I.nfraR.ed

Hey, what happened to Raymarch test?

Most probably mem controller. And ofcourse capable RAM. The board is capable though, but depends on the cpu a lot. Here I have 2x2GB dual channel 661MHz 5-5-5 I.nfraR.ed`s Memory Clock score: 661.3 MHz with a DDR2 SDRAM Highest on the bot for 790FX chipset with DDR2 is 725MHz.

Some people suddenly lost common sense?

Ah, for this competition CPUs with 512KB L2 cache were not that good, although I bet my 3.7GHz Venice would do fine. But 3000 and 3200 were a no go for sure, I was talking in general. For this competition you were looking for nice scaling San Diego 3700/4000, FX-55/57 San Diego for Cinebench, for others a X2 was needed - basically higher bin Opterons and FX-60. Perhaps some Toledo, Manchester and Clawhammer would do fine as well, but they are slower clock for clock. FX Clawhammer and even 4000+ have lower CB and being 130nm can take higher voltage, but are a little slower than San Diego. Some other CPUs would do good also, but these were the best choices.

Sometimes bad air chips fly on LN2 with really unexpectedly low CB. You should test all your CPUs on clod, no matter if they are bad on air. Well, this applies to less extent for lower multi chips like 3000+ and 3200+. I haven't found a single one that has an extraordinary low CB. And they need high FSB, so if they are really bad on air, then won't do much on LN2. By much I mean to break the records for these CPUs. Generally a CB around SS temps (-45/-50) is considered very good. One of these bad-air-good-on-LN2 is my 3800+ Venice. It could not even do 3GHz on air, but has -98C coldbug which is insane for these: AMD Athlon 64 3800+ (Venice) @ HWBOT So if Stelios' Opteron has a nice coldbug around even -40, then it will fly. For example my 180 has a CB around -8 and does ~3.3GHz for CB11.5. There are some other tricks for single core CPUs, but they don't help much. You can't lift that CB. You can only try to somehow increase temperature even on idle, because that's where it bugs mostly. When loaded it can usually do higher clock, due to the higher temperature. Which you compensate with voltage. PS: Unlocked chips (a.k.a FX) is best to be clocked with multi and low FSB (~225 max) and 250 divider for mem. This way you can get much lower on temperature, which translates to higher core clocks, but you all know that.

I noticed too there was a strange clock on the general page, like 12 hours more left compared to the counters on the stages. Basically it seems we had half a day extra, but not sure if just for this stage.

Someone tried HyperSLI patch then? We were able to run SLI on a 790FX chipset, which does not support it officially. Or some older driver? I have SLI-DR and Expert boards. Also one Ultra-D with chipset modded to SLI-D, but haven't actually tried. Might do in the evening, since I had the board setup, just no time to run it.

Nice scores everyone! I wanted to participate, but had no time and also have more important things to sort out. Hopefully I would be able to participate in some of the future competitions. I wonder why noone ran 2x7900GTX. Was it limited by the platform, was it buggy, was it not scaling or smth else? I have 3 of these, but didn't even had the time to mod 'em.

Thanks for the picture. This might actually help me hit even higher frequencies with my little 145 in the future. Seems to be a very tight fit, some elements are very close to the IHS edge, just like some s.939 chips. And I think it is advisable to leave the glue on the PCB, otherwise it might lead to excessive pressure on the DIE. You have a free bump

Damn, I would have taken the first GTX2 kit. We all know you can never have enough hypers .

Common, get it tested on LN2 already As you know, I'm waiting for you to test, since we have very similar CPUs with same batch, scaling, etc. And I won't be able to test mine on LN2 again until Apex is here.

Show us the 6.9G run already.

Good info, didn't know they have TIM under the IHS. I thought they are soldered just like the Phenoms. Do you have a photo of the naked chip?

Not to spoil the fun, but 4.2ghz at 1.5v is nothing extraordinary. I've sold two low-volt 4.5GHz+ (stock cooling) before and these were my rejects. For hw records they need to have other capabilities, not just pure core clock potential. Any reason to delid it? Aren't they "soldered"? Any thermal paste will be worse than stock and also needs maintenance, i.e. you need to replace the paste regularly.

My AMD clocks much higher.

Second try today - same thing. Can boot 6.5 1.73V no problem, can also set to 6800 same volts in Win. Can't run anything even at 6.0 when full pot - instant freeze on CB15 and CB11.5, have to restart from button on OCF. Before going full pot I tested again the saved profile at -90 6.0 1.6V and it passes no problem. Re-applied grizzly on DIE and IHS before the session, temps were staying under 0 this time. No cracks, no nothing, but same behavior under full pot. Definitely need to try another mobo before rejecting this chip. Also tried with XMP mem, low cache, 2T on mem, 1.8 CPU PLL, up to 2V VCC PLL, VCCIO up to 1.4, VCSSA up to 1.4. Tried different bioses, tried higher vCore. Mobo is running without radiaor on MOSFETs and sensors in bios show negative temps (both around -10) when going full pot. PS: Maybe I should stick to AMD, lol. Intel hates me.

Did you try to set vcore by offset? 0 or the minimal allowed. Basically load UEFI defaults, then change vcore to offset mode. Some bioses on big OCF give me different value: ~1.2, but by offset I have the same 1.104 across all bios versions.

I always pre-heat the pot to about 50 before applying tim. Then mount, cool down, start the system. And I always start running benchmarks at say -90, then gradually decrease temp and check if it scales. Didn't go full pot at once. Today after first attempt I tried heating to 70 as Allen suggested, then cool down again, start the system, but still the same. The CPU was very hot with stock tim before delid. I could not do much more than 5.1G CB15 because of temps. After delidding the temps seem ok. Maybe I just need to try on MOCF with CB Killer volts. Will try to get one soon. PS: Or it just doesn't scale under cold, as someone said about L643 batch.

6.2 -120 1.6V a little bit over 0, then full pot it stays under 0 when it actually passes. Full pot no problem booting and restarting, even saving bios settings, then doing stuff in windows, CB15 and CB11.5 freeze on the 1st set of squares.

My L643G I wrote about earlier, but after delid and AIO cooling, Grizzly on both, CB15 5.0/4.5 1.189V (measured with DMM) 5.2/5.0 1.328V VID is 1.104 according to OCF board. On LN2 and Z170 OCF - what I tested, it might be able to get higher with same settings 6.0 1.6V -100 6.2 1.6V -120 Then below -130 the "fun" begins. It starts freezing under load, nothing seems to help. Full pot same 6.0GHz profile I saved freezes when I start a bench. Pushing VCC PLL and CPU PLL to 1.8+ helps a bit, but still freezes on higher frequencies, 6.2 no go, while it is ok on higher temps. I will try to get MOCF to test. 6.0 cache booted, didn't test if CB will pass, I wanted to check cores first.

Wow, Corsair and good DDR4 mems

Looking at the photo, it seems the standoffs are not flush with the mobo. Don't know if that's normal though - don't have OBT.

Good one nevertheless! Yeah, my CPU can do 5GHz NB and that little boards is also very efficient (which is surprising for 9xx chipset), but I can only use my budget board for single-threaded benches, because of the weak VRM and enormous droop. UD5 is one of the best (if not the best) AM3 board. Maybe I should get one.

I have used this NB heatsink by Zalman

Old northbridge air cooler as a base, then insulated toilet-paper roll and sealed to the norhtbridge cooler: I.nfraR.ed`s wPrime - 32m score: 21sec 466ms with a Sempron 145