PowerColor Teases Radeon RX Vega Nano

Published by

Click here to post a comment for PowerColor Teases Radeon RX Vega Nano on our message forum
https://forums.guru3d.com/data/avatars/m/254/254725.jpg
Hey, I'll take anything with a GPU these days, it's been dull in the gaming graphics industry alright.
You can say that again.
https://forums.guru3d.com/data/avatars/m/16/16662.jpg
Administrator
Prince Valiant:

You can say that again.
Hey, I'll take anything with a GPU these days, it's been dull in the gaming graphics industry alright.
data/avatar/default/avatar05.webp
Fox2232:

That heatsink looks cheap. Does it eat like 120W? 8+6 pin does not match that cooling at all. Maybe for people who just buy it and slap liquid cooling on it.
8+6 pin?? That heatsink does not like a match for a full 6-pin πŸ˜€ πŸ˜€ 6-pin + PCIe= 75W + 75W = 150W
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
Pretty interesting - I wonder if this is something like a Vega 56 but with much lower clocks?
https://forums.guru3d.com/data/avatars/m/209/209146.jpg
schmidtbag:

Pretty interesting - I wonder if this is something like a Vega 56 but with much lower clocks?
The Sapphire Pulse has some pretty good components, loses a few power phases but still one of the best Vega models and it also has the higher-quality Samsung HBM2 modules which also clocks higher. πŸ™‚ (EDIT: Vega 56 chips only I believe.) https://www.reddit.com/r/Amd/comments/837fm8/sapphire_radeon_rx_vega_56_pulse_has_a_nanosized/ (Extender cooler though, similar to their Tri-X and Nitro Fury non-reference models.) If that's the same here then while the cooler is more reference looking well a decent after-market water cooler and perhaps a bios flash (voltages and other limits, worked on the Pulse from what I'm reading.) and this should be pretty good too, assuming it's not too stripped down PCB wise of course but I don't know if they're sticking closely to AMD's reference design but it was ultimately cancelled instead of AMD launching a Nano variant separately from the current '56 and '64 versions or if they're changing it up a bit more. (For both good and bad, some non-reference Vega models are not as good and others are pretty robust.) EDIT: On the downside the Pulse is the most expensive Vega non-reference model even compared against the Nitro+ Vega 64 full-length PCB. (Partially due to demand and partially from Sapphire setting a higher base price directly which then tends to increase even further depending on retailers but it seems they're fluctuating a bit currently.) Although I don't see anything about pricing in the above article so perhaps they'll place it a bit lower than current MSRP's for Vega though I wouldn't be surprised to see this going from anything between 600 to 800 USD or higher.
https://forums.guru3d.com/data/avatars/m/271/271560.jpg
i think it's great. it's perfect for sff gaming rigs at 1440p, add in Freesync with a ryzen+, and even if it's relatively expensive, the system value is present. it would be an awesome "steam box" console killer.
https://forums.guru3d.com/data/avatars/m/175/175902.jpg
Noisiv:

8+6 pin?? That heatsink does not like a match for a full 6-pin πŸ˜€ πŸ˜€ 6-pin + PCIe= 75W + 75W = 150W
You mistake dissipated heat in W and consumed W from electricity... Same name but not the same unit... To stay in AMD products: RX 460 have 6 pin + pcie (and so 150W possible under OC) but even overclocked the heat dissipated is bellow 75W (at least on the Sapphire i have tested)
data/avatar/default/avatar27.webp
rl66:

You mistake dissipated heat in W and consumed W from electricity... Same name but not the same unit... To stay in AMD products: RX 460 have 6 pin + pcie (and so 150W possible under OC) but even overclocked the heat dissipated is bellow 75W (at least on the Sapphire i have tested)
You are right sir. I WAS WRONG! You are correct in saying that the total heat (Joules, not Watts; heat is a form of energy) produced by the GFX card is smaller than the total electrical energy consumed by the card. To get the total heat (produced or dissipated - same thing; we are not accumulating the heat, else something would melt), To get the total heat we take the total electrical energy fed to the graphics card and then we subtract all the little consumers: the energy needed to run the fans (not all of it, because some if it gets transformed back to heat due to friction) energy consumed by RGB/lights (again not all of it; you've guessed it - some of it ends up as heat) the energy used on crackling noises, coil whine or any other audio whose source is the graphic card (but not the fans, that's already accounted for) All-in-all maybe 5W * the observed time interval. What remains is - the heat. Electrical power fed to gfx card = Heating power + ~5W Why? Because Electrical energy comes in, and it is transformed into Ohmic heating, frictional heating, moving air around, sound and light. There are no other energy transformations, no additional work. If you're imagining something like "useful computational work", it's already accounted for - moving electrons around ends up as Ohmic heating. To stay in AMD products: RX 460 have 6 pin + pcie If the heating power dissipated is 75W, then its using around ~78W of electrical power. And vice versa.
https://forums.guru3d.com/data/avatars/m/248/248994.jpg
Noisiv:

To get the total heat we take the total electrical energy fed to the graphics card and then we subtract all the little consumers: the energy needed to run the fans (not all of it, because some if it gets transformed back to heat due to friction) energy consumed by RGB/lights (again not all of it; you've guessed it - some of it ends up as heat) the energy used on crackling noises, coil whine or any other audio whose source is the graphic card (but not the fans, that's already accounted for)
Actually the card also "loses" some of the energy taken in due to having the cable connection to one or more displays (unless it's a bloody miner's rig). I have no idea how much, but if the card doesn't shine like a Christmas tree, wouldn't it be more than the energy consumed by the leds?
https://forums.guru3d.com/data/avatars/m/238/238382.jpg
It's so cute!!
data/avatar/default/avatar15.webp
Kaarme:

Actually the card also "loses" some of the energy taken in due to having the cable connection to one or more displays (unless it's a bloody miner's rig). I have no idea how much, but if the card doesn't shine like a Christmas tree, wouldn't it be more than the energy consumed by the leds?
Yeah I guess so... LEDs are like few tens of mW. DP/HDMI cable connection is few hundreds mW. Lets say a whole Watt and call it quits. Then again where did this Watt go? Does you video cable power your monitor - no it doesn't. So what form of energy did this 1W during 1 second = 1 Joule, finally end up as? That's right - 1 Joule of pure HEAT πŸ™‚
https://forums.guru3d.com/data/avatars/m/248/248994.jpg
Noisiv:

Then again where did this Watt go? Does you video cable power your monitor - no it doesn't. So what form of energy did this 1W during 1 second = 1 Joule, finally end up as? That's right - 1 Joule of pure HEAT πŸ™‚
Yeah, but that heat is entirely out of the PC case. Some of it is due to the cable's internal resistance, transferring into the room air, some of it will heat up the IC chip in the screen parsing the signal coming from the video card. So, as far as the video card cooling goes, that joule is free of worries.
https://forums.guru3d.com/data/avatars/m/145/145154.jpg
I love small things that do big jobs.