If they don’t publish the MSRP then what ever price you find it at is the RP.Let me know when the '1060' variant of Hopper actually hits retail availability at MSRP.
If they don’t publish the MSRP then what ever price you find it at is the RP.Let me know when the '1060' variant of Hopper actually hits retail availability at MSRP.
If they don’t publish the MSRP then what ever price you find it at is the RP.
This is actually Kyle's prediction, correct? GPU MSRP going away assuming there is not a drastic change in market conditions?
600+ watts is really hard to cool on-air alone, I am guessing all the high-end 550w + cards are going to come with some form of liquid AiO coolers. Otherwise, you are going to need some rather unpleasant fans to keep those temperatures down.
Hell, my space heater tripped the circuit at my old office when the heat broke in the middle of winter a few years ago.The 5000 series will require a 30amp run to your office. /s
I have a Dell T640 in my rack running some accelerator cards that are running in the 500w range. That system causes the AC to kick up a notch on full load and you can hear it on the other side of the building sounds like a small jet attempting liftoff. Datacenter’s require liquid on all accelerators over 600w for a reason, it’s just not feasible to do them on air.What kind of case airflow will it take to run one of these new cards if they really are close to 600W?
Everybody will be reviewing the gamers nexus case reviews to find the closest thing to a wind tunnel that exists.
600+ watts is really hard to cool on-air alone, I am guessing all the high-end 550w + cards are going to come with some form of liquid AiO coolers. Otherwise, you are going to need some rather unpleasant fans to keep those temperatures down.
I may be dreaming, but I am hoping for something in the 250w range that performs around the 3080ti, given that the mobile version of it manages to pull around 150w while only being ~30% slower than its desktop counterpart leads me to believe it is possible, the node shrink alone could spread that gap
But because of space limitations and how work is shaping up, I may just ditch the desktop altogether and look for a decent laptop with a 4070 class GPU in there.
The accelerator cards are all passively cooled running strictly off the chassis case fans which move a lot of air but are not quiet about it. But that case is running between 1800-2200w on the accelerators at full load, it's quite the space heater.I can cool my 600W 3090 FE on the retail cooler but it reaches 80C at max fan speed.
and that's with a thermal pad rework and liquid metal with an 0.2mm indium pad sandwiched on both sides between the LM (I think this gives temps somewhere close to in-between pure LM and thermal paste)
Going to try direct LM later without the pad as i suspect the pad is slightly reducing memory pad contact as my VRAM temps went up by 10C compared to regular thermal paste
The accelerator cards are all passively cooled running strictly off the chassis case fans which move a lot of air but are not quiet about it. But that case is running between 1800-2200w on the accelerators at full load, it's quite the space heater.
But the 3090FE is a 360w peak stock pushing into 450w depending on the OC.
3090 FTW3 had a 500w+ bios600+ watts is really hard to cool on-air alone, I am guessing all the high-end 550w + cards are going to come with some form of liquid AiO coolers. Otherwise, you are going to need some rather unpleasant fans to keep those temperatures down.
I may be dreaming, but I am hoping for something in the 250w range that performs around the 3080ti, given that the mobile version of it manages to pull around 150w while only being ~30% slower than its desktop counterpart leads me to believe it is possible, the node shrink alone could spread that gap
But because of space limitations and how work is shaping up, I may just ditch the desktop altogether and look for a decent laptop with a 4070 class GPU in there.
The coolers on those things are huge though, triple slot coolers are already annoying, quad and extra wide aren’t something I want to be the normal…3090 FTW3 had a 500w+ bios
3090Ti FTW3 already hits close to 500W (450w stock) + 6% boost available.
These rumors are always funny. NV and AMD will not lose an opportunity by naming a card which out performs their previous generation by 30% a mainstream card... They will just give it an enthusiast name eg: 4080, slap a higher price tag on it and call it a day. These companies are out to make money![]()
At what point do they switch to AIO cooler for some of these...which is a whole different realm of "not mainstream" as far as graphics cards are concerned anyway.The coolers on those things are huge though, triple slot coolers are already annoying, quad and extra wide aren’t something I want to be the normal…
Honestly I’d say for anything with power draw greater than 450w, and next gen is as good a time to start as any. It could tie in pretty well with features on the new PSU connectors.At what point do they switch to AIO cooler for some of these...which is a whole different realm of "not mainstream" as far as graphics cards are concerned anyway.
Even then, will need a large rad. Aint no simple 120mm for this kinda power draw lol.Honestly I’d say for anything with power draw greater than 450w, and next gen is as good a time to start as any. It could tie in pretty well with features on the new PSU connectors.
No, it's gonna be a triple, but you only need maybe a double for a CPU if you go that route as well so we are looking at a future where cases are going to need the ability to fit a pair of separate AiO at least 1 being a 360 and the other being a 240, that or they may just start putting out more offerings that are pre-fit with water blocks.Even then, will need a large rad. Aint no simple 120mm for this kinda power draw lol.
Some people don't want massive hulking blocks hanging off their motherboard. Thank God for watercooling keeping things sensibly sized.Genuinely curious, why? Unless you have other pcie components. Most people never use more than 1 physical pcie slot (for the gpu) on their board anyways. I modded my 3080 and it's basically a 4 slot card now. The silent operation is worth the footprint. Hell, I think a tiny little 2 slot card by itself on an ATX board almost looks goofy. Too each their own though.
For work, and hobby reasons I have a fair number of addon cards in my towers, GPU for sure, but I also have an FPGA, sound card, dual SFP+, and PCIe Storage.Genuinely curious, why? Unless you have other pcie components. Most people never use more than 1 physical pcie slot (for the gpu) on their board anyways. I modded my 3080 and it's basically a 4 slot card now. The silent operation is worth the footprint. Hell, I think a tiny little 2 slot card by itself on an ATX board almost looks goofy. Too each their own though.
I am hoping mine doesn't die at the 5 year mark.I will keep holding onto my 1080TI.
Do you have the EVGA 5 year warranty?I am hoping mine doesn't die at the 5 year mark.
Nope.Do you have the EVGA 5 year warranty?
but then the 3090ti came out with a msrp after that prediction
most gamers these days are just popping in a gpu and calling it a day.
Nvidia never announced an MSRP for the 3080 12GB.The 3050 and 3080 12gb has well I think?
A yes, was looking at the wikipedia entry page that had a valu for the launch price without validating that it was really the case.Nvidia never announced an MSRP for the 3080 12GB.
For sure, it's getting crazy.Understandable, that's what I figured. You're definitely in the minority though, most gamers these days are just popping in a gpu and calling it a day. Regardless, with the direction mfg's are headed of just brute forcing performance with raw power, the high end cards are still likely going to be these monstrous 3+ slot designs. Hell, some AIB's are shipping with their own anti-sag supports/brackets already.
Which would make a simple tower heat pipe system quite effective (noise and performance wise), I would think with how big the gpu chips is to cool.
I feel the percentage of people that would go with a well made large like that, but silent and superb performance GPU cooling without liquid would not be zero.
Considering what tend to make the most heat in a system, the current gaming pc design is really far from optimal, even bordering on ridiculous.
I will keep my 1070....I just hope I can get a laptop with better battery life, but that doesn't seem likely with the next 4000 series. Give me high performance with DLSS on low power and I will pay handsomely!I will keep holding onto my 1080TI.
I will keep my 1070....I just hope I can get a laptop with better battery life, but that doesn't seem likely with the next 4000 series. Give me high performance with DLSS on low power and I will pay handsomely!
I have an Alienware with a 2070 and an Intel i9 and the battery life is ok...but nowhere in the ballpark of an M1. That's the only reason I'm considering selling it as I'm on the go way more than I thought I would be.I had a Razer Blade with a 2070 MaxQ and a 9th gen i7. Great battery life. Amazing build quality.
I’m not sure it ever was just an Ampere refresh, it takes years to design a GPU obviously they make iterative changes as they go. Especially if you are completely changing up manufacturing and processing. But seeing a few leaks and tossing your designs wouldn’t be feasible.