TerraRaptor

Sweet chip)

Superb!

Great score, awesome!

1.92.2 x32 works fine for me with nf7-s, a7n8x-e, an35n. Most probably it is not socket issue alone but a combo of OS and socket. PS. Great attitude, running test for 80 minutes for team cup means a lot about your dedication to your team.

Always love to see FX being benched that high frequency

SSDs i use for benchmarking are 60GB and I won't be able to install any modern game (>100GB size) to bench:) PS. There is also a problem with game versions - usually bug/performance fixes during the lifespan of the game affect performance a lot. It is completely different to 3DMark benchmarks in that aspect. Take EVE online as an example - it is still EVE after 17 years yet very different game if you compare 2003 vs 2020 (or 2022?).

Given cooling used, this is very strong score

С возвращением к активности)

Is it infineon be-6? Pretty tight tRAS then, mine won't do that.

Yep, digitalbath sent me that link some time ago, thanks guys. The reason I questioned it was vtt level mismatch (50% default) used here. Will test that mod too soon (though smd soldering was always a challenge).

Just curious what vtt is? Is it dram vtt?

Yeah, great SS cpu.

How much did it gain from air to LN2?

Pretty strong cpu, good score!

1% is still too much. Do you? I'm quite sure you won't catch me if I start cheating. I'm nowhere close to be hardcore gamer yet I was playing Asphalt8 for a couple of years - rankings in one of the most popular games were full of cheaters. Even online events had lots of those. I think that most AAA titles with online functionality (so that they offer human vs human competition) also implement quite serious protection against cheating - auto aim, trainers etc and still they have a lot of such issues. And those protections also make things more complex for usual gamers, so why hwbot shouldn't implement the same protection for its 15y heritage?

Simplicity in general is also simplicity to cheat and take high spots. HWBOT was and is database of results in different benchmarks. Making simple validation will ruin the validity of this database - people are people and cheaters will exploit that simplicity to gain top rankings. Look at Aquamark - first only screenshot of am3 window was required, then screenshot in am3 pasted into paint plus cpuz/gpuz, then wrapper - just because there were too many cheaters using simplicity of early validation rules.

I also thought OC is out of interest for most young folks in my country until I recently joined local HW channel in Telegram - people do overclocking with their setups, test 'em in games and discuss interesting modding projects. The thing is HWBOT is not interesting for them. Interest in OC doesn't mean participation in hwbot. Casual overclocking/benchmarking is alive, hwbot should just find the way to invlove these people into community.

Great score!

Was it just ucode update? I tried to do it myself with no success - either wrong version of MMTool or ucodes.

Checked with Ground, it was bios from ObscureParadox not working correctly in pifast for older cpu.

Not sure if it was you or Ground to share L3014-aware bios with me but it was insta-fail in pifast (and only pifast) with stock cpu (cel 430?), those custom bioses are dangerous sometimes - I reinstalled system twice, downloaded pifast 20 timed and thought board is getting semi-dead)

You had to use pre-45nm bios to cut 150-200ms

Great score TAGG!

Типа "бумажный листок" в твоем написании, ну или книга в экселе (вкладки внизу)

Умеете, могёте!