CNL-U might be dead and I wouldn't be too surprised if it is, but there is strong evidence (that Richtek voltage regulation document) that it isn't.
I dunno, I think it's more likely Coffee. If that's the case, I have to wonder if that means for sure that Cannonlake-U is dead and only Y will come out.
Well, as if the die failing to be cooled wasn't enough, apparently all of the motherboards have absolutely awful VRM cooling:
https://www.youtube.com/watch?v=f7BqAjC4ZCc
It isn't even easy to attribute this to cost savings - this can only be explained by being rushed to market.
Wow. I can't believe that. Its not just the VRM's overheating. He clearly mentioned a far more fatal issue which is that many boards feature an 8 pin CPU power connector and if OCing that power hungry 10 core (or maybe even the 8 core?) too much current gets pulled through that 8 pin cable and it could get hot enough to CATCH ON FIRE.
This release is dead. Intel truly borked this one. Even if I was ready to buy an X299 platform, after seeing this I'd be scared to death and would certainly wait at least 2 months. Seriously guys, this is actually dangerous. If people leave an intense workload running overnight and they don't pay attention, they could burn their damn house down with one of these things.
Yeah because 1.25A above spec on the PCI-E for a GPU, which has two sources of power, is the same as a CPU trying to pull 300W from a single 8-pin.If RX480 did not burn anyone home by pulling way more out of the PCI-E slot than the ATX 24 pin connector can handle, you can be petty sure an 8-pin CPU power is not going to catch on fire.
Thats said, OC SKL-X is really pushing the EPS-12V specs.
Well, as if the die failing to be cooled wasn't enough, apparently all of the motherboards have absolutely awful VRM cooling:
https://www.youtube.com/watch?v=f7BqAjC4ZCc
It isn't even easy to attribute this to cost savings - this can only be explained by being rushed to market.
RX480 wasn't pulling 250-300w on a connector designed for 150w. That's going to get worse, much worse on the 12-18c parts.
The concern is real this time, and the fire hazard is there.
Those temps!Well, as if the die failing to be cooled wasn't enough, apparently all of the motherboards have absolutely awful VRM cooling:
https://www.youtube.com/watch?v=f7BqAjC4ZCc
It isn't even easy to attribute this to cost savings - this can only be explained by being rushed to market.
Thanks. Why doesnt Intel use at least that instead of regular non-conductive TIM, if they are not going to solder? Price, i guess?
The polymer TIM Intel uses might not have the best thermal conductivity, but it will outlast most other TIMs aside from indium solder. Anything an enthusiast puts in there usually needs to be redone periodically.
The one where you implied (if not outright stated) that server room admins care nothing about power usage and heat output.
So, it would seem that despite Intel's claims, most of the existing boards will not support higher than 10 core count CPU's as they are currently being pushed to the point of spontaneous combustion with only 10 cores. If the clocks are to be acceptable on the 12+ core parts, new boards will have to be released with a much higher power spec and none will have a single 8 pin connector. None of these boards could possibly withstand 16 cores at 4ghz.
That's what I was thinking. There is no way the current X299 boards are going to run the higher core count Skylake-X CPUs reliably. What is going on at Intel?
This made me curious, so made a quick attempt to find out the info with decent source material.The 8-pin EPS-12V is not 150W, is way over 200, i dont remember exactly the power rating now.
The guy is probably just mistaken or there's something we don't know.
Uh, they will in spec.There is no way the current X299 boards are going to run the higher core count Skylake-X CPUs reliably.
Possible, but as was said, they are still fine for running stock, just hilariously bad for any OC.Like intel changing clocks and hence power specs at last minute.
It's not Panic, it's Chaos. We've first seen it with renaming of Skylake refresh as Kaby Lake.I can only describe what I see from Intel as Panic.
That doesnt really explain it though. The boards should obviously be designed to handle overclocking as well.Like intel changing clocks and hence power specs at last minute.
They actually don't care about power usage or heat output. They care about power efficiency. If a chip produces 1 KW but does the work of 10+ 100W chips, they won't care.
However, continuous power ratings need to be lower. The best I could find is this article, which mentions 155W and 235W respectively. (though I think the 8-pin value should be higher)
Don't take the 235W figure as certain, all I was able to find out for sure (documented) is we're talking about 336W max power rating on the 8-pin connector, at max 30C over room temperature.Is the 235W figure @ 30C? If things start getting hot, you'll start to see the effective continuous power rating for the connector go down.