|
Post by Vinster on Feb 9, 2020 23:01:45 GMT -5
I never would had thought that to make a difference...
Vin
|
|
|
Post by ground on Jul 12, 2020 7:12:23 GMT -5
Finally went and tried the bloodrage I got almost a year ago. Seems to work alright, all air: valid.x86.fr/h9kb35
|
|
|
Post by Vinster on Jul 12, 2020 12:46:31 GMT -5
Did it also come with the NB Water Block? or just the air block? It's a nice board.
Makes me want to fire up one of my X58's and see what I can do with them from what I've seen you post. Generally the Gigabyte V1.0 X58A-UD5 or my V2.0 X58A-UD3R and even my Classified III e770 won't get the CPU's I have over 210-220blk... this is when going for high clocks, not FSB. so Multi's are at 21-22. beyond 200 I usually have to start clocking down memory ratio's and then start getting issues clocking up QPI and balancing temps... Vin
|
|
|
Post by ground on Dec 1, 2020 4:47:57 GMT -5
|
|
|
Post by Vinster on Dec 2, 2020 13:02:33 GMT -5
wee bit
Vin
|
|
|
Post by ground on Dec 6, 2020 21:29:32 GMT -5
|
|
|
Post by ground on Jan 14, 2021 7:22:43 GMT -5
Found an alright 975. 32M on weak Cascade (-80 load), seem worth a try on R3F with LN2. I think I got 5500 to finish too but can't find screenshot. No waza no tweaked uncore no tweaked mem, just wanted to see if the chip is worth keeping at all. 1M on same Cascade Bit of 32M testing on air: 4000 core, 1000 mem, 3600 uncore, 7-7-6 no waza 8:52.937 abload.de/img/852nowazajoko4.png4000 core, 1000 mem, 3800 uncore, 7-7-6 no waza 8:51.484 abload.de/img/851nowazaoakj8.png 4000 core, 1000 mem, 4000 uncore, 7-7-6 no waza 8:49.250 abload.de/img/849250nowazaitjnr.png4000 core, 1000 mem, 4000 uncore, 7-7-5 no waza 8:48.437 abload.de/img/848437nowazan3j2h.png4000 core, 1000 mem, 3800 uncore, 7-7-6 waza 8:44.797 abload.de/img/844waza9vjbt.png4000 core, 1000 mem, 4000 uncore, 6-7-5 waza 8:42.719 abload.de/img/842719wazaylj7w.png
|
|
|
Post by ground on Jan 31, 2021 14:48:10 GMT -5
|
|
|
Post by ground on Feb 2, 2021 17:30:21 GMT -5
So, bored and disappointed me (since my x58a-oc might have died) tried to kill either a stick of reject Gdie or an L5640 with VTT VDIMM delta. I started at 1.4V VTT and 2.0V VDIMM, which obviously worked. At 1.4V VTT and 2.75V VDIMM I was running into issues posting; upping VTT to 1.48V solved that issue so I was able to post 2.8V which seemed to be the limit. No success in killing nor being able to post any higher then 2.8V VDIMM.
I think this at least confirms my suspicion that the risk of degrading an IMC with sensible vdimm seems quite unlikely at this point. Obviously still not gonna recommend it but I'll no longer worry much running about 2.2V even on my good IMCs. Will try the same on bloomfield tomorrow.
|
|
|
Post by Vinster on Feb 3, 2021 17:41:30 GMT -5
wait, what? you tried to kill it and couldn't or you tried and succeeded?
Vin
|
|
|
Post by ground on Feb 4, 2021 7:23:14 GMT -5
wait, what? you tried to kill it and couldn't or you tried and succeeded? Vin I tried killing an IMC (since the claim has been >0.5V difference between Vdimm and VTT is dangerous) by increasing vdimm way beyond agreed "safe" limits, all the way to 2.8V (which didn't post but didn't kill anything). Nothing was killed in the process, not the outcome I expected tbh. Will try again and see if I can damage bloomfield with vdimm once I find the time.
|
|
|
Post by ground on Feb 4, 2021 19:14:04 GMT -5
1M testing with different mem: BBSE + BDBG 2666 8-11-7-22 tight 2.05V abload.de/img/snaphsot0008y6jy8.pngHyper 2100 6-7-5-20 tight 2.05V abload.de/img/snaphsot0011m0k4q.pngHyper run was a lucky one, later tries I only got 10.031 as well. Didn't get any steps in between. Tomorrow gonna start testing more memory relevant benches (read: 8M or 32M) with high clocked BBSE, BDBG or PSC vs Hyper at tight timings.
|
|
|
Post by ground on May 2, 2021 9:33:35 GMT -5
valid.x86.fr/t35p12 DDR3 3652 11-15-15-24 on x58. 2.45V Vdimm, CPU at -60, memory on air. Memory limited now.
|
|
|
Post by ground on Jun 13, 2021 15:46:26 GMT -5
|
|
|
Post by ground on Jun 26, 2021 15:47:17 GMT -5
|
|
|
Post by Vinster on Jun 27, 2021 10:58:22 GMT -5
Impressive. how you keep pulling this stuff off amazes me.
Vin
|
|
|
Post by ground on Jul 12, 2021 17:11:23 GMT -5
|
|
|
Post by MetalRacer on Jul 12, 2021 19:19:21 GMT -5
That's awesome stuff man!
|
|
|
Post by cbjaust on Jul 12, 2021 21:19:12 GMT -5
so close on the SPi 1M score with less frequency all round. I'm guessing you might revisit that one.
|
|
|
Post by BlindGrimReaper on Aug 4, 2021 23:12:33 GMT -5
So what happened you giving up after all that troubles of getting what seems like an impossible goal to be honest fair play to you in achieving
|
|
|
Post by Vinster on Aug 5, 2021 13:33:48 GMT -5
So what happened you giving up after all that troubles of getting what seems like an impossible goal to be honest fair play to you in achieving what do you mean "giving up" we all bench and tinker at our leisure. if you look at the dates of ground's posts, the dates have gaps. it's his process, it works. to each their own. Don't criticize others.
Vin
|
|
|
Post by BlindGrimReaper on Aug 5, 2021 16:22:56 GMT -5
I am not criticizing anyone I know it alot of work to get to the goal that the chap got and I know he put blood, sweat and time to get the results he wanted. Anyway change the subject for abit. I have a xeon x5675 pairing with asus p6x58d-e as I need to know what the safe max voltages in the bios as I am having a hard time finding the information regarding to set it up for overclocking
|
|
|
Post by Vinster on Aug 6, 2021 22:26:32 GMT -5
that's a partially tricky question. Make an account and start a new thread. Add details on cooling and there are ppl here that can provide suggestions.
Vin
|
|
|
Post by ground on Aug 11, 2021 1:54:09 GMT -5
I don’t get why there is an issue to me deciding that a category is DONE for me, given I got the platform record + my best chip died doing so? My other chips are nowhere near it (the better ones are low 290s). I’m not stopping x58, I’m stopping the hunt for higher bclk. I’d not call “stopping after getting the record” giving up.
Safe-ish volts for ambient are, assuming you can cool it, 1.5V Vcore, 1.5V QPI/DRAM. DRAM as far as you feel safe feeding the memory itself, I’ve never managed to damage a CPU with exceedingly high vdimm, even at 2.8V.
currently working on optimal mem setup for pyprime for teamcup, so not sharing my findings until TC is over. So far it looks like hyper at cl6 is fastest (1030mhz 6-6-4-19 super tight subs), but there is a lot more left to test. W3680 is playing nice, >5500 on chiller
|
|
|
Post by ground on Aug 18, 2021 16:04:06 GMT -5
|
|
|
Post by ground on May 9, 2022 15:45:01 GMT -5
|
|
|
Post by Vinster on May 11, 2022 21:49:53 GMT -5
that's insane. I'm sure that took quite a bit of time.
Vin
|
|
|
Post by maddmutt on May 25, 2022 16:33:20 GMT -5
I don’t know how I didn’t find this area sooner :-( I would have asked a nicer Question here instead of here ( warp9-systems.proboards.com/thread/928/intel-1366-x58?page=3) So from reading this area - It’s my i7-920 that is holding back my R3E and E760 MB. I’ll just start grabbing some off E-B as they are $10-$12 ea.
|
|
|
Post by maddmutt on May 26, 2022 14:14:48 GMT -5
I’m going to have to do some more testing tomorrow… Same setup in BIOS all I did was install my i7-960 and a GTX-690 (I had just ordered it to replace my HD-3650 that I had been using). I was able to boot straight to Win 7 @ 235 BLCK. The PCI only needed 106 was 105 but crashed loading win 7 @ 227 BLCK. Could I have over shot the PCI @ 110 @ 220 BLCK & 115 @ 225-227 BLCK which caused the i7-920/950/965 NOT to BOOT/POST faster than 227??. OR It could be the GTX-690 is better than the HD-3650 in this setup for OCing?? I do love the i7-965ee as it has a MAX Multi of 50 x 200 = 10GHz :-) I’ll never see that ;-) I’m using the Corsair Dom GT - 2133 @ 14xx (2x6 Multi) I’ll give a full break down of the EVGA e760 BIOS setting tomorrow.
|
|
|
Post by Mr.Scott on May 26, 2022 15:00:05 GMT -5
More PCIe, 110 or so.
|
|