|
Man, no wonder intel didn't do this earlier. The power usage really is bizarre, especially comparing the 9900k and the 9700k. 100mhz and turning HT on adds FIFTY watts of peak power usage? Like... what? In the end, its performance in stuff like encoding is pretty outrageous. It's beating the 10 core 7900X and 12 core Threadripper 1920X in a lot of those tests which is crazy. But for gaming, there really is no point. I bet most people that buy this will use it for gaming, and it kinda seems like a waste there. Really, the most logical gaming part of the new ones is the the 9600K... the only thing that would make it better would be HT enabled. Huh, wonder if such a thing exists.
|
#
?
Oct 19, 2018 17:09
|
|
|
|
| # ? May 30, 2024 12:54 |
|
|
feeling very good about being a delidded 8700K haver right now
|
|
#
?
Oct 19, 2018 17:09
|
|
|
.
sincx fucked around with this message at 05:55 on Mar 23, 2021 |
|
#
?
Oct 19, 2018 17:40
|
|
|
TheFluff posted:feeling very good about being a delidded 8700K haver right now Feeling like a dope that I didn't jump on the 8086k deal a couple weeks ago
|
|
#
?
Oct 19, 2018 17:40
|
|
|
I haven't felt the need to upgrade my 6600k (running with a braindead OC to 4.5ghz) yet, and I really don't see any reason to even consider the 9th gen. The wait continues.
|
|
#
?
Oct 19, 2018 17:44
|
|
|
I'm probably going to get the 9900k but this does seem like a bad year for hardware, everything is slightly faster and 50% more expensive. I can get a 9900k for 620CAD through work so the intel tax for me is $240 compared to a 2700k with a taichi motherboard. edit: I forgot the 2700x comes with a cooler and I was getting a d15 for the intel and forgot a rebate so the intel tax is more like $380 drat. Perplx fucked around with this message at 18:29 on Oct 19, 2018 |
|
#
?
Oct 19, 2018 18:20
|
|
|
fargom posted:I haven't felt the need to upgrade my 6600k (running with a braindead OC to 4.5ghz) yet, and I really don't see any reason to even consider the 9th gen. The wait continues. Yeah I have a serious case of the upgrade bug and my stupid lizard brain is constantly nagging me to buy a 2700x but I really can't justify it since only 1 game has hit all 100% on my 6600K at 4K60 (AS:Origins), the rest is bottlenecked by the GPU
|
|
#
?
Oct 19, 2018 18:32
|
|
|
so are the reviews out now? seems like from what y'all are saying if you get a 9xxx series chip you better run some crazy watercooling setup or everything will melt. is that accurate?
|
|
#
?
Oct 19, 2018 18:40
|
|
|
axeil posted:so are the reviews out now? seems like from what y'all are saying if you get a 9xxx series chip you better run some crazy watercooling setup or everything will melt. is that accurate? Tom's Hardware used a watercooler. They OC'd to 5.0 at 1.33v, used a -200MHz AVX offset, and prime95 hit a sustained 95C. The thing sucks a ton of power, as it's essentially an 8th gen with two more cores and additional cache. It was using 250 watts at load with these settings. The soldered lid seems all but required for the thing to exist. The 9900k at 5.0 performs almost exactly the same as the 8700k at 4.9 in most gaming tests.
|
|
#
?
Oct 19, 2018 18:44
|
|
|
The i5-9600K is most likely perfectly fine under air cooling. For gaming there is very little reason to buy anything more expensive than that. If you want 16 threads the Ryzen 7 2700 offers an amazing deal at $250. The only reason to buy a 9900K is if you're doing something incredibly specific and weird like streaming a very single thread-bound game like DotA2, using x264 encoding on the CPU, and you're comfortable with blowing several hundred dollars extra instead of using very slightly lower quality encoding settings.
|
|
#
?
Oct 19, 2018 18:55
|
|
|
Once it actually gets here, I'll be pairing a 9900 with an h150i. I'll let you all know what temps are like when I do (I'm rocking an old Corsair 800D with bitFenix Pro fans on the back and bottom intakes).
|
|
#
?
Oct 19, 2018 19:07
|
|
|
eames posted:Well here we go, 10 Celsius temperature difference between solder and liquid metal at 4.8 GHz 1.25V, This video is very interesting. The die thickness is a big head-scratcher.
|
|
#
?
Oct 19, 2018 19:12
|
|
|
Maybe the heatspreader is thinner and so they needed to increase the die thickness? Dunno. Increasing the package thickness is good because they thinned it out on Skylake leading to deformed CPU packages in some cases.
|
|
#
?
Oct 19, 2018 19:30
|
|
|
TheFluff posted:The i5-9600K is most likely perfectly fine under air cooling. For gaming there is very little reason to buy anything more expensive than that. If you want 16 threads the Ryzen 7 2700 offers an amazing deal at $250. The only reason to buy a 9900K is if you're doing something incredibly specific and weird like streaming a very single thread-bound game like DotA2, using x264 encoding on the CPU, and you're comfortable with blowing several hundred dollars extra instead of using very slightly lower quality encoding settings. I play a lot of Paradox games and they seem to slow down horribly with my i5 4690k so I wanted to at least go to an i7, possibly an i9 if it really helps.
|
|
#
?
Oct 19, 2018 19:35
|
|
|
Cygni posted:In the end, its performance in stuff like encoding is pretty outrageous. It's beating the 10 core 7900X and 12 core Threadripper 1920X in a lot of those tests which is crazy. But for gaming, there really is no point. It does seem cool that the 9900k allows turbo boosting of 2X cores making it actually useful in real computers that have background tasks. Maybe next generation will work out well.
|
|
#
?
Oct 19, 2018 19:36
|
|
|
redeyes posted:Maybe the heatspreader is thinner and so they needed to increase the die thickness? Dunno. Increasing the package thickness is good because they thinned it out on Skylake leading to deformed CPU packages in some cases. In the video they sanded down the die and the heatspreader still worked fine, and they got much better temperatures. He speculates that they needed it thicker for the soldering process, but if that's the case, Intel traded a dollar for 75 cents with the switch from paste to solder.
|
|
#
?
Oct 19, 2018 19:41
|
|
|
I'm conflicted - I'm still within the return window for most of what I've already bought. I still have about a week before 30 days is up on the 2x16 kit I ordered, and my 9900K hasn't shipped yet. Even if it did, I ordered from Amazon and I have a 30 day return window where I can just take it back to one of the two Amazon Books stores around DC. Really the only thing I have coming that can't be easily returned is the Evolv X I just kind of impulse ordered, and that I honestly want to keep because it just looks like a really, really nice case. I'm also past my window for the PSU I bought, but honestly I needed a new one of those anyway and got a really good price on it. ~Decisions, decisions~. Though I do wonder if this is why EVGA's taking their sweet time getting the Z390 DARK out - they realize making a top-of-the-line hobbyist overclocker board for the 9900K might be a fool's errand since it doesn't really have much higher to go without using exotic cooling. I honestly wonder if dropping to the 9700K and going with the ASUS Z390 WS Pro isn't the better course of action. 
|
|
#
?
Oct 19, 2018 19:47
|
|
|
axeil posted:I play a lot of Paradox games and they seem to slow down horribly with my i5 4690k so I wanted to at least go to an i7, possibly an i9 if it really helps. I don't think it does. AFAIK most Paradox games (with the possible exception of HoI4) are mainly bottlenecked on singlethread performance. OC your 4690K to 4.5GHz if you can and if you're not satisfied with that get a 9600K and overclock it to 5GHz. Six cores are still more than enough for the foreseeable future. TheFluff fucked around with this message at 19:51 on Oct 19, 2018 |
|
#
?
Oct 19, 2018 19:49
|
|
|
TheFluff posted:I don't think it does. AFAIK most Paradox games (with the possible exception of HoI4) are mainly bottlenecked on singlethread performance. OC your 4690K to 4.5GHz if you can and if you're not satisfied with that get a 9600K and overclock it to 5GHz. Six cores are still more than enough for the foreseeable future. Hm, thanks for this. I'm going to wait and see if in a few months anyone has done any HOI4/Stellaris/CK2 benchmarks between all the chips to make my decision. The i9 does seem like overkill though after reading through Anandtech's review.
|
|
#
?
Oct 19, 2018 19:59
|
|
|
BIG HEADLINE posted:I'm conflicted - I'm still within the return window for most of what I've already bought. I still have about a week before 30 days is up on the 2x16 kit I ordered, and my 9900K hasn't shipped yet. Even if it did, I ordered from Amazon and I have a 30 day return window where I can just take it back to one of the two Amazon Books stores around DC. Really the only thing I have coming that can't be easily returned is the Evolv X I just kind of impulse ordered, and that I honestly want to keep because it just looks like a really, really nice case. I'm also past my window for the PSU I bought, but honestly I needed a new one of those anyway and got a really good price on it. There's always the 8700k, which is in stock in at amazon for MSRP. And there are plenty of of good 370 boards out there that are cheaper than the 390 stuff.
|
|
#
?
Oct 19, 2018 20:04
|
|
|
il serpente cosmico posted:There's always the 8700k, which is in stock in at amazon for MSRP. And there are plenty of of good 370 boards out there that are cheaper than the 390 stuff. I'd stay with a Z390 board. I do think I'm going to drop down to a 9700K, though. The thermals seem way better. Of course, now I have to hope the AMEX points I used on the 9900K through Amazon get refunded, so I can use those in conjunction with a ~$95 gift code. BIG HEADLINE fucked around with this message at 20:12 on Oct 19, 2018 |
|
#
?
Oct 19, 2018 20:06
|
|
|
Seeing all this makes me glad I didn’t wait and just jumped on the 2700X train right away.
|
|
#
?
Oct 19, 2018 20:09
|
|
|
Work/Watt is in line with everything else. Same amount of work in a shorter period of time means higher heat output during that shorter period AVX stress tests as a measure of power consumption are still dumb though.
|
|
#
?
Oct 19, 2018 20:13
|
|
|
Are the thermals on the 9700k much better than the 9900k? Haven’t seen much info between the two so far.
|
|
#
?
Oct 19, 2018 20:15
|
|
|
BIG HEADLINE posted:I'd stay with a Z390 board. I do think I'm going to drop down to a 9700K, though. The thermals seem way better. Has anyone seen what the thermals on a 9900K look like without HT on?
|
|
#
?
Oct 19, 2018 20:15
|
|
|
il serpente cosmico posted:He speculates that they needed it thicker for the soldering process, but if that's the case, Intel traded a dollar for 75 cents with the switch from paste to solder. Intel never wanted to switch, the reviewers and consumers put a lot of pressure on them with the toothpaste memes and all. I’m not going to lie, I also wanted soldered CPUs back but so far it doesn’t look like it was worth it. Zen 2 will be super interesting because Intel the 9900K feels like the 4 GHz P4 all over again. Core count seems maxed (can’t cool 10 cores at 5 GHz with air in that socket), overclocking headroom is approaching zero, 95W TDP with 250W power consumption, etc.
|
|
#
?
Oct 19, 2018 20:18
|
|
|
After sitting on an i7-3770k for the past 6 years I finally decided to get the 9900k because I can definitely feel the CPU chugging along even though I have a 1080 Ti. Until I read the reviews this morning, that is. Guess I'm going with the 9700k instead.
|
|
#
?
Oct 19, 2018 20:25
|
|
|
B-Mac posted:Are the thermals on the 9700k much better than the 9900k? Haven’t seen much info between the two so far. AT has some power charts using POV Ray, which you can extrapolate to thermals:   https://www.anandtech.com/show/13400/intel-9th-gen-core-i9-9900k-i7-9700k-i5-9600k-review/21
|
|
#
?
Oct 19, 2018 20:29
|
|
|
Oh man that power climb
|
|
#
?
Oct 19, 2018 20:35
|
|
|
B-Mac posted:Are the thermals on the 9700k much better than the 9900k? Haven’t seen much info between the two so far. Anandtech showed a full ~70W less under full load, and they got 5.3Ghz out of their 9700K on a Z370 board. The 9900K showed 220W+ and they didn't even bother including a graph as to where its thermals went, they just had a follow-up page where they left-handedly said it's "technically" the fastest gaming chip but also titled the page "For the Gamer Who Wants it All." 
|
|
#
?
Oct 19, 2018 20:36
|
|
|
Hmm maybe I’ll get the 9700k if thermals aren’t complete poo poo compared to the 9900k.
|
|
#
?
Oct 19, 2018 20:41
|
|
|
I swear one of the reviews had a stock turbo table and the 9700k's all core was 100 MHz less than the 9900k. Then add in ~30% for hyperthreading doesn't ~160 going to ~220 seem like right where it should be for a test like that?
|
|
#
?
Oct 19, 2018 20:42
|
|
|
Winks posted:I swear one of the reviews had a stock turbo table and the 9700k's all core was 100 MHz less than the 9900k. Then add in ~30% for hyperthreading doesn't ~160 going to ~220 seem like right where it should be for a test like that?
|
|
#
?
Oct 19, 2018 20:46
|
|
|
Anime Schoolgirl posted:Last time this kind of HT powerclimb nonsense happened was all the way back in the Prescott days The 8700k's power consumption is way higher, double, than the 8600k with the difference of a couple hundred MHz and HT. If anything the jump seems more reasonable for the 9700k->9900k than that one. vvvv I didn't even notice the 8086k up there because I expected it to be next to the 8700k. That's strange. The jump to the 9600 is reasonable. What's up with their 8700k though? Winks fucked around with this message at 06:07 on Oct 20, 2018 |
|
#
?
Oct 19, 2018 20:51
|
|
|
Winks posted:The 8700k's power consumption is way higher, double, than the 8600k with the difference of a couple hundred MHz and HT. If anything the jump seems more reasonable for the 9700k->9900k than that one. 8600k to 8086k is 25% bump in power usage on that AT chart, with a 200mhz all core turbo difference 9700k to 9900k is a 28% bump, with a 100mhz difference (as an aside, the bump from 9600k to 9700k is a wild 49% for only 25% more cores. but there is a 300mhz clock difference as well)
|
|
#
?
Oct 19, 2018 21:06
|
|
|
8086ks are binned 8700s.
|
|
#
?
Oct 19, 2018 21:11
|
|
|
various posters posted:benchmark stuff Now let’s compare some 1440p CPU benchmarks using a RTX2080ti... oh. 
|
|
#
?
Oct 19, 2018 21:22
|
|
|
eames posted:Now let’s compare some 1440p CPU benchmarks using a RTX2080ti... oh. Guru3d and that german ComputerBase site have some tests with a 2080ti e: oh unless you are saying there is no difference up there, cause yeahhhh  ... guru3d only did 3 games in their testing but they were all GPU limited at 1440p even with a 2080ti
Cygni fucked around with this message at 21:32 on Oct 19, 2018 |
|
#
?
Oct 19, 2018 21:27
|
|
|
The clincher for me is that it seems the 9900K *cannot* be used reliably with high-end air. Period, full stop. It supposedly auto-throttles to 4Ghz whenever it hits its thermal limit, unless you remove that limit in BIOS, at which point if it goes too high it simply crashes, which I'd have to worry about too much even with a Noctua D15. Moving to a 9700K knocks $100 off the top, which I can then roll into either a better motherboard or feel better about a 2080 buy. And yes, the 2700X option is there, but sorry, even though the numbers are impressive, I like monolithic dies and Infinity Fabric reminds me way too much of HyperTransport from the bad old days and gives me a "duct tape and baling wire" feeling. Intel might be eating poo poo right now, but the 9700K seems like a worthy successor to my 2500K.
|
|
#
?
Oct 19, 2018 21:29
|
|
|
|
| # ? May 30, 2024 12:54 |
|
|
Cygni posted:AT has some power charts using POV Ray, which you can extrapolate to thermals: This is still about equal to the work gain from HT, even in ancient povray which has rubbish scaling. 25%ish watts for 25%ish ops. E: I am bad at charts, 50% ish watts is of course less fun.  To cherry pick the render test with the best scaling,  Of course, Corona possibly has a larger power draw diff precisely because HT is doing more relative work. Running amperes through transistors isn't free. Conversely, if you're running a game that barely uses HT the power draw will be closer. Xerophyte fucked around with this message at 21:34 on Oct 19, 2018 |
|
#
?
Oct 19, 2018 21:31
|
|
