Change power limits from 11400 to 0 on Gigabyte B560 DS3H
Change power limits from 11400 to 0 on Gigabyte B560 DS3H
Thank you for the feedback. I was considering the m-atx model but ended up exploring the MSI option. I plan to use a Bluetooth dongle for my headphones since I’ll be near the desktop, ensuring good connectivity.
ah okay, you said b660 ds3 in the thread so that's the board i looked at I do find it funny that people get really hung up on VRMs. Sometimes it's warranted, but often it isn't. Most new boards use DR MOS nowadays, and have fairly competent power ratings. But I've gone with much weaker than often recommended...I picked up a Gigabyte Z370 Aorus Gaming 5 which I thought was a 10-phase VRM because of the appearance of the chokes...but it is actually a 4 + 2 phase with doubled components (not actual doublers). So the CPU really was only a 4 phase. I have had an 8700k overclocked to 5ghz at 1.38v since 2017 and it's been fine, VRM temperatures no higher than 77 during full tilt all-core load.... Others have reported 9900k at 5ghz was fine too. On a 4 phase.
The B660 board stands out as a powerful option, especially with its 50A DrMOS-es. It supports high speeds like 12600k, 12700k, and 12700k with ease. Most likely it will be the first one reviewed by GamersNexus due to these specs. Other B660 and nearly all H610 boards (except a few budget models) can handle 12400 without throttling, as demonstrated in the linked videos. Even without Dr.MOS-es, the MicroATX B660M DS3H works well at 12400, along with the Asus Prime B660M-A and Asrock B660M PRO RS. The differences between these models are outlined here: https://geizhals.eu/?cmp=2661683&cmp=2660678&active=1
You're finding it hard to understand the differences between the boards. The videos seem to highlight performance variations, especially between the DS3H and B660 chipsets, even though they tested similar models. I thought the power consumption would be similar just because they’re the same board with different chipsets. For my build, I’m using an i5 12400 and expected it to work well since it uses less power than the higher-end CPUs. However, the unboxed content suggests the i5 might still face some performance issues. This is my first build, so I’m not sure how much the power delivery components affect results. I plan to use the PC mainly for video editing, gaming, and programming, maybe streaming occasionally.
It's worth noting that I recently purchased some components for my BIL, such as the ASUS PRIME B660-Plus D4 and an i5-12600K. We chose the i5-12600K instead of the 12400 because of the E-cores—it was a preference. Considering Alder Lake offers better efficiency than earlier models, I believe an 8-core configuration should easily manage performance at both normal and maximum speeds. What are your thoughts?
There's some merit as I said, but it depends. Most of it is reliant on the motherboard stock settings, and then bad pairings. The B560M DS3H without power limits handled the blender test fine on the 11400F at 4.2ghz and only 78c, which is more than acceptable during a full load test. It even handled the i7 when adhering to the old PL2 spec, and while it DID throttle during a 1 hour blender test, it still managed almost 1ghz more frequency (4.1ghz vs 3.2ghz). Additionally, in most use cases, you aren't going to see the same load applied as a 1 hour render test. In games, the i7 likely won't throttle. And B660 has about double the VRM solution as that board. B560M DS3H: B660 DS3h:
The i5 performed well on the B560, which is a low-end chipset, and might even handle an i7 in gaming scenarios. I wouldn't suggest it, but for argument's sake, the B560 was definitely suitable for a locked i5 without power constraints, and the B660M offers a bit more performance, especially the i5-12400, which is significantly more energy-efficient than the i5-11400.
I believe I understand your perspective. Likely I won’t push anything that would negatively impact stress testing these benchmarks. The peak performance I expect from the chips should align well with gaming and streaming demands. My main worry was whether the CPU could maintain stable boost clocks without jitter. Based on what I’ve seen in this discussion, it seems unlikely to be a problem with the B660M board, especially considering the 12400’s efficiency and lower power consumption.
I’d definitely invest the extra $20-30 in a higher-quality motherboard, since it offers improved power delivery and additional features such as better IO. I wouldn’t hesitate to go for the B660M model, but I’d likely spend a bit more just to ensure reliability. Most of my feedback has been theoretical rather than practical.