I.nfraR.ed

Tried the CH5 with Deneb at same settings as CH3 and 970A-UD3. Confirmed it's 10sec slower. Still waiting for the newer 890FXA-UD5 and 990FXA-UD3.

Yes. didn't work well, waiting for the newer one. Meanwhile, something interesting. Didn't expect Crosshair V would work so well with Thuban. There's room for improvement. 890FXA-UD5 could not run 1100 at all, only booting sometimes. 1140 boots on CHV, but can't get it past initial loop, maybe cold on the CPU would help, might try with the chiller in full throttle. Hypers... 1000 6-7-5 looks fast, but couldn't finish a run, needs more work. 1067 to 1100 7-8-7 in theory should be easy with my GTX2, but on this platform was a no go. On socket 775 they work well at 1000 6-7-5 http://hwbot.org/submission/3991700_ or 1085 7-7-6 on s.1155: http://hwbot.org/submission/2280242_i.nfrar.ed_superpi___32m_core_i5_2500k_5min_51sec_609ms Even 2200 8-8-8 doesn't work, while GTX2 are rated 2250 8-8-8 1.65V. Will wait for both boards to arrive, test them and then decide which one would be the best to use for Thuban. Not gonna bother with LN2 yet.

That's not a solution. I would still want to access some of my older scores for comparison purposes, e.g. a 5G, 4G, competition score which all got bested by e.g. a 6.8GHz score. It had worked this way before. What if a certain score is linked in a thread? You delete it to "fix" a flaw in the code and it's gone - link doesn't work anymore. What happens if it is a past competition score?

970A-UD3 seems a little faster. The CH3 with that bios is much slower at 235MHz FSB and 1:4 divider. Waiting for newer revision 890FXA-UD5 and rev1.1 990FXA-UD3. Meanwhile I might play with 1090T + PSC/BBSE.

OS is the same, but new installation, maybe not set up correctly. Managed to kind of match the old result with CH3, but only on higher HTRef, not sure what's up. 890FXA-UD5 with the same 555 doesn't want to boot mem higher than 1664 (overclocking failed, FSB back to 200), no matter timings, dividers and voltages. Tried all bioses, have to try with another CPU where it was working before. If I drop the mem divider at say 235 FSB, thus lowering the mem freq it boots. This CPU can do 966 6-6-6-18.

With the SuperTalent kit I can run same settings as on the 970A-UD3 and Crosshair III. 890FXA-UD5 is a little faster than the 970A, but nothing dramatic in terms of eff. Mem overclock is a little harder on both CH3 and UD5. No way I can drop off 10 seconds with just a board/bios change. Plus I can't even beat my old 4G score, where I didn't even tried hard: http://hwbot.org/submission/2990939_i.nfrar.ed_superpi___32m_phenom_ii_x4_965_be_16min_16sec_641ms

I think that chip is a 6.9GHz 32M material, unless it doesn't scale with more volts.

A whopping 14 seconds faster than my 5G run. I didn't spend much time on that, perhaps a single run, but still a big difference which don't think I can come up close at all. Wonder what I'm missing. Also interesting that at loop #13 you're already half a second faster compared to the failed run, although frequencies are exactly the same. What bugs me is the real memory vs available real memory and no way to figure out how you get that efficiency. Obviously still a lot to learn PS: Waiting for 2 more boards, the 890FXA-UD5 rev 2.0 only likes one set of SuperT Hypers I have. The first board that didn't like the GTX2... I always thought my eff is decent enough, especially compared to most of the guys at the top of the AM3 rankings. You can destroy the K10 32M if you find a good enough CPU.

Doh, didn't pay attention there's an additional rule in the stage. Not that I care much, but don't see the reasoning either. Maybe for the ES rule, but I've never seen and ES board in AMD submissions or a board that never went retail like some of the GB LN2 boards for Intel. Anyway...

I can't even touch my old scores, something must be wrong with the OS. Crosshair III is faster than 970A-UD3, but not much. Found 890FXA-UD5, hopefully not DOA, will see tomorrow. Does anybody know what is the difference between board revisions? rev2.x is rated for 1866+ DDR3, while 3.x is 2000+, some color difference and white vs black socket. Apart from that I can't see a difference.

Mine are still fine, but they've rarely seen 2+ V. For AM3 it's not really needed - can handle tight timings at much lower voltage. One of the sticks have "died" 3 times. Solved it by re-flashing the SPD, but originally I thought it's dead. One day a gigabyte board booted with both modules installed (only 2GB detected), but the "dead" stick was visible in windows with no SPD shown in cpuz and flashing tools. Re-flashed it from the good stick and it worked. Now I know how to resurrect it. Wonder if there are people who experienced the same and trashed the supposedly dead sticks... PS: Mind you most boards refused to boot with the broken SPD. But the 970A-UD3 didn't care :D. Must restist . I still have your 4x2GB ST 2200C8. Then another 2200C8, 3x2GB OCZ Blade 2000C7, some mediocre HyperX 2000C8.

Several people asked about mine through the years, but I will never sell them (3 sticks) ?

Wow, you beat me by whole 6 seconds. I should really play with my older boards.

450 HTT is 900 mem and 6.5GHz with Sempron 150 (I have a promising one), so there's still a headroom for memory frequency. Based on my testing on air, if it scales ok, then ~170cb might be possible. Don't know about Geekbench, but mem should definitely help. Not sure if the IMC would be a bottleneck, though. Have to test. Looking forward to the new scores as well, I can only watch at the moment. Don't forget Sabertooth is one of the worst boards for AM3 CPUs, at least for Pi, Geekbench and CB might be different. Some examples (not perfect, but you get the point). Both are excellent benchers: DDR2 + 4969MHz https://hwbot.org/submission/3252830_noxinite_cinebench___r15_sempron_145_119_cb DDR3 + 4662MHz https://hwbot.org/submission/2518880_strunkenbold_cinebench___r15_sempron_145_120_cb

Nice one! In theory sempron + ddr3 would be untouchable for a Xeon. Hopefully I will try mine soon, already have it running up to 450+ HTT with DDR3 board, just need LN2 It also scores much higher in other benchmarks

I have a Koolance LN2-V2. PM me if interested and we can arrange a low price, since it's only taking space. I am from Bulgaria.

Well, it might be due to board layout, hidden latencies and timings that can't be set manually. I don't exactly why, but crosshair iv is just slower. You might try some old bios, but I could never match Crosshair III even with higher memory clocks. It was a long time ago, so I might have missed something back then, however that's my observation. Haven't tested all the boards out there though. I also have a 890G Asus board which is even slower and has less options in bios. The problem with newer Gigabyte revisions is the microcode which is Orochi.PI and not the old AGESA anymore, so old AM3 CPUs run in "compatibility" mode. That's what I think. The 990FXA-UD3 rev3 I have is a POS, looses dual channel above ~1700 mem even with loose timings, can't clock FSB high, some settings don't stick and overall has a crappy bios. I hate that board. Unfortunately I have dislocated my shoulder 2 weeks ago, otherwise I would test CHIV, CHV and some giga boards. PS: I set noexecute no matter if it is exposed in bios or not.

They have the old trusty Award bios interface, while newer revisions have the new flashy, slow and buggy graphical UI where you can also use mouse.

Older revisions of 9xx giga boards are surprisingly efficient (non-uefi bios). Not sure if all of them, but at least some.

I guess it is a non-uefi board.

Practically - yes. But not every board and not every revision. 890 ud5, ud7 are good, some of the 9xx boards too, but I think it greatly depends on revision and we know Gigabyte has ton of them... My 990FX-UD3 rev 3 is the worst board I've ever had. From Asus - Crosshair III. Not sure about other brands. Crosshair IV isn't bad, but still slower. The board I am using is very tricky with awful vdroop and only good for single-threaded benches, not really suitable for multithread like Cinebench and wrpime.

Board. Most people were benching on non-efficient boards, some even on an awful boards for Deneb, e.g. Sabertooth or M5A99X. I was expecting some LN2 scores from you, though

To be honest, looking at country cup scores, @Bullant is the only one with a nice efficiency in 32M.

Hmm, I was looking at my scores and thought something like this should be possible: http://hwbot.org/submission/2324360_i.nfrar.ed_pifast_athlon_xp_m_3000_(barton)_48sec_930ms Below 50s in any case, maybe timings too slack?

What about 257-258 x11? Should be faster even with the lower CPU freq?