Unfortunately I don't think the new beta drivers include a gaming mode anymore
Just installed them last evening and it doesn't have it anymore. Also the driver package is almost half the size, 17.6 was 550ish MB, this is 350MB...
I've seen people install the RX drivers instead of PRO and seems to work better with gaming, although I'm interested in the Pro aspects of this card as well. Hopefully someone in this forum can help with a few answers, at the moment I'm feeling a bit lost...
Thank you so much for replying.
Could I please ask you to open a separate thread on the Community Forum about that?
I am guessing it can get difficult for the AMD Community Support team to follow questions inside other main topics.
Also since you own a Vega FE and I am just asking questions at the moment, your support question should carry much higher priority and weight than mine.
Please ask about those Beta Drivers and why Gaming mode is not there, and when the Vega FE will include RX Gaming Drivers?
RE: I've seen people install the RX drivers instead of PRO and seems to work better with gaming.
That is a great piece of news but I would really prefer not to have to reinstall drivers for a different product with a different VBIOS if I do take the plunge and buy a Vega FE Liquid or Vega FE with a waterblock. It is a big investment for me at the moment which is why I am taking so much effort to investigate.
RE: Although I'm interested in the Pro aspects of this card as well.
Me too. Absolutely. Although they are not Certified Drivers, I think the Vega FE card is a great idea.
To be honest I wish the Radeon Technology Group would allow people with RX Vega 64 to purchase a separate upgrade Software Licence (& maybe VBIOS Flash) to allow to turn on
the Radeon Pro features in future, essentially turning their RX Vega 64 into an Vega FE 8GB.
If they got it right it would be a big selling point for Gamers who are interested in getting into more of the Pro type aspects in the future.
RE: Hopefully someone in this forum can help with a few answers, at the moment I'm feeling a bit lost...
Me too. I I totally understand what you mean.
I have the funds, I need the Compute Power, I want to stay with AMD, but the Gaming Performance/Power is a concern.
Currently the pricing is way above what I was expecting, and I think everyone needs some more in depth answers.
Could I ask a a question please?
Have you run your own Compubench / Luxmark / other OpenCL Performance Metrics yet?
I am trying to work out how much better a single RX Vega 64 or Vega FE will be than the cards I have been running at present.
I have just been running Compubench on a pair of my R9 Nano (FuryX versions) with the latest AMD Crimson ReLive WHQL Drivers for Windows 8.1 64bit.
My cards are reporting much better benchmark results than I see in Compubench for some reason (It only lists AMD Fury Series)...
I made sure to run with Crossfire off and I have tested both the Primary and Secondary Cards. I am running on PCIe 3.0 interface.
I don't know if it is the Drivers, or if Compubench is way out of date for the AMD Fury Series cards or what is going on.
I can see some data in Compubench for Vega cards, but now I am not sure how up to date that information is and it would be great to get a cross reference check.
Anyhow thanks again.
I am off to kick off a few runs of LuxMark next.
I've already opened a thread for my problems, sorry more of an FYI intervention there.
The gaming mode switch should automatically swap active drivers, and I believe as drivers mature, we will see a difference. For instance: current blender benchmark is at 1m24s which is not great at all considering Vega RX go as low as 35 on new drivers apparently(https://www.reddit.com/r/Amd/comments/6tnu9r/vega_64_20_faster_in_blender_than_gtx_1080ti/ )... bit of an alarm bell going on there, so I assume if/when these new drivers will be attached to FE suite, then it will make a difference when switching modes. But for now, 17.8.2 Beta is behind on performance. I am NOWHERE NEAR being able to replicate those results with the current drivers.
My system is a R7 email@example.comGHz OC, Asus Crosshair VI Hero, 2x8GB Corsair Dominator DDR4@3000MHz(2933 actual), 500GB SanDisk Ultra II SSD. I also have an EVGA 1080Ti FTW3 for comparison (FE came out after I bought the 1080Ti and it was too expensive at launch).
Let me know exactly what tests you want me to run, if you could post links to the software would be great, and I'll post the results this evening (London here btw...)
RE: I've already opened a thread for my problems, sorry more of an FYI intervention there.
No need to apologise ... thank you again for your information and help.
RE: Gaming switch ... yes I thought that the Gaming Switch / Mode was supposed to allow a Vega FE to stay on and run Stable, Production Ready "Vega FE Radeon Pro Drivers" for the Professional / Gaming Development Applications. I thought that the intention was the "Gaming Mode" would use a parts of the Vega FE Driver with the RX Vega Gaming Driver optimisation (the Beta Drivers perhaps?).
That is based on what Gamer's Nexus found out the intention is in any case.
At the moment they were calling Gaming Mode a Placebo Switch ...
Last time they tested it it made no difference at all to gaming performance and simply changed user options in the Crimsoin Software.
I do not know if that has changed at all ...
I am testing on i7-4790K@9% OC, Asus Z97 Deluxe NFC&Wireless, Crucial Ballistic LP Tactical DDR3@ 1600MHz, 1TB Crucial SSD.
I can supply exact OC levels later.
So far I am running on Windows 8.1 64bit.
I also have Window 7 64Bit Ultimate and Windows 10 Pro 64Bit if needed.
RE: Let me know exactly what tests you want me to run, if you could post links to the software would be great, and I'll post the results this evening (London here btw...)
1. Compubench Version 2.0 Desktop.
CompuBench - performance benchmark for various compute APIs (OpenCL, RenderScript)
2. LuxMark v3.0
LuxMark - LuxRender Wiki
I am testing Windows first and then I will be testing OpenCL on Ubuntu 16.04 and Fedora 26 but I need to wait for other work to complete as I need to get some other work completed for next Monday Morning. Hopefully I will complete the testing on Linux early next week.
Thank you again, I really do appreciate your offer of help.
While pricing was a concern for me initially as well, I have adopted a different mindset of late. Vega 64 can currently be found at Best Buy for $599, and still includes Prey and Wolfenstein. Now if you have no interest in the games, that may seem like a $100 markup, but in this age of cryptocurrency, older GPUs have retained their value more than ever before.
The R9 Fury X I have is currently selling on EBay for ~$400. That is nearly two thirds of it's MSRP two years after launch! So if I decide to sell my old GPU, the cost of an upgrade falls to $200 with two free games. I think that is a big part of the equation I didn't consider at first either. Demands outside of gaming may have driven up the cost of new hardware, but they have elevated older hardware as well. In my case the net effect has created the lowest cost to upgrade that I can remember.
ajlueke Agreed, if anyone is in your situation where they can and are willing to offset the rather inflated cost by selling older hardware, then yes, by all means go ahead, ride the wave while you still can, it's well worth the upgrade I'd say if only 200$ is actually being spent.
Price-performance for any newcomer gamer (purely gaming-wise) it's prohibitive for 64 and FE, borderline illogical in some cases. I am an AMD fan, don't doubt that, but if I was purely gaming, it wouldn't have been my final choice.
This digression aside, getting back to colesdav's conundrum, in a professional or creative environment, $100 price hike is not necessarily concerning considering yield of performance, which Vega seems to be more than capable of delivering, but for some reason fails to do so at this moment. It's a completely different investment perspective to mining.
colesdav I'll do some testing this evening for you and post the results once I have a complete batch. I'll also try and do a clean-up of the drivers and try using RX only to see if there's a difference. Bear in mind, we have different systems so performance WILL be different, I'd be amazed if it wasn't...