly it went.

I only have anger management issues around stupid people.

Get a life.
 
Yar, and according to the just released [H] review, overclocking is still borked. Valid point on the fan noise, sounds like that issue can be avoided using Chill.

H review did not use todays released drivers. They did (Adored TV) and Not an Apple fan both from YouTube.
 
I have an nVidia card in my system, true. It's the first nVidia card I purchased in ten years. I could have purchased a Vega 56 if it were released 13 months sooner than it was. If I had the money I would have purchased a Radeon VII and a ThreadRipper system but since my wife and I are purchasing house plans to have a house built that ship has sailed.

With that said, I never ever expect slack to be cut for piss poor releases. The fact that you attack me for being a fanboi for stating a simple fact (don't expect reviewers to cover up for your fuck ups) places you in the realm of fantasy and fanboi-ism well beyond my apparent bias. I also did not and will not purchase an RTX card; I guess I should go out and get one since I'm a closet fanboi.

Grow up.
So in a nutshell with you first impressions are always correct??? The world would be in a shitbasket far worse than it is now, if everybody followed that practice.
 
So in a nutshell with you first impressions are always correct??? The world would be in a shitbasket far worse than it is now, if everybody followed that practice.

I'm not paying money and/or time to give my reviews of impressions.

With that said, over time you have confirmed my first impression of you.
 
Can you sweethearts carry the bickering to PM? I know it's valentines day and all but enough with the PDA..

Just set mine to auto undervolt. May as well give it a try.
 
Can you sweethearts carry the bickering to PM? I know it's valentines day and all but enough with the PDA..

Just set mine to auto undervolt. May as well give it a try.

Please let us know how you get on. I'm always interested in the average users experience when it comes to things like this.
 
So in a nutshell with you first impressions are always correct??? The world would be in a shitbasket far worse than it is now, if everybody followed that practice.

AMD majorly fucked up their press and initial public driver release. They only have themselves to blame for the negative press at launch. AMD knew the drivers sucked and still gave them to the press and still launched the card with bad drivers. The drivers being better now will be reflected in newer testing reviewers do on these cards, but no one should be forced to take pity on AMD for AMD's own fuck up and re-do reviews because they unfucked their drivers after launch.
 
With wattman set to auto undervolt fans peaked at about 2350rpm around 75c. And then, once temps kept climbing, it actually REDUCED fan speeds to around 1900. I quit the test at that point as temps were still climbing and fans were not responding. This was with furmark 1920x1080, 8x AA.

My experience is, it's broken, and I wouldn't use it. Maybe it magically works better for someone else. Manually setting fan curves hasn't worked well for me either with wattman commanding much higher fan speeds at points along the curve than I have graphed.

FWIW, wattman has been a total piece of trash even with my asus rog strix vega64. This isn't a hardware issue, it's crap software, and I have always felt the experience with afterburner and nvidia graphics cards was more desirable.
 
These updates aren't magically making the hardware better. Retesting manual undervolting with 19.2.2 yielded identical results to 19.2.1 for me.

Wattman wasn't properly saving/applying settings previously such that any 19.2.1 adjustments (especially auto-tuning) should be thrown out. The Wattman issues are embarrassing for AMD and should have been resolved before the cards even reached reviewers. Probably best to shelve overclocking for a few weeks while they sort out the other issues.
 
These are the best settings I've come up with so far. This is after 15 minutes of furmark. That fan curve, which is mostly flat, so far hasn't resulted in any excursions above 2800rpm. Basically I have two speeds. Dead silent out of games, and 2800rpm in games. 950mv.
 

Attachments

  • IMG_20190214_222849.jpg
    IMG_20190214_222849.jpg
    501.6 KB · Views: 0
AMD majorly fucked up their press and initial public driver release. They only have themselves to blame for the negative press at launch. AMD knew the drivers sucked and still gave them to the press and still launched the card with bad drivers. The drivers being better now will be reflected in newer testing reviewers do on these cards, but no one should be forced to take pity on AMD for AMD's own fuck up and re-do reviews because they unfucked their drivers after launch.


I am not a fanboy. I agree that AMD stumbled a coup[le of times with the Radeon 7. First the should have delayed the availability for 1 week so the driver issues would be solved. That would have mitigated a lot of the initial coolness to many reviews. They also did not produce enough cards and that was probably by plan since they are making only a few dollars on each card, I believe what actually prompted Lisa Su to go ahead with this project was the delays in Navi. Navi was pushed back almost 90 days because a design flaw found in the first engineering samples after initial tape-out. They had to fix the flaw and then tape-out a second time. I suspect poor leadership on the project allowed that to happen. You remember they announced a major departure from RTG about 2 months ago. That guy was in charge of the project and I am sure he got the boot for his failure. These problems do happen but since AMD is already almost 2 years behind Nvidia because of their lack of research and development from when they were cash starved, they can't afford these f--ckups at this time. I hope the new management in RTG has a better grip on things so this does not happen again. So I am well aware of the warts and all, but the only way the graphics market is not plagued by stagnation and gross monopoly pricing is if their is vibrant competition. I am actually glad Intel is going to produce discreet graphics. it will force Nvidia and AMD to step up their game, be better in their execution of projects, be more innovative, and keep prices at a more reasonable level,
 
AMD majorly fucked up their press and initial public driver release. They only have themselves to blame for the negative press at launch. AMD knew the drivers sucked and still gave them to the press and still launched the card with bad drivers. The drivers being better now will be reflected in newer testing reviewers do on these cards, but no one should be forced to take pity on AMD for AMD's own fuck up and re-do reviews because they unfucked their drivers after launch.

It is NOT a question of pity. It is a question of presenting an accurate statement of the current state of Radeon 7. When facts change, reality changes, and users have a right to know the new reality so they can make tjheir proper choices base on current fact. I guess you riding on your high horse have never screwed up or stepped in doo-doo??? That would be a first case in human history. What matters is what happens after an error is made. Is it repeated or is there a positive learning experience. This is NOT 6 months later. The errors should not have happened , Radeon 7 was released 7 to 10 days early. But AMD has cleaned up their errors and turned out a product not nearly as bad as first impressions. So you don't think users here and on other review sites have a right to know the current level of Radeon 7 performance? That would be cavalier and disingenuous. Let us criticize what is wrong and embrace efforts that quickly fix error.
 
Has anyone posted about this DigitalTrends article showing Radeon VII beating 2080 Ti at some tests?

https://www.digitaltrends.com/computing/amd-radeon-vii-review/

Something is wrong with their testing honestly... they have the Ti only 12% faster than a 2080 in Fire Strike

They appear to be using the overall score instead of GPU (not a good metric when comparing just the GPU) but still with a Ti I get a 22.7K overall score. I have the same CPU as them too and it's not overclocked ( 1950X). I'm heavily CPU bottlenecked in FS also, which hurts the overall score.
 
Well mine is on it's way from NewEgg. ASRock branded one popped into stock so I nabbed it. I have had good luck with their RMA process recently on a mobo and their gpu warranty is 3 years, so figured I'd roll the dice since they are all the same cards anyway. Now just need to figure out if my son gets my Asus Vega 64 to replace his GTX 1070 and sell the 1070? Might as well , would be the first time ever that I'd be all AMD graphics cards in my house. Daughter is running a RX 570, but her case is too small I think for the Strix.
 
AMD majorly fucked up their press and initial public driver release. They only have themselves to blame for the negative press at launch. AMD knew the drivers sucked and still gave them to the press and still launched the card with bad drivers. The drivers being better now will be reflected in newer testing reviewers do on these cards, but no one should be forced to take pity on AMD for AMD's own fuck up and re-do reviews because they unfucked their drivers after launch.

I'm not sure how much I want to agree with os2wiz on anything as he seems overly zealous, but I will say that no tech review site "has to" change anything they've posted or do anything. That's their prerogative.
However, considering that all of these sites are businesses and they generate their money through traffic and advertisements, it's probably advantageous to have a reason/excuse to create more content that people are interested in (such as updated benchmarks with new drivers).
You could argue that from the tech-journalist standpoint, having a bunch of drama (clickbait) and eventual fixes is preferable to having a smooth launch. Just so they can keep reporting on the story as it progresses and therefore continue to generate traffic. I wouldn't call any of that pity. It's business sense.

AMD releasing faulty drivers (which I don't dispute) is a totally different story. One that any tech site should be indifferent to.
 
Last edited:
Reporting on a story and redoing a review are two very different animals, especially when, in the case of [H]ardOCP, it's not a simple 'launch game, hit benchmark button, collate data after some period of time elapses'. The reviews here, which is why I put so much weight on them, are real-world actual gameplay runthroughs. Kyle is continuing to post in his review thread. He is undoubtedly going to be presenting an overclock article soon which will show a vast improvement to the product.
 
I am not a fanboy. I agree that AMD stumbled a coup[le of times with the Radeon 7. First the should have delayed the availability for 1 week so the driver issues would be solved. That would have mitigated a lot of the initial coolness to many reviews. They also did not produce enough cards and that was probably by plan since they are making only a few dollars on each card, I believe what actually prompted Lisa Su to go ahead with this project was the delays in Navi. Navi was pushed back almost 90 days because a design flaw found in the first engineering samples after initial tape-out. They had to fix the flaw and then tape-out a second time. I suspect poor leadership on the project allowed that to happen. You remember they announced a major departure from RTG about 2 months ago. That guy was in charge of the project and I am sure he got the boot for his failure. These problems do happen but since AMD is already almost 2 years behind Nvidia because of their lack of research and development from when they were cash starved, they can't afford these f--ckups at this time. I hope the new management in RTG has a better grip on things so this does not happen again. So I am well aware of the warts and all, but the only way the graphics market is not plagued by stagnation and gross monopoly pricing is if their is vibrant competition. I am actually glad Intel is going to produce discreet graphics. it will force Nvidia and AMD to step up their game, be better in their execution of projects, be more innovative, and keep prices at a more reasonable level,

Where did this come from?

I am not aware of a "design flaw" after the "initial tape-out".
 
I forgot.

That’s the excuse FakeTV used for why Navi didn’t show up at CES


That's just like the digitimes business model. Make a up loads of rumor when 95% of them turn out falls keep making up new rumors about why the first rumors didn't come true...
 
Bummer. Overdriventool doesn't support Radeon VII yet. Would be a good alternative to wattman for on the fly fan control.
 
I forgot.

That’s the excuse FakeTV used for why Navi didn’t show up at CES

Yeah, the excuse were particularly weak this time and don't make make sense considering the pricing of any recent AMD card.

7nm die + 8gb of GDDR6 at this point in time is not a 249 dollar product.

If Radeon VII is somehow a loss at 699 or cost lol, switching to GDDR6 and having a slightly smaller die isn't going to save 450 dollars.

8gb of GDDR6 is 90 dollars and the price of HBM at a maximum is 300 dollars. Maybe 90 vs 130 for die for radeon VII and your selling at near cost if AMD were going to launch Navi now.

Like Nvidia, they are better selling an RX 580 at 200 than a Navi GPU at 250 since the former is profitable and they have excess quantities.

The 249/200 dollar Navi parts adoredTV proposed would have destroyed the value of AMD Polaris/Vega inventory. As a result, the pricing was absolutely fake for Navi. Not saying performance is impossible but pricing certainly was given AMD wants to make money and not lose it.

The leaks were manufactured as a juxtaposition to make the pricing of the RTX series look even more ludicrous than it already was.
 
Yeah, the excuse were particularly weak this time and don't make make sense considering the pricing of any recent AMD card.

7nm die + 8gb of GDDR6 at this point in time is not a 249 dollar product.

If Radeon VII is somehow a loss at 699 or cost lol, switching to GDDR6 and having a slightly smaller die isn't going to save 450 dollars.

8gb of GDDR6 is 90 dollars and the price of HBM at a maximum is 300 dollars. Maybe 90 vs 130 for die for radeon VII and your selling at near cost if AMD were going to launch Navi now.

Like Nvidia, they are better selling an RX 580 at 200 than a Navi GPU at 250 since the former is profitable and they have excess quantities.

The 249/200 dollar Navi parts adoredTV proposed would have destroyed the value of AMD Polaris/Vega inventory. As a result, the pricing was absolutely fake for Navi. Not saying performance is impossible but pricing certainly was given AMD wants to make money and not lose it.

The leaks were manufactured as a juxtaposition to make the pricing of the RTX series look even more ludicrous than it already was.

Exactly.

It doesn't pass the smell test.
 
Back
Top