The 15-game ‘Premium’ VR Review – Pascal vs. Vega benchmarked for the Oculus Rift

Martini1991 · 4 Jan 2018 at 12:27

melmac said:
The 17.11 drivers were very good. But, you know that the Adrenalin drivers have been very poor. Lots of new features but they needed more testing time.

The only reason I am suggesting that the Site use 17.11.4 is because they are the only ones that work with the Red Devil Vega cards the review used.

Regardless the best way to test to remove any bias is to use the latest driver. There's no stable vega driver so you can't do latest stable. If it's a poor showing for AMD then it is what it is.

melmac · 4 Jan 2018 at 13:49

Martini1991 said:
Regardless the best way to test to remove any bias is to use the latest driver. There's no stable vega driver so you can't do latest stable. If it's a poor showing for AMD then it is what it is.

Not sure I agree. For either Nvidia or AMD, I think the best driver should be used. 99% of the time, that will be the latest driver, but sometimes drivers are released that are particularly poor. It doesn't happen often. I have seen review sites like HardOCP to this from time to time, use an earlier driver if there are issues with the latest driver.

melmac · 4 Jan 2018 at 14:16

Gregster said:
You have gone off on one a bit needlessly in truth. Comparing latest drivers is the way forward or even comparing an older NVidia driver to a newer NVidia driver would be cool but you getting 'a little bit frustrated at the timing of the review' strikes me as silly and I personally appreciate the work that people put into testing.

Gone off on one? What? I haven't gone off on one at all. What nonsense are you spouting? Comparing an older Nvidia driver to a newer Nvidia driver is cool but me expecting a site to do the same thing with an AMD driver is "needlessly going off on one" Hahahaha

I was waiting months for them to do a follow up review, then they finally get around to it and they are using drivers which aren't great. They have used the latest drivers, if they had done the review a few weeks ago the latest drivers would have been 11.4. That's what I meant by bad timing. That's it. That's all, I haven't criticised the site, there is no drama here that you and Googaly are trying to attribute to my comments.

Gregster said:
I personally appreciate the work that people put into testing.

A bit of a weird comment, Do you think that the site owners are reading this thread? Are you trying to score brownie points or something? I haven't said anything bad about the site so not sure why you said something so random?

Oh Wait, I get it, you think people who are critical of something can't appreciate it. A bit of childish attitude don't you think?

Gregster · 4 Jan 2018 at 14:21

melmac said:
The 17.11 drivers were very good. But, you know that the Adrenalin drivers have been very poor. Lots of new features but they needed more testing time.

The only reason I am suggesting that the Site use 17.11.4 is because they are the only ones that work with the Red Devil Vega cards the review used.

Were they the 17.11 or 17.11.2 or the 17.11.3 or the 17.11.4 drivers that should be tested? As a tester myself, it becomes a complete nightmare scenario to test each and every driver to see what gives the best performance. Maybe 17.11 gave great performance in Batman but not such good results in Elite Dangerous, so should they keep swapping and changing to suit the fanboys?

As for your last quote, I won't stoop to that low with a response.

melmac · 4 Jan 2018 at 14:34

Gregster said:
Were they the 17.11 or 17.11.2 or the 17.11.3 or the 17.11.4 drivers that should be tested? As a tester myself, it becomes a complete nightmare scenario to test each and every driver to see what gives the best performance. Maybe 17.11 gave great performance in Batman but not such good results in Elite Dangerous, so should they keep swapping and changing to suit the fanboys?.

Trying to be smart Gregster and failing hard. If you read the post I quoted you would have your answer. 17.11.4's are the only ones that work with the custom cards used in the review.

Gregster said:
As for your last quote, I won't stoop to that low with a response.

What, there is nothing wrong with my previous reply to you. IF you appreciate the work sites do on testing, good for you. It was just a bit of an odd statement considering I never said one bad thing about the site doing the review.

Oh BTW, did you apologise to the Mod in the other thread that you accused of been AMD biased, when he was right all along?

Gregster · 4 Jan 2018 at 14:47

melmac said:
Trying to be smart Gregster and failing hard. If you read the post I quoted you would have your answer. 17.11.4's are the only ones that work with the custom cards used in the review.

What, there is nothing wrong with my previous reply to you. IF you appreciate the work sites do on testing, good for you. It was just a bit of an odd statement considering I never said one bad thing about the site doing the review.

Oh BTW, did you apologise to the Mod in the other thread that you accused of been AMD biased, when he was right all along?

Not trying to be smart at all and you seriously missed the point. I mention 4 drivers in the 17.11 branch and as a tester myself, it becomes a nightmare testing each and every driver and hotfix. Some hotfixes reduce performance, whilst some will add. Some add stability and no performance gains, some add performance gains but have stability issues and some lower performance but add stability. Not sure how you missed that part but then again, not once did I accuse you of lambasting the site and you somehow managed to get "you said bad things about the site" from me saying "I appreciate the work people do in testing" (because I have done it on many occasions and it is a real chore). Not sure why you have gone off on a tangent either but I always apologise if I feel I was wrong.

maonayze · 4 Jan 2018 at 15:00

Woah!!! Hold yer horses there Mel & Kim......err sorry....Mel & Greg. Calm down, calm down. Come on guys we all know that no drivers are perfect for every living game or program at any one time from either company. A good driver for one game could possibly screw up another game very easily. It's just noticed by players of those particular games, whatever they may be at the time. I imagine that creating a driver set is like firing at moving targets. And our moving targets are made up of lots of different parts in different combinations. I'm sure it aint an easy job. Lets not get into a slanging match like the other recent thread....blimey that was crazy and even though I was tempted...I resisted the urge to go bounding in (2 strikes already

). Has the season of good will to all men passed that quickly.

Gregster · 4 Jan 2018 at 15:05

I have remained polite under duress

Mark Poppin · 4 Jan 2018 at 15:41

We actually do read some comments on our articles as we see the links and views generated from here. Thank-you for your comments.

As to the drivers we picked, we have determined long ago to always pick the very latest drivers for our major reviews or there are accusations of bias and "cherry picking" drivers. AMD's VR performance has significantly increased over the past 4 months since we lasted tested VR. It's impossible to say that 11.4 is overall better than 12.2 for all 15 games that we picked. And we saw good stability in VR with 12.2.

We plan to revisit this VR evaluation every 3-4 months at BTR, and each time we will increase the size of our benching suite.

nashathedog · 4 Jan 2018 at 15:47

melmac said:
Dammit, they used the Adrenalin drivers There is a drop in performance in VR games using the 12.1 and 12.2 drivers. The 11.4 drivers perform much better. For instance in Robo Recall I can run the game with 1.3 pixel density in high settings and get no frame drops using 11.4. But when I use the Adrenalin drives I have to turn the pixel density down to 1.

What amazes me is how often this type of blunder happens with so call tech review sites that are meant to be bang up to date with current tech news, They should have known to not use the Adrenalin drivers or wait.

As for AMD they should know that releasing drivers that contain lots of new software changes that are likely to be buggy just as the staff are off on their Crimbo holidays isn't the best thing to do.

r22snapper · 4 Jan 2018 at 16:13

Mark Poppin said:
We actually do read some comments on our articles as we see the links and views generated from here. Thank-you for your comments.

As to the drivers we picked, we have determined long ago to always pick the very latest drivers for our major reviews or there are accusations of bias and "cherry picking" drivers. AMD's VR performance has significantly increased over the past 4 months since we lasted tested VR. It's impossible to say that 11.4 is overall better than 12.2 for all 15 games that we picked. And we saw good stability in VR with 12.2.

We plan to revisit this VR evaluation every 3-4 months at BTR, and each time we will increase the size of our benching suite.

Nice one, and thanks for the response. I have to agree that most recent drivers is the only fair way of doing it.

Gregster · 4 Jan 2018 at 19:06

Mark Poppin said:
We actually do read some comments on our articles as we see the links and views generated from here. Thank-you for your comments.

As to the drivers we picked, we have determined long ago to always pick the very latest drivers for our major reviews or there are accusations of bias and "cherry picking" drivers. AMD's VR performance has significantly increased over the past 4 months since we lasted tested VR. It's impossible to say that 11.4 is overall better than 12.2 for all 15 games that we picked. And we saw good stability in VR with 12.2.

We plan to revisit this VR evaluation every 3-4 months at BTR, and each time we will increase the size of our benching suite.

Good work and hope I get some brownie points

Razor Time · 5 Jan 2018 at 17:08

Mark Poppin said:
We actually do read some comments on our articles as we see the links and views generated from here. Thank-you for your comments.

I was wondering if you next round of VR Benchmarks you could a few newer or more popular games like:

Robo Recall
Lone Echo
Arktika.1
Elite Dangerous
From other Suns
Arizona Sunshine

Mark Poppin · 5 Jan 2018 at 20:07

Razor Time said:
I was wondering if you next round of VR Benchmarks you could a few newer or more popular games like:

Robo Recall
Lone Echo
Arktika.1
Elite Dangerous
From other Suns
Arizona Sunshine

Absolutely. From the conclusion of the review:

We are playing and will add Lone Echo, Raw Data, ARKTIKA.1, Robo Recall and Arizona Sunshine to our next VR review for a total of 20 games.

I'll also consider adding From Other Suns and perhaps Elite Dangerous which I have in my library. Ultimately, I intend to benchmark 25 VR games as my regular benching suite.

I can't stop by here as often as I would like. If you have any further requests you can email me -
apoppin@gmail - or else comment on the article. I get regular notice of new comments although it may take me awhile to reply. Thank-you.

I just got a Red Devil RX Vega 64 and I'm back to PC game benching with 35 games and it will be awhile before I can get back to VR.

Razor Time · 5 Jan 2018 at 22:55

Mark Poppin said:
Absolutely. From the conclusion of the review:

I'll also consider adding From Other Suns and perhaps Elite Dangerous which I have in my library. Ultimately, I intend to benchmark 25 VR games as my regular benching suite.

I can't stop by here as often as I would like. If you have any further requests you can email me -
apoppin@gmail - or else comment on the article. I get regular notice of new comments although it may take me awhile to reply. Thank-you.

I just got a Red Devil RX Vega 64 and I'm back to PC game benching with 35 games and it will be awhile before I can get back to VR.

Excellent and good to see so many games (25 VR) lined up for the next round up.

JediFragger · 5 Jan 2018 at 23:46

Gregster said:
I have remained polite under duress

That's your main problem here G, just go ape**** like the rest of them would

Gregster · 6 Jan 2018 at 11:08

JediFragger said:
That's your main problem here G, just go ape**** like the rest of them would

LMAO

Gregster · 6 Jan 2018 at 11:35

https://overclock3d.net/news/gpu_di...ve_headset_with_new_year_s_resolution_tweet/1

A new Vive with a bigger resolution? I will buy one.

JediFragger · 6 Jan 2018 at 14:30

Gregster said:
https://overclock3d.net/news/gpu_di...ve_headset_with_new_year_s_resolution_tweet/1

A new Vive with a bigger resolution? I will buy one.

Nah, I'll buy one when it's as simple as slipping on a pair of sunglasses tbf. Until then, no dice

melmac · 6 Jan 2018 at 14:44

Gregster said:
Not trying to be smart at all and you seriously missed the point. I mention 4 drivers in the 17.11 branch and as a tester myself, it becomes a nightmare testing each and every driver and hotfix. Some hotfixes reduce performance, whilst some will add. Some add stability and no performance gains, some add performance gains but have stability issues and some lower performance but add stability. Not sure how you missed that part but then again, not once did I accuse you of lambasting the site and you somehow managed to get "you said bad things about the site" from me saying "I appreciate the work people do in testing" (because I have done it on many occasions and it is a real chore). Not sure why you have gone off on a tangent either but I always apologise if I feel I was wrong.

You accused me of going off on one and called me silly for been frustrated at the timing of the review. Your whole comment was silly in the line of the conversation because I wasn't going off and one and what's wrong with me been a little bit dismayed at the timing of the review especially when I know how bad and buggy the Adrenalin drivers are. If anything I am more frustrated with AMD and how badly they messed up with these drivers.

You also completely misunderstood my posts. Didn't you read the bit where I said the latest drivers should be used 99% of time? I am only saying that websites shouldn't use drivers which are broken and with serious problems when doing reviews. It's not bias, it's common sense. It's not rocket science Greg. Company releases Drivers A, then B, then C, Then D. D driver turns out to be messed up, what drivers do tech sites use in their review, Driver C. In this case, Adrenalin drivers faulty, what was the latest driver before the Adrenalin drivers? 17.11.4. Use that in the reviews.

You see, you guys think, you and Googaly, that I am complaining about the review and the drivers used because AMD are behind. But, that's not it at all. I have been complaining about the Adrenalin drivers long before this review came out. Ask Nashathedog, who I have quoted below, I have been saying that there is something seriously wrong with the Adrenalin drivers since they have been launched. They are buggier than all the previous drivers, they are slower and some people have actually thought their cards were faulty because the drivers are so bad. Every day new reports of games not working. And now culminating in not just been buggy for certain games, but breaking nearly all DX9 games. Seriously, Do you really think that tech sites should be using drivers this bad in reviews? No way, they should be advising people to roll back to the 17.11 drivers until AMD sort the current ones out. I thought the same when Nvidia released the drivers around April 2016, can't remember the exact numbers, but the ones that were causing computers to crash and blue screen.

nashathedog said:
What amazes me is how often this type of blunder happens with so call tech review sites that are meant to be bang up to date with current tech news, They should have known to not use the Adrenalin drivers or wait.

As for AMD they should know that releasing drivers that contain lots of new software changes that are likely to be buggy just as the staff are off on their Crimbo holidays isn't the best thing to do.

You know first hand how bad the drivers are. You were even thinking about returning your card again!! For the life of me I can't understand what the problem is with what I am suggesting. Use the latest drivers for your reviews, unless the drivers are crap, like the Nvidia drivers last year or the Adrenalin drivers. Then just use the ones directly before that or wait for a fix to come out for the current ones.

AMD should know better, the DX9 thing has really blown up in their face. They have done so much good work in the past few years of getting drivers out quick for game releases, solving bugs quickly without introducing new ones and generally producing good quality drivers. They actually were slowly changing people's opinions and their drivers were been called better than Nvidia's. All that hard work is gone now with one bad driver release. It looks like they rushed the Adrenalin drivers out without proper testing just to get them launched in December. You are right, they should have waited until January.

And sorry for the late replies, I wasn't online for the last day and half due to family emergency.