Intel Core i9 7900X processor review

Processors 199 Page 1 of 1 Published by

Click here to post a comment for Intel Core i9 7900X processor review on our message forum
https://forums.guru3d.com/data/avatars/m/197/197287.jpg
So, if this thing requires this much power, what is the 18 core going to require..?
data/avatar/default/avatar12.webp
So, if this thing requires this much power, what is the 18 core going to require..?
A Fermi 😀
data/avatar/default/avatar06.webp
Pair it with a Hawaii GOU and heat the whole house.
Dude I can live without my central heating... WOHOOO!!!! All in all the x299 is a Fiasco, the KLX is garbage that is slower than the KL and more expensive, the 7900x is a hot MESS , the boards are full of bios and firmware bugs , the only shinning light is the 7820K.. TR will bury this easily
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
Power consumption should lower with temp due to leakage.
To further clarify - this isn't an Intel problem, this is a physics problem. The hotter the CPU gets, the more voltage it needs to remain stable. But the more voltage you put into it, the hotter it gets. So it's a self-promoting issue. As you push a CPU harder, the vcore will drop, where you need to add even more voltage to compensate. When you've got 10 cores with a crappy IHS, the temperature+voltage problem amplifies itself really quickly.
https://forums.guru3d.com/data/avatars/m/237/237771.jpg
Dude I can live without my central heating... WOHOOO!!!!
It's not global warming, God is an Intel user.
To further clarify - this isn't an Intel problem, this is a physics problem. The hotter the CPU gets, the more voltage it needs to remain stable. But the more voltage you put into it, the hotter it gets. So it's a self-promoting issue. Another thing to consider is the vcore will drop under heavy load, where you need to increase the voltage in order to maintain stability. This then causes the temperature-voltage loop issue I just mentioned, but in addition to that, stability drops as temperatures rise. So, this further feeds the problem. When you've got 10 cores with a crappy IHS, the temperature+voltage+stability problem amplifies itself really quickly.
I understand this I don't think you understand what I was asking. I would like to see a before and after delidding with an identical overclock to see if lowering the temperature brings the power consumption in line.
data/avatar/default/avatar18.webp
it's not global warming, god is an intel user.
lmfao :d
https://forums.guru3d.com/data/avatars/m/80/80129.jpg
Hmm, 10 core has two FMA's per core, 8 and lower only have one. I find that weird not that it effects me. I wish there were more in depth reviews on the 8 core -- sucks that the EU tech reviewers got shafted by Intel, EU sites usually have the best reviews =\
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
I understand this I don't think you understand what I was asking. I would like to see a before and after delidding with an identical overclock to see if lowering the temperature brings the power consumption in line.
I know what you were asking, I was just clarifying for those who may not fully understand the situation; my statement wasn't made out to you specifically. Anyway, I too would like to see results after delidding.
https://forums.guru3d.com/data/avatars/m/236/236670.jpg
That thing is Fast!...Nice review Boss!
https://forums.guru3d.com/data/avatars/m/237/237771.jpg
I know what you were asking, I was just clarifying for those who may not fully understand the situation; my statement wasn't made out to you specifically. Anyway, I too would like to see results after delidding.
Gotcha:thumbup:
data/avatar/default/avatar06.webp
Bad marketing. You don't release a space heater in the middle of summer.
https://forums.guru3d.com/data/avatars/m/237/237771.jpg
Bad marketing. You don't release a space heater in the middle of summer.
Well it was originally set to release right before winter to be fair.
data/avatar/default/avatar35.webp
Well it was originally set to release right before winter to be fair.
True 😀
https://forums.guru3d.com/data/avatars/m/156/156133.jpg
Moderator
I think Loophole won the internet with his post lmao. But damn those temperatures though! You need this in a server room with AC blasting and a custom loop just to keep it cool enough to run! Hopefully one would assume these are highly binned cores, so they should be able to handle high voltage and high temperatures.
data/avatar/default/avatar13.webp
I think Loophole won the internet with his post lmao. But damn those temperatures though! You need this in a server room with AC blasting and a custom loop just to keep it cool enough to run! Hopefully one would assume these are highly binned cores, so they should be able to handle high voltage and high temperatures.
If you are referring to the space heater in the summer, that one was mine 😀
https://forums.guru3d.com/data/avatars/m/231/231931.jpg
Dude I can live without my central heating... WOHOOO!!!!
Eh? Your whole system is a power hog too. Anyways this is a engineering sample. Could change with retail plus on H20 or delid the power consumption will be inline with 6950x likely.
data/avatar/default/avatar30.webp
Eh? Your whole system is a power hog too. Anyways this is a engineering sample. Could change with retail plus on H20 or delid the power consumption will be inline with 6950x likely.
Just check the toms hardware,, there is no way around ir the CORE i9 is a power hungry hot mess
https://forums.guru3d.com/data/avatars/m/80/80129.jpg
Just check the toms hardware,, there is no way around ir the CORE i9 is a power hungry hot mess
FYI AVX2 & FMA are enabled on the Prime95 test for Toms. So it's not an apples to apples comparison against Ryzen. The gaming loop power test they did is more in line with real world results. There is a 35% power gap there, but the 7900X has two additional cores and obviously performs slightly better. So honestly it's not that bad so long as you aren't power testing it with a full AVX load. The 8 core should be fair better in torture tests because it only has one FMA per core. I don't know what cinebench uses though, isn't much information on it.
data/avatar/default/avatar10.webp
With increasing turbo modes/names, I'd like to see some kind of test on turbo availability/sustainability. What I'm trying to say is, test the % of time the cpu can be at a given clock speed. For instance, the i7 on my laptop has a turbo clock for 1 core of 3.8Ghz, but even in winter (15C on room) it quickly hits the power and/or thermal limits. So in the end the 3.8Ghz is more of a PR item than a real thing. Of course this get smoothed on benchmarks since they run for long enough but I bet a LOT of people will think something in the lines of "ooh, 10 cores and a super high turbo for gaming, take my wallet", when in reality, the cpu will probably run at almost base speeds avg while gaming.