Zen 2 / Ryzen 3000 per-CCX OC

well duh, but almost double the price for a fucking 100MHz bump? Who the hell pays for this?

Just looked it up, all 9900kf make it to 4.8, 92% make it to 4.9, 31% make it to 5.0, and 4% make it to 5.1

literally the top 4% of all chips make it that high, this site functions off of removing the gambling aspect from chips. they have at points had i7's cost nearly 1000$ because they got one that clocked stabley to 5.4ghz

I mean look at this shit. A base price 9900k is what now, $470? $480? I thought the entire thing about the 9900k that intel shills push was "muh OC headroom". Why the fuck are people paying price premiums for 4.8GHz capable 9900k? Unless this is 4.8GHz all core at some stupidly low voltage.

yea I get it. Rarity and they take all the guesses out. But c'mon now.
>that $929 5.1Ghz capable 9900k listed as sold out
Goodness me. More money than sense it appears.

Attached: y-tho.png (817x618, 286K)

Record breaking benchmarks here we go. I wonder how much improvement gaming will get with these OCs.

>who pays for it
People who buy $500 Intel CPU over $200 AMD because muh 10% difference. They will shell out extra $400 if that means they can get an edge over other Intel owners as well.

because professional overclocking is somehow still a thing, and it beats having to buy 10+ chips to not even get one that clocks so high.

Stability at these speeds are subjective. The site probably has a more stringent requirement for stability at a given clock speed, and their ratings are based off a -3 avx offset.
Most/all 9900k can do 5ghz without avx offset if you compromise for stability in edge workloads like blender and small fft p95.

if I had the money and could take guesswork out of buying a cpu, I would likely pay for a near msrp cpu just so i know I don't get a lemon, and pay for them to delid it just so I don't have to risk fucking it up or buy the tools to do it.

it seems smarter to me this way then buying intel thermal jizz, only able to 'oc' to 4.5 because the pads are so fucking bad it heats up like a motherfucker, and then have the pad fail and the thermals are constantly at 100 even when it downclocks, but this is my little brothers experience with intel we are talking about.

I'm also curious about this. If you could get these types of overclocks as a daily use, I would buy Ryzen right now.

I don't know how people consider that "stable". If I have to worry about what sort of workload I'm doing to retain stability, it's not stable. Period.
I suppose. I don't knock people for buying what they want, but I've never seen such blatant idiots throwing money away. Fringe prices for fringe use cases I suppose.

9900k and 9700k are soldered

only 35% can do 5ghz.

Attached: chrome_2019-07-17_21-15-53.png (1052x652, 191K)

little brothers is a haswell I forget what it is exactly, and since they can delid it regardless of solder, that would probably be a good thing to do seeing intel fucked that up as well.

>Stability at these speeds are subjective.

It all depends on what you mean by "idiots." Is it idiocy if I buy a $3 million dollar chiron over a $1000 beater/reliable Honda? Who can say.

yes you can make it to os at 5ghz, congrats, I want my system to not fail and loose shit.

sorry to tell you, but a 100MHz boost for nearly a 100% price premium isn't comparable to a $3 million car vs some beater. Again, if you have money to throw away on stupid shit, go ahead and drop almost a grand on a 9900k.

>t. butthurt ryzenfag
Now you're just strawman. I lost the silicon lottery with my 9900k that does 1.38v LLC6 5ghz.
Even under the most strenuous workload in my daily use (rpcs3, RDR uses 130W+, 90C package temps) it comes nowhere near to crashing.

and I watch intels 'oh, all cpus get to 5ghz, don't worry' bullshit and look at my brothers that cant do more than 4.7 and wonder what people consider stable. I don't consider pre 2000's yours shit crashes all the time useable, and them saying 35% seems to be far more believable when you hear everyone bitch about their cpus not hitting 5

>I don't consider pre 2000's yours shit crashes all the time useable, and them saying 35% seems to be far more believable when you hear everyone bitch about their cpus not hitting 5
We share the same opinion then. It's not surprising that many people lie on the internet. I take it with a grain of salt whenever somebody claims 'my xGhz oc is stable at y volts.'

>Even under the most strenuous workload
>rpcs3, RDR
So nothing actually strenuous. Try rendering or encoding. Your CPU is not considered stable if it's stability is base on workload.

It's stable enough for me but I'm not fooling myself/anybody. If I needed actual stability, I would stick to stock clocks with an undervolt.

>der8auer
isn't his GF an actual whore?

The difference is overclocking is a real hobby rather than collecting garage queens.

Attached: 10380680_720024251374401_6828985503802789635_o.jpg (2048x1367, 307K)

just seems like a waste of time to me. The chance of crashing during whatever it is you're doing due to instability isn't worth the extra 4% performance.

That's a fair opinion. The oc to 5ghz gave +6% more performance at a cost of +50% powerdraw over stock+undervolt so I'm sticking with the 4.7 all core.

here's a reddit guy

Attached: 1535276727789.png (284x379, 33K)

maybe hardcore overclockers who use LN2 . i bet that on LN2 that chip can do over 7 ghz

5% at best .

Holy shit
I think this is how PB2 is actually supposed to function, boosting the best quality CCXs during lightly threaded workloads.

Basically paying for bragging rights, damn those guys with money sure like to waste it on minor things such as emotions.

Status symbol drives 95% of human action for 99% of humans.