• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

The RX Vega 64 Owners Thread

Associate
Joined
4 Oct 2017
Posts
590
Location
Australia - Sunshine Coast
Why would they add x2 aa to the 4k benchmarks?
It's the default Ultra settings. He generally does just default settings for testing.

Interestingly Forza is still a single core CPU hog so having a ThreadRipper is not an advantage but a detriment.

GPU limited for only 93.1% of the run and 122FPS average vs HWU's 130 for a LC Vega64. Can guarantee my GPU was running harder at ~1744MHz most of the time. So yeah, roll on i7 9700K for single core limited games....
 
Last edited:
Soldato
Joined
8 Mar 2010
Posts
4,967
Location
Aberdeenshire
It's the default Ultra settings. He generally does just default settings for testing.

Interestingly Forza is still a single core CPU hog so having a ThreadRipper is not an advantage but a detriment.

GPU limited for only 93.1% of the run and 122FPS average vs HWU's 130 for a LC Vega64. Can guarantee my GPU was running harder at ~1744MHz most of the time. So yeah, roll on i7 9700K for single core limited games....
I see. As for single core limitation I'm not bothered about that. Its not really why I have one but even still, sooner or later threadripper will blow past that 9700k as soon as the first 16 core games start coming out. For now the performance is more then acceptable for me.
 
Associate
Joined
4 Oct 2017
Posts
590
Location
Australia - Sunshine Coast
I see. As for single core limitation I'm not bothered about that. Its not really why I have one but even still, sooner or later threadripper will blow past that 9700k as soon as the first 16 core games start coming out. For now the performance is more then acceptable for me.
Oh hell yeah, I'm getting the 9700K for gaming with older and single core limited games. It's likely to be the last of the Ringbus series CPU's so kinda fitting it'll be a retro gaming PC for me. It's also for iRacing as ThreadRipper is just inadequate at shoving 5.5GHz at a single core problem :D

Should be fun testing the different GPU's on the 9700K and the TR4 especially as games develop into the Multi-Threaded era.
 
Associate
Joined
19 Sep 2016
Posts
343
Location
Jensen's Leather Jacket
*innate rambling*

More and more I tinker with the Strix 64, the more I like it...

So I have it undervolted a little, nothing too crazy but it now at least hits the stated boost clock of 1,590Mhz, Target Temp set to 65°C (which incidentally helps with that one lonely naked Mosfet on the VRM -Strix owners will know..) and when coupled to the Freesync Monitor and using Radeon Chill to keep it within the monitors 40-75 range its an absolute joy to play with.

funnily enough, even on Shadow of the Tomb Raider with things cranked up it very rarely needs to clock upto the max boost to hit the monitors 75fps limit which means a tiddly power draw and temperatures often in the mid 50's..

But the biggest compliment I can give to the Vega 64 is the fact that the GTX 1080 is boxed and under my desk picking up dust :)
 
Soldato
Joined
22 Nov 2009
Posts
13,252
Location
Under the hot sun.
*innate rambling*

More and more I tinker with the Strix 64, the more I like it...

So I have it undervolted a little, nothing too crazy but it now at least hits the stated boost clock of 1,590Mhz, Target Temp set to 65°C (which incidentally helps with that one lonely naked Mosfet on the VRM -Strix owners will know..) and when coupled to the Freesync Monitor and using Radeon Chill to keep it within the monitors 40-75 range its an absolute joy to play with.

funnily enough, even on Shadow of the Tomb Raider with things cranked up it very rarely needs to clock upto the max boost to hit the monitors 75fps limit which means a tiddly power draw and temperatures often in the mid 50's..

But the biggest compliment I can give to the Vega 64 is the fact that the GTX 1080 is boxed and under my desk picking up dust :)

You can always replace the thermal pads on the VRMs with 3mm thick Thermal Grizzly. And since you will be there, replace the thermal paste on GPU & HBM with Thermal Grizzly liquid metal :p
Or at least Cryonaut.

Boomstick has made a discussion in here how he moodified his Strix to improve perf.

Also give a try to the settings found few pages back. They are very efficient (240W), providing really good perf, while they keep the card cool.
 
Associate
Joined
5 Oct 2018
Posts
13
Hello, i'm new here but i've been following this 3d for a long time while waiting for my Sapphire Vega 64 Nitro+ to arrive. 2 days ago I finally received it, but I found some big problems, since you are certainly much more experienced than me, maybe you can help me:

1) COIL WHINE: I have a Seasonic gold 80+ 750w power supply, I put two different 6 + 2 PCIE cables that go from the power supply to the video card. I disassembled and reassembled the cables at least 10 times to prevent them to touch, but nothing, every time I open Heaven bench (for example) you hear a strong high pitch coil whine. I read that it may depend on the PSU, but in my case I hoped that Seasonic was good enough.There is a way to make it stop? with time maybe it decreases? Or UV?

2) UNDERVOLT: I tried as I read on this forum in different replies to perform a slight undervolt, with the only overclocking of the HMB (to 1100mhz) leaving the p6 p7 frequencies untouched. So i went p6 1000mv - p7 1050mv HBM (floor voltage): 1000mv - Power limit + 50%. During Heaven it crash with the screen that turns black and I have to do a reset. If I set these settings instead: HBM mhz 1080 - p6 1050mv - p7 1100mv HBM (floor voltage): 1050mv - Power limit + 30% it complete the benchmark without crashing but I do not get big increases in performance (3460 with 69 ° 1920x1080 - ultra - extreme tassellation), the preset TURBO (3380 with 69 °). Am I doing something wrong? I have seen in this 3d some heavier undervolt / overclock.

3) GAME PERFORMANCE: After the undervolt (p6 1050mv - p7 1100mv HBM: 1050mv - Power limit + 30%) I tried to enter in Destiny 2 to see the differences with my previous video card (Rx470) and to my incredible disappointment I find out that the FPS (with the same graphic settings) have not really increased: with the rx470 I had fps that ranged from 70 - 80 now i have 60 - 90. the range min fps - max fps has increased, but honestly I expected a much bigger difference. Do I have to reinstall the game after changing the card?

For the sake of completeness I say that for the change of card I used DDU in safe mode to remove the drivers of the rx470, then I installed the Nitro + and finally the latest Adrenaline drivers (18.9.3).

My pc has

Ryzen 2600 overclock at 4.0 ghz
Ram 3000mhz
Motherboard: asus rog strix b350 f gaming

Someone can help me? I've waited so long for this card that now I want it to work :)
 
Last edited:
Soldato
Joined
8 Mar 2010
Posts
4,967
Location
Aberdeenshire
Hello, i'm new here but i've been following this 3d for a long time while waiting for my Sapphire Vega 64 Nitro+ to arrive. 2 days ago I finally received it, but I found some big problems, since you are certainly much more experienced than me, maybe you can help me:

1) COIL WHINE: I have a Seasonic gold 80+ 750w power supply, I put two different 6 + 2 PCIE cables that go from the power supply to the video card. I disassembled and reassembled the cables at least 10 times to prevent them to touch, but nothing, every time I open Heaven bench (for example) you hear a strong high pitch coil whine. I read that it may depend on the PSU, but in my case I hoped that Seasonic was good enough.There is a way to make it stop? with time maybe it decreases? Or UV?

2) UNDERVOLT: I tried as I read on this forum in different replies to perform a slight undervolt, with the only overclocking of the HMB (to 1100mhz) leaving the p6 p7 frequencies untouched. So i went p6 1000mv - p7 1050mv HBM (floor voltage): 1000mv - Power limit + 50%. During Heaven it crash with the screen that turns black and I have to do a reset. If I set these settings instead: HBM mhz 1080 - p6 1050mv - p7 1100mv HBM (floor voltage): 1050mv - Power limit + 30% it complete the benchmark without crashing but I do not get big increases in performance (3460 with 69 ° 1920x1080 - ultra - extreme tassellation), the preset TURBO (3380 with 69 °). Am I doing something wrong? I have seen in this 3d some heavier undervolt / overclock.

3) GAME PERFORMANCE: After the undervolt (p6 1050mv - p7 1100mv HBM: 1050mv - Power limit + 30%) I tried to enter in Destiny 2 to see the differences with my previous video card (Rx470) and to my incredible disappointment I find out that the FPS (with the same graphic settings) have not really increased: with the rx470 I had fps that ranged from 70 - 80 now i have 60 - 90. the range min fps - max fps has increased, but honestly I expected a much bigger difference. Do I have to reinstall the game after changing the card?

For the sake of completeness I say that for the change of card I used DDU in safe mode to remove the drivers of the rx470, then I installed the Nitro + and finally the latest Adrenaline drivers (18.9.3).

My pc has

Ryzen 2600 overclock at 4.0 ghz
Ram 3000mhz
Motherboard: asus rog strix b350 f gaming

Someone can help me? I've waited so long for this card that now I want it to work :)
I suggest to begin with you run a 3d mark test at power save, balanced and then turbo mode to start with then we can can figure out tweaking after the results.
 
Soldato
Joined
22 Nov 2009
Posts
13,252
Location
Under the hot sun.
Hello, i'm new here but i've been following this 3d for a long time while waiting for my Sapphire Vega 64 Nitro+ to arrive. 2 days ago I finally received it, but I found some big problems, since you are certainly much more experienced than me, maybe you can help me:

1) COIL WHINE: I have a Seasonic gold 80+ 750w power supply, I put two different 6 + 2 PCIE cables that go from the power supply to the video card. I disassembled and reassembled the cables at least 10 times to prevent them to touch, but nothing, every time I open Heaven bench (for example) you hear a strong high pitch coil whine. I read that it may depend on the PSU, but in my case I hoped that Seasonic was good enough.There is a way to make it stop? with time maybe it decreases? Or UV?

2) UNDERVOLT: I tried as I read on this forum in different replies to perform a slight undervolt, with the only overclocking of the HMB (to 1100mhz) leaving the p6 p7 frequencies untouched. So i went p6 1000mv - p7 1050mv HBM (floor voltage): 1000mv - Power limit + 50%. During Heaven it crash with the screen that turns black and I have to do a reset. If I set these settings instead: HBM mhz 1080 - p6 1050mv - p7 1100mv HBM (floor voltage): 1050mv - Power limit + 30% it complete the benchmark without crashing but I do not get big increases in performance (3460 with 69 ° 1920x1080 - ultra - extreme tassellation), the preset TURBO (3380 with 69 °). Am I doing something wrong? I have seen in this 3d some heavier undervolt / overclock.

3) GAME PERFORMANCE: After the undervolt (p6 1050mv - p7 1100mv HBM: 1050mv - Power limit + 30%) I tried to enter in Destiny 2 to see the differences with my previous video card (Rx470) and to my incredible disappointment I find out that the FPS (with the same graphic settings) have not really increased: with the rx470 I had fps that ranged from 70 - 80 now i have 60 - 90. the range min fps - max fps has increased, but honestly I expected a much bigger difference. Do I have to reinstall the game after changing the card?

For the sake of completeness I say that for the change of card I used DDU in safe mode to remove the drivers of the rx470, then I installed the Nitro + and finally the latest Adrenaline drivers (18.9.3).

My pc has

Ryzen 2600 overclock at 4.0 ghz
Ram 3000mhz
Motherboard: asus rog strix b350 f gaming

Someone can help me? I've waited so long for this card that now I want it to work :)

Did you use DDU to wipe the drivers?
What is the resolution you gaming at? Have you tried to use power save mode only?

Also can you use the overlay to see GPU usage? Can you run other games or 3d mark?
 
Associate
Joined
5 Oct 2018
Posts
13
I suggest to begin with you run a 3d mark test at power save, balanced and then turbo mode to start with then we can can figure out tweaking after the results.

Did you use DDU to wipe the drivers?
What is the resolution you gaming at? Have you tried to use power save mode only?

Also can you use the overlay to see GPU usage? Can you run other games or 3d mark?

Yes, DDU in safe mode, then change card and last Adrenaline driver install. I'm gaming at 1080p 144hz, in a couple of hours i'll do the test in 3d mark - timespy at power save, balanced and then turbo mode as Lokken said and then report here the results. Thanks you
 
Soldato
Joined
22 Nov 2009
Posts
13,252
Location
Under the hot sun.
Associate
Joined
5 Oct 2018
Posts
13
I suggest to begin with you run a 3d mark test at power save, balanced and then turbo mode to start with then we can can figure out tweaking after the results.

I did the benchmarks in 3dmark demo - Timespy:

Powersaving: 6131

https://ibb.co/dBpUGz

Balanced: 7116

https://ibb.co/cMzkpK

Turbo: 7234

https://ibb.co/fskVOe

In the balanced and turbo benchmarks the gpu utilization was always at 98 - 99%

This is my power draw in turbo mode according to Gpuz: 220w (i have seen here someone with 243w, is mine right?)

https://ibb.co/nEukpK

With those Wattman settings: HBM mhz 1080 - p6 1050mv - p7 1100mv HBM (floor voltage): 1050mv - Power limit + 30% i had in Heaven just a few more points compared to the Turbo mode. I think that even in 3d mark timespy the result will be the same

On Destiny 2 you should be looking something around these values with the 2600

https://www.gamersnexus.net/game-bench/3097-destiny-2-launch-cpu-benchmark-ryzen-vs-kaby-lake

No, i'm not even close to those numbers. I have high fps spikes (with higher coil whine), but i also have very low fps spikes. Don't know if its related to Destiny 2 game, i'll try with some other games and see
 
Last edited:
Soldato
Joined
8 Mar 2010
Posts
4,967
Location
Aberdeenshire
I did the benchmarks in 3dmark demo - Timespy:

Powersaving: 6131

https://ibb.co/dBpUGz

Balanced: 7116

https://ibb.co/cMzkpK

Turbo: 7234

https://ibb.co/fskVOe

In the balanced and turbo benchmarks the gpu utilization was always at 98 - 99%

This is my power draw in turbo mode according to Gpuz: 220w (i have seen here someone with 243w, is mine right?)

https://ibb.co/nEukpK

With those Wattman settings: HBM mhz 1080 - p6 1050mv - p7 1100mv HBM (floor voltage): 1050mv - Power limit + 30% i had in Heaven just a few more points compared to the Turbo mode. I think that even in 3d mark timespy the result will be the same


No, i'm not even close to those numbers. I have high fps spikes (with higher coil whine), but i also have very low fps spikes. Don't know if its related to Destiny 2 game, i'll try with some other games and see
Do you have access to the full version of timestrike so you can put it in windows mode and just loop test 1. Then you can play with the settings and see what works best.
Are the temps good and the clock speeds holding up? Maybe actually dropping p7 a little might help overall, it certainly does with me. I can set 50% PL if p7 is at 1600mhz or less. I also found that the lower the voltage the less drops in clock speed I get but that's probably because of the temperature. What happens when you max out the fan speed?
 
Associate
Joined
5 Oct 2018
Posts
13
Do you have access to the full version of timestrike so you can put it in windows mode and just loop test 1. Then you can play with the settings and see what works best.
Are the temps good and the clock speeds holding up? Maybe actually dropping p7 a little might help overall, it certainly does with me. I can set 50% PL if p7 is at 1600mhz or less. I also found that the lower the voltage the less drops in clock speed I get but that's probably because of the temperature. What happens when you max out the fan speed?

No, i don't have the full access to 3dmark becouse before now I've never needed to do thorough extensive benchmarking.

Since this graphic card has 3 months of life but I have bought it used I do not know what use it had been made of by the previous owner, or if he made changes to the bios (perhaps for mining). Is the 220w power draw in turbo mode correct for a Sapphire Nitro + non-LE? are these benchmark results in line with other vega 64s or are they lower?

Thanks you
 
Soldato
Joined
8 Mar 2010
Posts
4,967
Location
Aberdeenshire
No, i don't have the full access to 3dmark becouse before now I've never needed to do thorough extensive benchmarking.

Since this graphic card has 3 months of life but I have bought it used I do not know what use it had been made of by the previous owner, or if he made changes to the bios (perhaps for mining). Is the 220w power draw in turbo mode correct for a Sapphire Nitro + non-LE? are these benchmark results in line with other vega 64s or are they lower?

Thanks you
In that case just watch the clock speeds as the full test runs and keep an eye out for drops. The score in Turbo doesn't seem too far off. I think the power draw is ok I have similar results in that the reported power draw is never over 220w. Perhaps a bios thing. Also I'm getting only a slightly higher score in Timespy with the cpu at 4ghz and the best settings I could muster on gpu at the time. I'll try it again to compare but your score looks alright.
 
Associate
Joined
5 Oct 2018
Posts
13
In that case just watch the clock speeds as the full test runs and keep an eye out for drops. The score in Turbo doesn't seem too far off. I think the power draw is ok I have similar results in that the reported power draw is never over 220w. Perhaps a bios thing. Also I'm getting only a slightly higher score in Timespy with the cpu at 4ghz and the best settings I could muster on gpu at the time. I'll try it again to compare but your score looks alright.

About the temperatures, with my custom settings, after 40 minuts of benchmarking i get 69° on the gpu, 75/76° on HBM and 92° on Hotspot, a bit too high for a UV

I'll try to lower the p7 frequency from 1632 to 1600 with PL +50% and try to increase hbm to 1100mhz while lowering abit the voltages. Maybe can you give me your Wattman settings, so i can try them?
 
Soldato
Joined
8 Mar 2010
Posts
4,967
Location
Aberdeenshire
About the temperatures, with my custom settings, after 40 minuts of benchmarking i get 69° on the gpu, 75/76° on HBM and 92° on Hotspot, a bit too high for a UV

I'll try to lower the p7 frequency from 1632 to 1600 with PL +50% and try to increase hbm to 1100mhz while lowering abit the voltages. Maybe can you give me your Wattman settings, so i can try them?
Yea that's definitely going to be holding back the card a bit.
Your likely to get better results using a profile similar to someone who doesn't have the reference card like I do but if you want to try mine are;
1533mhz on p6, 1577mhz on p7, both are 1050mv. HBM voltage has to be equal to them and set the OC for it. 50% pL 62c temp target and I manually set the min fan speed to 2500rpm but with the max value set to max. Yours should be better but you can try.
 
Back
Top Bottom