MSI team overclocker TSAIK has successfully clocked AMD's Ryzen 7 5800X3D CPU above the critical 5 GHz mark. Interesting for a PU with a locked miltiplier.... Overclocker Exceeds 5 GHz on AMD's Ryzen 7 5800X3D
Now I really wish AMD hadn't locked the multiplier on it, even if the voltage is locked. There's already conspiracy theories going around that AMD did it so it wouldn't be close in performance to Zen 4. There's no way for a motherboard without an external clock generator to get a base clock overclock, is there?
It looks a bit like the extra cache sabotages the heat transfer somewhat, because there is a extra layer on top of the chip, so it acts a bit like the Intel delid years. Stock performance is fine, but extra heat from overclocking is hitting the chip harder, then it would a 5800x, because the heat does not flow as easily away from the hot spots.
It would have been preferable if AMD had locked voltage, EDC, PPT and TDC. Instead of the multiplier. This would allow a decent control on the chips temperatures, while allowing a decent margin for OC.
If this was a 5800x and could do 5.15 on 1.2v i would not consider it even a golden sample.. Or platinum or diamond sample ... Maybe mythril sample
Proccess node has improved a lot; newer 5800X are more likely to reach those clocks compared to the ones released a year ago
My 5800x does 4.6 all core on 1.185 and supposedly is a golden sample 5.15 on seems a bit too much of an improvement . On the other hand I hope you are right I am not against people getting such good silicon I am not petty like that ,.seeing that this is an ocer that works for MSI I suspect other than the 5800x3d being higher binned on top of that it is likely he was given the best MSI had their hands on ! That said if an average 5800x3d can hit 4.9-5.0 on 1.2 that would be AWESOME !
As far as I know the 5150 is CPU-Z validation only, no load at all. You can probably boot and validate 5000 at 1.2v too, but it will crash at load.
I wonder why the PCIe speed issue is never mentioned. If BCLK Ocing was a thing we'd all be doing it.
The article somewhat implied it: That doesn't mention PCIe specifically, but that's because it's not just PCIe that becomes unstable.
Well yeah HH noted it but the overlocker doesn't mention that. it can't really be called anything more than a dog and pony show.
True, but I think the point was to show that these chips had overclocking headroom, whereas AMD claimed otherwise.
I liken it to trying to jam a 300HP nitrous kit into a Toyota Prius but I suppose you could assert that the chips might have the room. The bigger question is does the 3d cache have that room.
Oh ok that makes much more sense! I really assumed it can complete at least one cinebench loop ;p that's why I said this is mythril sample ;p
If this chip could run 24/7 @5150mhz then i think it would be close to Zen4 (gaming wise). I don't believe that's why they locked it down though. I think the cache becomes unstable with an OC so that's why they had to lock it down. I wonder if this means all Zen4 chips will be locked down too, still this is only Gaming scores as i believe 5800X3D even with an OC would only beat 5800X in all other benchmarks by a small amount as we saw in HH's review.
I bet it's under LN2, not because of heat output but because extreme cold can get you stable at lower voltage.