• 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.

Golden Vega 56?

Associate
Joined
3 Mar 2012
Posts
293
Hi All

Recieved my new Sapphire Pulse Vega 56 couple days ago, I really seem to have some decent clocks for mega low voltage. Check my screenshot. This is stable at folding and gaming. During folding core runs at ~1609mhz, memory 950mhz and ~960mv. During gaming core clock runs at ~1600mhz and memory 950mhz again at 960mv. Power draw is ~160W and my temps are great, core, memory and hotspot run between 55c-70C with fans at 70%.

Could this be considered a golden 56 considered the relatively decent clocks at the voltage?

Thanks

Ekbc4C3.jpg
 
Memory looks to be 900 in your screen shot.


Sorry I upped the memory this afternoon and it's running at 950mhz. Would you say that's golden given the voltage? I can't see many other reports of similar people with same clocks at the low voltage?
 
Just tried your settings @ 960mv and my core holds 1560MHz during Superposition benchmark. I need to push my volts up to 1020mv to hold >1600MHz. Not sure how Superposition compares to folding load/utilisation. The coolers on these things are great:).
 
Looks average or just above average to me, not like it's doing 1700+ @ 1100mv :D


Righto I'll give that a try see if I can do it. I had it getting just shy of 24k graphics score at 1690mhz at 1000mv few days ago but apparently the score was abit low for the clock speed.
 
This is quite normal, the Pulse are very good clocking cards as Vega responds to low temps, so the improved cooler helps a lot there.
We see them generally all sustaining 1600-1650MHz in gaming, the golden ones sustain 1650-1700MHz core with some exceptional ones holding 1700+ on the pulse cooler, switch these to water and we have seen them around 1800 at which point they are Vega 64 / 1080/1080ti performance. Which at £280 really shows you the insane bargain these cards.

We generally leave the memory alone but of course.
 
Core clocks looks very good, but my very old Vega56 (first batch from OCUK) can run memory at 1090mhz on Vega64 bios. Got a EKWB on it. I wonder how yours will perform water cooled. Shame its custom PCB.


Currently testing 1635mhz core and 950mhz memory at 960mv. Can't push memory past 950mhz on this bios atm. It's Samsung memory though so I could flash it.
 
This is quite normal, the Pulse are very good clocking cards as Vega responds to low temps, so the improved cooler helps a lot there.
We see them generally all sustaining 1600-1650MHz in gaming, the golden ones sustain 1650-1700MHz core with some exceptional ones holding 1700+ on the pulse cooler, switch these to water and we have seen them around 1800 at which point they are Vega 64 / 1080/1080ti performance. Which at £280 really shows you the insane bargain these cards.

We generally leave the memory alone but of course.

qYxdY6F.jpg

Latest testing results, any good? I've also got p6 and p7 both set to same voltage, don't know if that makes a difference I've read they should be different.
 
Last edited:
Looking good.

You may get higher fps with slightly higher voltages. These cards actually back-off the MHz if voltages are deemed too low. If you do not already have it, download 3DMark Superposition, then run above settings with 1050mv, 1075mv and 1100mv. If you get the same scores/sustained MHz with all three voltages, then 1050MHz is completely fine. If you achieve higher scores with higher volts, then 1050mv will be on the limit. I use Superposition to determine the minimum voltage needed to obtain sustained-MHz. The sweatspot for my card appears to be P6 1612MHz / P7 1677MHz @ 1050mv, although I then bump volts up to 1075mv just to be sure I am not too close to the edge of stability. At this setting I obtain a stable ~1630MHz within everything I run. I could push higher, but power-consumption and fan-noise start to become bothersome as I use a single-fan inlet/outlet MATX case.
 
I take it you mean Unigine Superposition?

I really need to mess around more with my wife's V56 Pulse. Was going to slap it on my test bench but then got side tracked when it (the Test bench) wouldn't post (cue 2-3 hours of head scratching and swearing).

Just checking, with the above is the power limit already set to +50% (150%) before messing around with undervolting?
 
I take it you mean Unigine Superposition?

I really need to mess around more with my wife's V56 Pulse. Was going to slap it on my test bench but then got side tracked when it (the Test bench) wouldn't post (cue 2-3 hours of head scratching and swearing).

Just checking, with the above is the power limit already set to +50% (150%) before messing around with undervolting?
Yes, sorry Ungine Superposition. You can also use 3DMark but I find Superposition pushes GPU's pretty hard and is a good stability test tool. Power Limit at +50% is also safe. It allows the card to pull more amps from the 12v line if/when needed, rather than cut fps/MHz to meet AMD's advertised TDP.
 
Back
Top Bottom