MP 1,1-5,1 EVGA PowerLink with AMD GPUs

[MIKX]

You can plug a 6 pin male socket into the RIGHT SIDE of the Vega 56's 8 pin other female socket. .. look at the pics in post #1 in this thread again.

This is for the 2 X SATA to 1 X 6 pin cable .. this goes to the RIGHT SIDE of the other Vega 56's 8 pin female socket.

In any case powering a HD 7970 or a Vega 56 . . . power is power.

If not enough power your GPU can shut your cMP down or worse, burn traces on the motherboard under high load.
[automerge]1583272633[/automerge]

 

[flyproductions]

Is this not balanced, on top of the balancing the EVGA does?

No, in this configuration the balancing is done by the EVGA. But in fact like this you are just using it as a (much) more expensive and (in my opinion) less elegant version of the dual-mini-6pin to single male 8pin -> single female 8pin to dual male 6+2pin cabeling...which techincally does the exact same.

 

[KingCachapa]

No, in this configuration the balancing is done by the EVGA. But in fact like this you are just using it as a (much) more expensive and (in my opinion) less elegant version of the dual-mini-6pin to single male 8pin -> single female 8pin to dual male 6+2pin cabeling...which techincally does the exact same.

thanks! Understood.

Considering the slot blockage... pardon my ignorance but reading around I can't come to a conclusion – is the reference Vega 56 card a viable option compared to the Sapphire Pulse version? I'm debating returning the one I got and finding a reference one for the size, but I'm unclear if this is a good idea performance-wise + eventually placing a Squid 4 slot NVMe above it....

Should I just keep the Sapphire Pulse and throw it on slot 4 and start rearranging my storage?

 

[krakman]

thanks! Understood.

Considering the slot blockage... pardon my ignorance but reading around I can't come to a conclusion – is the reference Vega 56 card a viable option compared to the Sapphire Pulse version? I'm debating returning the one I got and finding a reference one for the size, but I'm unclear if this is a good idea performance-wise + eventually placing a Squid 4 slot NVMe above it....

Should I just keep the Sapphire Pulse and throw it on slot 4 and start rearranging my storage?

I'm running the reference Sapphire vega 56 + EVGA power link + Bios switched to low power setting + HW acceleration enabled.

works fine - never had a hard shutdown - the fans spin up when exporting video from FCPX but the rest of the time it is quiet.

The EVGA powerlink is usually available on ebay at bargain prices and contrary to other comments it is does have an extra capacitor circuit in there to smooth out the current draw. It fits onto the GPU nicely so you only need two mini 6pin to 6pin cables which reduces clutter inside the MAC.

Final point - you still get slot 2 free to install your squid.

 

[KingCachapa]

I'm running the reference Sapphire vega 56 + EVGA power link + Bios switched to low power setting + HW acceleration enabled.

works fine - never had a hard shutdown - the fans spin up when exporting video from FCPX but the rest of the time it is quiet.

The EVGA powerlink is usually available on ebay at bargain prices and contrary to other comments it is does have an extra capacitor circuit in there to smooth out the current draw. It fits onto the GPU nicely so you only need two mini 6pin to 6pin cables which reduces clutter inside the MAC.

Final point - you still get slot 2 free to install your squid.

thanks @krakman ? I already purchased everything including the Vega 56 and EVGA, but the Vega 56 I found unfortunately was the big Pulse (I went with it because of price but it seems to be pointed as the better variation of the Vega 56 too).

Does the low power BIOS affect the card in performance? Sounds great.. I might return my Pulse and get that reference Vega!

 

[krakman]

thanks @krakman ? I already purchased everything including the Vega 56 and EVGA, but the Vega 56 I found unfortunately was the big Pulse (I went with it because of price but it seems to be pointed as the better variation of the Vega 56 too).

Does the low power BIOS affect the card in performance? Sounds great.. I might return my Pulse and get that reference Vega!

IIRC in the early days of GPU compatibility the Mac driver had a bug affecting the fan on the reference cards, I think it ran full speed all the time and it made the card draw all its power from the mini 6pin plugs and nothign from the PCIE slot. This bug did not affect the pulse card which is why it was the preferred choice. Since then the fan bug has been fixed and the card now draws some power from the PCIE slot so unless someone points out something I missed the reference has better spec than the Pulse???

Low power bios is something like -20% power draw = -5% performance Toms hardware or engaget did a detailed review a few years back you will need to google it.

 

[KingCachapa]

IIRC in the early days of GPU compatibility the Mac driver had a bug affecting the fan on the reference cards, I think it ran full speed all the time and it made the card draw all its power from the mini 6pin plugs and nothign from the PCIE slot. This bug did not affect the pulse card which is why it was the preferred choice. Since then the fan bug has been fixed and the card now draws some power from the PCIE slot so unless someone points out something I missed the reference has better spec than the Pulse???

Low power bios is something like -20% power draw = -5% performance Toms hardware or engaget did a detailed review a few years back you will need to google it.

Got it, yeah you are making sense, and I've just been told low power bios makes it perform even better than fullpower, at least on our cMPs, however I just saw this post that someone reported losing boot screen when switching to lowpower bios (obv. in a flashed 56). ? only thing spec. wise standing out is the cooling is better on the Pulse but I'm thinking by the time I have problems with this card if ever, the VII prices will have gone down and I can upgrade later.

OK, I've done more testing on the card, remember that this is a MacVidCard Radeon Vega 56 with the Boot screen. I remembered the small BIOS switch on the card (low power setting or Full power). I flipped the switch towards the I/O connectors and run Luxmark again. Here are the results: the consumption is a little lower but the score is higher(26695), than the full power setting. Secondly, there is no Boot Screen, as with any regular PC card.
I, then, flipped the BIOS switch back towards the power connectors on the card and run the test again. I posted the results once again below. I got the BOOT screen back, the power consumption went up as before (Post #332) but the Luxmark score went down (24413).
So the card is flashed in FULL POWER normal mode (not power saving mode), since I get no BOOT screen with the switch set on that mode.
I'm not complaining about the card, I like the performance and the fact that it is pretty quiet, plus the BOOT screen ( I honestly missed it for awhile)

Just realized that the WATT consumption has been reduced to minimal on the screenshots (sorry about this). In normal BIOS setting, the consumption was around 86 WATTS for the 2 ports) while for the low power mode setting, the WATT consumption is identical to the one on post #332, sorry about that once again

 

[krakman]

Got it, yeah you are making sense, and I've just been told low power bios makes it perform even better than fullpower, at least on our cMPs, however I just saw this post that someone reported losing boot screen when switching to lowpower bios (obv. in a flashed 56). ? only thing spec. wise standing out is the cooling is better on the Pulse but I'm thinking by the time I have problems with this card if ever, the VII prices will have gone down and I can upgrade later.

Interesting to read that report from a flashed vega 56 with boot screen, although there is no telling what tinkering has taken place in the flashed ROM, my unflashed card doesn't have boot screen (unless you install Opencore) Luxball score = 26422

In real world scenario you probably won't notice the difference between the pulse and reference card. The fan in my card rarely speeds up and on ocasion when it does its not as loud as the mac Pro cpu fans.

What are planning to use the GPU for? Video editing? games?

The real question is can you live without the second 16x pcie slot? I have an old 4x mSATA board with 3TB of raid SSD it maxes out at 800Mbs and its fast enough for my video editing needs (multicam weddings/conferences)

I am also waiting for the VII to drop in price! ???

 

[tommy chen]

I am also waiting for the VII to drop in price!

you will have to wait longer, the new and used prices go through the roof : -(

 

[KingCachapa]

Interesting to read that report from a flashed vega 56 with boot screen, although there is no telling what tinkering has taken place in the flashed ROM, my unflashed card doesn't have boot screen (unless you install Opencore) Luxball score = 26422

In real world scenario you probably won't notice the difference between the pulse and reference card. The fan in my card rarely speeds up and on ocasion when it does its not as loud as the mac Pro cpu fans.

What are planning to use the GPU for? Video editing? games?

The real question is can you live without the second 16x pcie slot? I have an old 4x mSATA board with 3TB of raid SSD it maxes out at 800Mbs and its fast enough for my video editing needs (multicam weddings/conferences)

I am also waiting for the VII to drop in price! ???

View attachment 897417

& @IndioX yeah I hear you on the pricing. @krakman I really don't want to lose the ability of using the 2nd PCIe slot because it's one of the two fastest connections we have in there. I want to keep the gpu on Slot 1 (bottom) and a multi NVMe (x4) PCIe card on Slot 2. I also have a PCIe SSD (SATA from OWC) that I'd like to use off of Slot 3, and last but not least I have the USB PCIe on Slot 4.

I do a range of visual work from regular design to motion graphics and video editing / grading etc.

This works for me because of $$ invested so far and the coronavirus is affecting my workload at the moment – so I'd like to steer clear of having to do even more mods that involve more time and money. They say it's already affecting the prices of parts and such. With this setup I have all my HDD bays filled as well for more storage, which I use for mostly backup and inactive / infrequently used files. My further backup process step would be to then backup as well to an external device but I haven't quite gotten there yet.

I'm trying to figure out what's the fastest viable connection for me to bring in media INTO the system (right now my fastest connection to bring in footage from external SSD superfast drives / various CF / SD cards is USB 3.x or WiFi :/). Any advice on this point?

Not to go too off topic but which 4x mSATA board are you using? I have 2 2.5inch SSDs I'd like to put to use but currently only one is using max speeds (the other is in an HDD bay via adapter).

 

[Norbert Mikołajczyk]

Hi guys!

I read the entire thread, it looks very promising as I want to use power link to feed my Radeon VII.
I want to use 3xSata to 8pin + 2x6 pin to 8 pin + Pcie. Will this be enough to fully load the slightly OC'ed RVII without shutdowns or backplane premature failure? Or the 2xSata to 8Pin will be enough instead of 3xSata?

Also where I could find 3xSata to 8-pin, I'm searching the Amazon and no results...

Best regards

 

[tsialex]

Hi guys!

I read the entire thread, it looks very promising as I want to use power link to feed my Radeon VII.
I want to use 3xSata to 8pin + 2x6 pin to 8 pin + Pcie. Will this be enough to fully load the slightly OC'ed RVII without shutdowns or backplane premature failure? Or the 2xSata to 8Pin will be enough instead of 3xSata?

Also where I could find 3xSata to 8-pin, I'm searching the Amazon and no results...

Best regards

It's a very bad idea to power a VII the way you intend. Even using all the SATA ports, you are still powering the GPU from the backplane power circuit, all power that go to the SATA ports are originating on the backplane.

One thing is to power a VEGA56 from the backplane, another totally different is a VII. Make a Pixla's mod.

 

[Norbert Mikołajczyk]

It's a very bad idea to power a VII the way you intend. Even using all the SATA ports, you are still powering the GPU from the backplane power circuit, all power that go to the SATA ports are originating on the backplane.

One thing is to power a VEGA56 from the backplane, another totally different is a VII. Make a Pixla's mod.

Thanks!

 

[AlexMaximus]

Thanks!

Hi there,

It's really not very difficult to do it, just follow the instructions on one of the mentioned websites above to do the Pixlas mod. If you need further help with the mod, just pm me. All that it is, is to properly install the yellow connectors on the cable in the right order as described. You can use some spare cable and test with a view ones there before you start to do the actual thing. Just take your time. The most important thing is to put the other cable into the mac pro before you crimp the opposite connectors on it, otherwise, you need to fumble the finished cable with connectors through the very tight hole in the corner. (not shown below) This takes as long as the whole mod if you forget it.

 

[Norbert Mikołajczyk]

Hi there,

It's really not very difficult to do it, just follow the instructions on one of the mentioned websites above to do the Pixlas mod. If you need further help with the mod, just pm me. All that it is, is to properly install the yellow connectors on the cable in the right order as described. You can use some spare cable and test with a view ones there before you start to do the actual thing. Just take your time. The most important thing is to put the other cable into the mac pro before you crimp the opposite connectors on it, otherwise, you need to fumble the finished cable with connectors through the very tight hole in the corner. (not shown below) This takes as long as the whole mod if you forget it.
View attachment 964611

Thanks, I see now that’s the only good way to go ?

 

[Emen Mali]

can I use the PowerLink EVGA with a Red Dragon 5700xt in a Mac Pro 4,1->5,1? Or should I just do the Pixlas Mod?

 

[flyproductions]

can I use the PowerLink EVGA with a Red Dragon 5700xt in a Mac Pro 4,1->5,1? Or should I just do the Pixlas Mod?

I would strongly recommend the pixlas mod! I have an oced GTX 1080 with a tdp slightly over 200 watts. And even i do see peaks of 95 watts to the mainboard connectors, using the PowerLink. Your card's tdp is 225!

 

[Emen Mali]

I would strongly recommend the pixlas mod! I have an oced GTX 1080 with a tdp slightly over 200 watts. And even i do see peaks of 95 watts to the mainboard connectors, using the PowerLink. Your card's tdp is 225!

Just saw the how to on YouTube, so I ordered what I needed for the mod. Don't want to risk my soon to be 12 yr old Mac Pro!