ShadowChaser

Did it to a 7600K and an 8700K, it's pretty darn foolproof. You might be able to get it cheaper from ebay - I got mine for $5. All you gotta do is put the cpu in the slot, line up the ihs with the block, and start cranking. I did practice it on a C2D but turns out even though it wasn't compatible it still worked fine. I've since used it to delid various other intel CPUs to be used as showpieces.

The cheapest 'not complete garbage' motherboard in that bracket this side of the puddle is the AsRock B450(m) Pro4, which regularly retails for less than $70. Truth be told any B450 motherboard with a VRM block will be able to handle a 1600X provided you run it at stock speeds and have some ambient airflow over the heatsink. https://uk.pcpartpicker.com/product/zym323/asrock-b450m-pro4-f-micro-atx-am4-motherboard-b450m-pro4-f

No, but he asked which would be faster and quieter, and all else being equal the chip that draws less power will be the quieter system.

interestingly enough folding at home identifies max-q variants. I don't know how they do it though. You can also take a look at power consumption as the max-q variants typically have a similar power draw to the gpu tier below them. ie a 2060mq would pull about the same power as a 1660 in general

Speak for yourself. In a thermally or power constrained environment, everything counts. No, it's not as bad as a GPU, but GPUs don't have to deal with craptastic IHSs and shitty TIM. 53W on a CPU warrants a higher tier of cooler, plain and simple. I personally care about performance per watt and that's why I actually bother to tune my components. Since you have the ability to brute force your cooling and power it won't matter as much. For me it does.

time to replace that light bulb then, unless it must be so powerful

I'd personally say so, but ymmv, of course. There just seems to be a much smaller effective voltage range for ryzen compared to comparable intel chips

Indeed. Power and thermal management matters in a <4L chassis. Looks like I got something to do this coming weekend, then I was hoping that running a benchmark would give me a good idea of what I could expect compared to a similarlly specced system but since the scale is so weird I guess it doesn't matter

It doesn't underclock better, per se, it just manifests itself as a better underclocker given the IPC and overall efficiency of Zen with its 7nm process. You'd be hard pressed to find a ryzen chip that scales well below 0.8v, I'd imagine, while most intel chips, especially mobile ones, will take 0.6v and scale almost perfectly.

I downclocked both to heck, actually. Saved about 50W of power between the two compared to stock. Then I wondered how that affected my performance in game and such If there was a way for me to measure performance loss that didn't involve me downloading and installing every game I'd play and compare the same situations side by side I would have done it, or I could just be dumb and missed something that does exactly that. It wouldn't surprise me XD also I'm pretty sure 4.0 on a 3600 is a not an overclock, technically speaking, since it can all core boost to 4.2 iirc.

here's just my own experience: My 8700K runs at 5.0Ghz 1.25v under a 120mm AIO, and draws approximately 110W under load My 3600 runs at 4.0Ghz 1.12V under a tiny 80mm SFF cooler, and draws approximately 57W under load. The 8700K is about 10-12% faster in multicore applications while drawing almost 50% more power. If power efficiency is truly a concern I really would just get something Zen 2. If you need something for multicore get the chip with more cores. If you need the maximum single core performance then maybe get the 8700K? Though a Zen 2 cpu will usually be only a few percent slower than the fastest Intel CPUs.

I recently acquired a Ryzen 5 3600 to be used in my Xbox sleeper build, and out of necessity given the tiny cooler I had to undervolt it. Now I'm no stranger to undervolting or over/underclocking with efficiency in mind, but what I can acheive with this chip compared to a comparable Intel CPU still surprises me. I must admit that undervolting Ryzen is more difficult than Intel's recent offerings, having done it to U-series, H-series, and K-series chips for most of the 14nm lineup. I really wish I could set individual core clocks, grr, but that's not the point. This is the point: Cinebench r20 Scores R5 3600: 3520/462 (multi/single) i7-8750H: 2568/404 (multi/single) And the 3600 drew less power than the roughly equivalent i7, averaging 57W PPT compared to 60W+ on the i7 (45 watt my ass) In order for me to achieve a similar score I would need to use my i7-8700K, clocked at 5.0Ghz and drawing double the power. Yes, it does win by a fair margin at 3900ish/500ish but my goodness the performance per watt of even a relatively poor bin of Zen 2 is amazing. At this point I might as well swap my 8700K for a 3700X or maybe a 3900X if I can find one for a good price. They barely lose out in single core and would give me a lot more head room for my other non-gaming hobbies.

I was tuning my new build today to try to get my GPU to consume less power as I'm redlining my PSU in terms of how much wattage I'm pulling from it and it's running really hot. I'm unfamiliar with how the scores in Time Spy and Fire Strike scale, and given that I'm expecting to see some performance loss with a power optimized overclock compared to a performance optimized overclock I'm wondering if any lovely peeps who know more about how these benchmarks work can give me some insight as to how much a score difference matters. My system currently runs 6575 in Time Spy (avg of 3 runs) and 15370 in Fire Strike (avg of 3 runs). I expected these scores to be lower than a comparable system, complete list of specs below, but since I don't know how these scores scale I have no idea if I'm losing a lot or a little. It seems like I should be hitting ~7000 in Time Spy and ~16000 in Fire Strike from a quick look at typical OCed results The build is running: Ryzen 5 3600 - 4.0Ghz all core @ 1.12v EVGA GTX 1660 Ti SC Ultra Gaming - 1950mhz core @ 0.9v, 7000mhz mem @ stock Corsair Vengeance LPX 2x4GB 3200mhz 16-18-18-36 1.34V MSI B450I Gaming Plus AC Motherboard Sabrent Rocket 512GB Windows 10 Home, almost fresh install HD-Plex 200W DC-ATX + HP 230W Brick

Sounds like your CPU isn't running at its max speed. Is it at 3.6 Ghz when you play games? Your temps seem fine but it should definitely be pulling close to its tdp, if not more in a gaming workload.

I'm running a 1070Ti and an i7 on 450W, so you should have no problem. When you get stutters you should check and see if you are running into some sort of power/thermal limit, or just straight up bottlenecking, which is possible given your CPU choice.

I personally got a NCase M1 and the form factor is very similar. The only problem with these smaller cases is that it's usually terrible to build in, but they are great once they're built

it doesn't seem to be available anymore in the US

https://eu.evga.com/products/product.aspx?pn=11G-P4-2489-KR

The hydro copper ftw is a real product, and so is the hybrid.

it would have to match the PCB. I think EVGA has two hydro copper series for the 2080 Ti, one for the FE, XC, or SC, and one for the FTW series. Check what your card pcb is and I'd go from there. Good luck finding anyone with a spare cooler though it is possible since some people convert their cards to run on water.

AMD's naming scheme strikes again! ryzen 3000 is zen 2, ryzen 4000 is zen 3. Zen 3, or ryzen 4000, will not work on B450 but Ryzen 3000, or Zen 2, will run fine on B450. The 3100 is Zen 2 and should be fine on a B450 motherboard however given that B550 is coming soon and should slot into the same price bracket I really would not spring for B450 at the moment

unless it's a weird card the mounting holes are the same and so you'd be fine

just make sure it's a gpu that is supported

it's hard to tell from the pics but it looks like the sata block isn't higher than the pcie slot and so you should not see any interference

the rgb fans that come with the case should go into the RGB hub, which then goes to the motherboard. If the headers are standard then the psu rgb should also be able to interface with the rgb hub. If not then you will have to find a separate hub that can handle both the psu rgb and the fan rgb.