I.nfraR.ed

The total member count for my team is wrong. I have checked other teams, e.g. Overclock.net and their elite members count seems to be wrong as well, possibly inflating the total count, just as in my team. We don't really have an elite member either, the one that is currently listed as elite is a result of an old submission with "LHe" picked by mistake as the cooling method. I'm not aware of any other external elite members joining our team and then leaving. Link to team page: https://hwbot.org/team/xtreme_oc_team_bulgaria__/

I could eventually beat it, but was running out of LN2. Didn't even check what is the top score at that time, haha. I'm sure we'll see better scores.

5 sessions, 2 of which full pot with bad mount and horrible frequencies and voltage tolerance; one good mount at cb/cbb -182; one full pot with kryosheet, but a bit worse frequencies and the last one finally a good paste mount and no cb, cbb still at -182. Could eventually get better scores, especially in wprime where my XP failed miserably for some reason, but I'm not going to bother again. Wprime should have been 4.020 - 4.030 at that frequency, if my XP wasn't scoring 6.4xx.

The scary/crazy thing is it makes even XP kind of obsolete for some other older platforms (not sure if all that can run Win10)... I can easily beat scores people obviously spent ton of time tuning timings and running wazza.

As the title describes, in my opinion it would be convenient to have a filter to list all motherboards for a specific socket or chipet. Currently, you can only browse by manufacturer and then drill down to the socket and then chipset for that socket, so e.g. 790FX chipset is now the third inclusive filter: motherboard vendor -> socket -> chipset. In addition to that, and I guess related to the recent change in the db schema, if you open e.g. DFI LANPARTY DK 790FXB-M2RSH (https://hwbot.org/hardware/motherboard/lanparty_dk_790fxb_m2rsh/) and you click on the additional links on top, it will only show you the DFI motherboards that meet the search criterion. Thanks!

Thanks. I didn't spend much time on this as I really hoped for a better clocking chip.

Thanks. LN2 no need

21H2

It's also possible with this one, but a 150 at a similar FSB in unganged mode would be an "easy" gold, yes. I was expecting to hit a bit over 6.1GHz which would have been enough for the current #1, but ran into different issues, so that's what I have for the moment. Would need another session to max it out and I also want to run 32M on XP where it does a bit higher FSB in unganged.

Excellent. Well done!

I'm counting on you, ale! I don't have a more suitable board right now.

I reckon higher is possible, just wasted too much LN2 fighting with the other board. If someone manages to run high FSB with Unganged mode it should be possible to beat it with a lower clock. A more efficient board could also make a change as this one is pretty slow in everything. Core-wise that's definitely not the limit of the CPU, the board is holding me back. I might try brute-forcing it with DDR2 board.

For geekbench3 it seems to be, according to the rules.

Thanks. This board is very inefficient, but my Giga ud3 didn't work well on cold Hopefully I will have time to try again next month.

Why 2T though?

Killing it :) Great work!

Oh, wow. Great CPU, congrats!

I was using 130MHz PCIe clock on the DFI with that Sempron 145 some years ago. Not too confident it actually did anything, but the card can handle it no problem.

Matched the model number, nice!

What do you mean? PSU is a modern one, haven't hooked my 3.3V rail-modded one yet. VIO is 3.8 for these submissions.

I have added it in version 1.31. All BIOSes have this voltage, it was just not displayed in older ZenTimings versions.

I only have 3. Ordered last year and received recently. This one seemed to POST a little higher than the other 2, but I haven't tested them yet, apart from that.

It is really hard to understand what you are asking. It is not needed to have equal VDDIO, VDDQ and VDD. They also depend on CPU and memory sticks. For VDDIO people usually stay in the 1.30 - 1.40V range. VDDQ and VDD are not displayed correctly in ZenTimings when high voltage mode is enabled.

That is a great CPU, indeed!

Yes, the 0516s and 0530s are all good cold, if I'm not wrong, need to test them properly some time in the future. I've only tested single benchmarks for competitions. the FX53 is untested, but that's all I have - just one.