![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
sappyse107
Limp Gawd
- Joined
- Apr 21, 2000
- Messages
- 510
It has been a good long while since I really dove into the details on chipsets, or even cared about a CPU beyond its frequency for my stupid engine software. Thanks AMD.... So I have had a B350-F and a 2600X for a year now, and its been solid. The 3200mhz CAS14 RAM ran at its advertised spec, though at 1.4v instead of 1.35v, but I am also running the DRAM Calc spec for fast. The B450-F is supposed to be a little better on RAM but the 2nd onboard m.2 is what excited me the most. What I forgot about was the difference in reported temps on the CPU.
The same Trident Z RAM now does 3400 mhz at the fast calculation. Using Performance Test V10.0, the latency went from 51 to 36. Overall score of 2445 went to 3226. Memory Read Cached and Available were the 2 that were basically identical.
For the CPU, I relied on HWInfo64 and Folding@Home to provide the comparison I cared about, and this is where the reported temps bite me in the ass. Its a bit of info so ill try to keep it simple. H100x watercooling and a carbon thermal pad used.
B350-F. 65c usually, T/die would show 74-75 peak. Reported CPU temp was usually 58 or so. I couldn't undervolt at all without some rare system instability, so I kept that at stock. All core multiplier of 41.3 max with the -.0125 volt. 40.8 max with stock volt.
B450-F. 65c if I turn off core boost and PBO. Reported and T/Die are basically the same on this chipset. 39.0 all core. 70c if I turn on the same PBO spec as the B350, Multiplier 39.5. -.0250 volt. 73+c if I turn every boost option on, still around 39.5. -.0250 volt and I see 39.8-40.0. So far its solid/stable but its only been 2 days
Comparing actual voltage though, I think its good. It won't boot with 0.0375v. The B350-F never saw the radiator fans go over 60%. I had to change the curve for the B450-F so it will stay 50% to 70, then 60% to 75 before it goes full blast. It never went full blast, and I don't want the dB unless its pushing the CPU with the F@H.
If I missed the AURA lighting on the board, I would swap the 350-F back in. The B450-F just lights up the rear shroud that is covered by the GPU radiator.
I wish I had a Praxis wet bench. Looking forward to the B550-F
The same Trident Z RAM now does 3400 mhz at the fast calculation. Using Performance Test V10.0, the latency went from 51 to 36. Overall score of 2445 went to 3226. Memory Read Cached and Available were the 2 that were basically identical.
For the CPU, I relied on HWInfo64 and Folding@Home to provide the comparison I cared about, and this is where the reported temps bite me in the ass. Its a bit of info so ill try to keep it simple. H100x watercooling and a carbon thermal pad used.
B350-F. 65c usually, T/die would show 74-75 peak. Reported CPU temp was usually 58 or so. I couldn't undervolt at all without some rare system instability, so I kept that at stock. All core multiplier of 41.3 max with the -.0125 volt. 40.8 max with stock volt.
B450-F. 65c if I turn off core boost and PBO. Reported and T/Die are basically the same on this chipset. 39.0 all core. 70c if I turn on the same PBO spec as the B350, Multiplier 39.5. -.0250 volt. 73+c if I turn every boost option on, still around 39.5. -.0250 volt and I see 39.8-40.0. So far its solid/stable but its only been 2 days
Comparing actual voltage though, I think its good. It won't boot with 0.0375v. The B350-F never saw the radiator fans go over 60%. I had to change the curve for the B450-F so it will stay 50% to 70, then 60% to 75 before it goes full blast. It never went full blast, and I don't want the dB unless its pushing the CPU with the F@H.If I missed the AURA lighting on the board, I would swap the 350-F back in. The B450-F just lights up the rear shroud that is covered by the GPU radiator.
I wish I had a Praxis wet bench. Looking forward to the B550-F
