![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
All true, but packing all those components together in such a small space does not bode for great heat transfer. And already pushing 325w or 350w? Oof. I don't think Samsung has left a lot of headroom in these chips. Asus is already touting a 400W model. I would think that is about topped out.
3090/3080/3070
No clocks, no prices.....hmmm.
There will be 3+ slot cards as well I was told.
The cooler defines heat transfer, it doesnt diminish with PCB size.
Roger that.
It is only going to be an issue on FE cards anyway. And there will not be many of those for sale, so....
Shame, it's a great looking design. Gonna be interesting to see those thermals reviewed.Roger that.
Yeah, won't say I won't be missing this at all, but NVIDIA would not sample me anyway after GPP.
I understand better what you were trying to say, it wasnt about GPU power but heat from the smaller components packed in a small space.Roger that.
Thanks for letting me know.I understand better what you were trying to say, it wasnt about GPU power but heat from the smaller components packed in a small space.
NVidia take enough care to ensure air flow through the cooler over components that get too warm, I wouldnt worry.
If anything gets too warm still, it will have a small heatsink on it (or a pad to the large heatsink) or a larger component will be chosen that dissipates the heat over a larger volume (surface area).
They dont cock up that often
I think these new coolers look great. I just need to see how fast one warms my office up and how loud it sounds. I am already having 290X flashbacks.
lol, I fitted an Accelero Xtreme III cooler to my 290x, worked a treat but didnt change the amount of heat generated.Thanks for letting me know.
Yes it is. Good catch. These cards are going to be HOT.
Not sure how that will work out, Jensen stated that 3070 performance was basically equaling a 2080Ti, Nvidia 3070 is rated at 220w, Nvidia 2080Ti 250w. 90% per/watt increase looks more like 10%. Must have gotten the 90% and 10% backwards.
based on power transients which may overload the power supply, is rather entertaining. I wonder if they will include an audible alarm to go with that?
And, we actually built our own equipment to test transient power response. I do not think any other review site ever covered that, and that is what ASUS is focusing on. You can ask @[spectre] He knows a lot more about this than me.
Annnnnd now I need a new case, too.
Considering the new-found partnership they have with LG and their OLED TVs, I wouldn't be surprised if they were using a prototype.
Not sure how that will work out, Jensen stated that 3070 performance was basically equaling a 2080Ti, Nvidia 3070 is rated at 220w, Nvidia 2080Ti 250w. 90% per/watt increase looks more like 10%. Must have gotten the 90% and 10% backwards.
ASUS spending extra money and putting flashing light indicators on their 3000 series 400w Ampere (AMPS and more AMPS) card, selling it as a feature that is needed
The evga XC looks like it is exactly what you want. Someone posted a pic of it on the EVGA thread, can't tell if it uses a custom pcb.
I think pretty much every AIB card will be using a "custom" PCB.
Was gonna say their cheaper PCB, I'm guessing they will use better components for the FTW3.
So you think having a direct stream of GPU/heat airflow blowing directly on the CPU area is the same as being diffused with airflow coming into the case? (Serious question.) I can see several different ways to look at this and how air travels through the case and exactly how case airflow is set up.
Hot air flow density distribution at the cpu heatsink air intake would be different between these two gpu cooling setups especially in a closed PC case. That would be my guess.
Exactly what I was thinking. Put a blowdryer an inch from your hand, then pull it back several inches.....diffusion matters.
I would argue this is even worse for water cooled systems. More likely there is dead air around the CPU and the power components in water cooled systems than air cooled. (Not in mine, but yours.
)
I would argue this is even worse for water cooled systems. More likely there is dead air around the CPU and the power components in water cooled systems than air cooled. (Not in mine, but yours.
Honestly, I expect it to make a small impact on the CPU temperatures compared to current cards because some of the heat is exiting out the back of the card, not into the case. Sure, the right fan in the picture above is exhausting into the CPU fan, but think about current air-cooled card designs for a second... where does all of the heat go? Out the top and bottom of the card, bouncing off the MB and case side panel and into the CPU cooler fan, ultimately, then out the back case fan(s). Yes, these new cards will create more heat because of higher wattage compared to current-gen cards, but again, this extra heat output is lessened by the fact that some of the heat (say, 30%?) is exiting the rear of the GPU card directly and not into the case. So you have 70% of 350W = 245W of heat entering the case with a 3080 vs 100% of 225W = 225W of heat entering the case using a 2080Ti. An extra 30W of heat will increase CPU temps by what.... 1-3 degrees C when using a decent air cooler (NHD-15 class)? If you have a decent airflow case and you aren't near any thermal limits for the CPU, then it shouldn't matter anyway. Obviously if only 10% of the heat is exiting the rear of the GPU card, then a lot more heat will be exhausted into the case and into the CPU cooler's fan, which would definitely impact CPU temps to a higher degree.
So you know that "half" of the heat is dissipated by each fan path? I think that NVIDIA has been lacking transparency here for a reason.
Maybe, maybe not. Reviewers that test in cases is the only way to tell besides having one in hand personally.
No, not at all. A diffused airflow after case heatsoak is actually what I am thinking.
