Unless you guy like playing craps, or have a Six Code CPU enroute. I wouldn't bother with 43,44, or 49. The Microcode changes made for the Gulftown/Westmere are not working well with the Bloomfeild processors. So you Folks with Quad cores, 920,960,965,and 975, probably be better of with the bios prior to bios 43_beta. The microcode thing is an intel thing, Not EVGA. So until Intel works out the "kinks" in the microcode, it's best to go back to the bios that best works for you.

Guys, can I flash back to SZ2Z on my 758 from 49?

I can't wait any more, I need VT. :\

Governator

Nereus

I think it's been recommended not to load saved profiles on a new BIOS, but to enter them manually after flash. That may have been part of the problem (this is just a suggestion, not a criticism).

This is true and a good suggestion at that. BTW I love your avatar

firerx

Unless you guy like playing craps, or have a Six Code CPU enroute. I wouldn't bother with 43,44, or 49. The Microcode changes made for the Gulftown/Westmere are not working well with the Bloomfeild processors. So you Folks with Quad cores, 920,960,965,and 975, probably be better of with the bios prior to bios 43_beta. The microcode thing is an intel thing, Not EVGA. So until Intel works out the "kinks" in the microcode, it's best to go back to the bios that best works for you.

firerx, what issues are you referring to exactly? I'm on 44 and have none.

If you have CxE Function enabled with Turbo Mode the 2x bin multiplier for Core 0 (22x on the 920) will kick in more often than it did with previous BIOS (infact it pretty much never kicked in during normal use in previous BIOS's), when the 22x multiplier activates it can cause problems with peoples overclocks (people are experiencing BSOD's when it activates).

If you dont use CxE Function then you wont notice this issue, also if you have CxE Function enabled but Turbo Mode disabled it wont happen either.

firerx

Unless you guy like playing craps, or have a Six Code CPU enroute. I wouldn't bother with 43,44, or 49. The Microcode changes made for the Gulftown/Westmere are not working well with the Bloomfeild processors. So you Folks with Quad cores, 920,960,965,and 975, probably be better of with the bios prior to bios 43_beta. The microcode thing is an intel thing, Not EVGA. So until Intel works out the "kinks" in the microcode, it's best to go back to the bios that best works for you.

 
THANKS FOR THAT! I asked about that, Oh, about 6 or 7 pages ago. 

Zsoltom

Is this new 49 bios supported by Xeon X5640 processor?

firerx

If you asking about the Westmere-EP's , You'll have to wait for Shamino to answer that one. I know they were working on it. But unsure of the results

Actually, there are two hardware mods needed to run those. Iirc the mods disable one of the QPI links.

Hmm... I wonder if this is why I and another user reported lower GPU scores in Vantage using 5970s.... PCI-E bandwidth bottlenecking would explain it.

You guys should PM and email Jacob directly with your findings before they release a new BIOS.

Balthazor

Hmm... I wonder if this is why I and another user reported lower GPU scores in Vantage using 5970s.... PCI-E bandwidth bottlenecking would explain it.

Darn I didn't notice if this is what was specifically happening to me when I had the 49 bios loaded.    I honestly don't feel like putting it back on just to downgrade right after.  :(

I didn't check the PCI-E bandwidth when I had the 44, 48, or 49 BIOSes installed, but under all three I was experencing noticeably lower GPU scores in Vantage when compared to the S61J (with all settings being the same.)
 
I would only write Jacob if I reinstall 49 and test it myself, because at this point it is just a hypothesis.  And because 49 was interfering with my GPU performance I'm not using it currently, and I don't really feel like testing it.  Maybe next BIOS version, if I'm still seeing the same issue, I'll test it more thoroughly, but there comes a point when I just want to use my computer and not spend my time trying to figure out why a vendor's update is interfering with the system.

lcabezas

lam_das

lcabezas

You guys should PM and email Jacob directly with your findings before they release a new BIOS.

Like that's ever helped...Evga techs don't read their private messages and they certainly don't read these forums after they have released the BIOS.

Jacob answered me about the VT problem, maybe he will answer you guys. I say just bug the crap out of him and other EVGA tech till they answer you lol. I do believe though they they hardly ever read forums after they released the new bios because their lack of input in the past.

I have posted in the forums for every release of BIOS and sent PM to evga techs describing S3 issues in details. I never got any response from them. I even called tech support and they refuse to acknowledge the S3 issues. As far as evga is concerned, there is no problem with S3. So, that leads me to the conclusion that they don't read PMs and forums.

Alucard666

This bios killed my Ram....

http://forums.evga.com/tm.aspx?m=243297

seems somehting has changed on the Memory front even in 49, the Round Trip Latency setting for each channel has changed over previous BIOS's if left on AUTO, more specifically for me they are looser than previous BIOS's, maybe this was necessary to get the 14x mem multi (1866) working again (since it was broken in 43), not sure how that could damage memory but it indicates something changed in regards to memory and since memory damage has occured after a BIOS update it may show a link.

rwoodburn1

Having a lot of issues with this bios.VTT is way higher than what you set in bios so is memory voltage.I know the built in windows tests are not that great but when I run winsat mem, my memory score dropped from over 30000 to 26000 with same settings of 7 8 7 20 1t shows 2006 frequency in bios and 998.5 frequency in cpuz. I am also getting random bsod where I was stable before.I am getting a little woried about sticking the new chip in I ordered.BTW I am using 965 extreme

I am not sure if you are questioning it but CPU-Z showing half the speed is normal.

DDR means Double Data Rate, which means the Memory can do 2 operations per clock, so if the Memory Clock speed is 1000MHz (close to your 998.5) then the DDR speed would be 2x1000 = 2000 (close to your 2006) and its sometimes listed as MHz. CPU-Z shows the Memory Clock whereas the BIOS shows the effective DDR speed which is double the Memory Clock for DDR3.

The reason CPU-Z may not show exactly half is due to it using the actual BClock, there will be a small variation from what you set the BClock to in the BIOS and the actual speed produced.

The reason for your BSOD's where you were stable before could be down to Round trip Latency for all 3 channels, if you dont set them manually they will change anyway on every restart, the BIOS recalibrates them but also the new BIOS sets them ttoally different to previous ones, it should make them looser which normally should be ok but the slight change may be throwing somehting off.  Its best to see what the BIOS usually sets them to and then manually enter them

For example for my DDR3 2000 8-8-8-24 the Round Trip Latency for the3 channels are
Channel A RTL = 60
Channel B RTL = 62
Channel C RTL = 64

but if i leave them on AUTO the majority of the time they will be 60/62/64 but sometimes they will get calibrated differently, for example 59/61/62 as you can see the difference between each channel is no longer the same, this can sometimes mean the difference between passing a stress test and getting an error.

On a high Memory clock or OC its sometimes advisable to find what the BIOS sets them to on AUTO and then add an additional 5 or 6 to them, so instead of being 60/62/64 i would set them to 65/67/69 for example.  This can sometimes make things stable, Round Trip Latency can be pretty important to memory stability on the X58 platform.

rwoodburn1

Thanks Moltenlava.Yea I am aware of ddr lol,but before my mem was something like 1002 not the 998 like it is now.I will give the latency a try,I normally leave it on auto.I am not overclocking much just trying to get stable till my 980 arrives.Appreciate the help.

sorry rwood, didnt really look at the name, just read the post, if i did i would have realised you wouldn't need to be told that :).

Boy, I'm thinking I'll stay with 44 for now. I also liked S61J but both it and 44 have been fine for me, nothing out of the ordinary.

I don't know anything about the most recent update to the "759" bios, I haven't tried it yet since I am still trying to get an undrestanding of this "44" version first.  I was wondering if anyone could help me out on this:

1. My bios settings are different than what E-leet is displaying for my memory settings
    a. It is set to 2:10 a speed of 2133mhz in the bios, it reads a setting of 2:8 1683mhz in E-leet 
        and in everything else I look at it through.
    b. the voltages set in the bios are also reading differently as well.  set to 1.64v in bios, reads 1.71v in E-leet

2. I was told by tech support that you should go off of the CPU temp and not the "core" temps when monitoring thermal activity.  I just wanted to know what was more preferred since my CPU temps usually only spike around 72c after about 7 or so hours of Prime95.  But the "core" temps are another story every one knows I am sure.  I just would like to know what most people prefer and what to do to stay on the safe side of things. 

Case: Corsair Obsidian 800D
CPU: i7-920 @ 4.42ghz Koolance CPU-350AT
PS: Silverstone ST-1500
MOBO: eVGA e-759 bios 44v
MEM: G Skill Perfect storm 6gb
     2x3gb @2133mhz 8-8-8-21-60 1T
SC: Creative SB Champ Ed.
GPU: X3 GTX 275
SSD: OCZ vertex Turbo x2 Raid 0
HDD: X3 Seagate Mom. 500gb
HDD: X1 WD cavair 250gb
Mon: x1 Dell 24" x1 25.5" Westinghouse

Sorry about that, will ask the question on another post.....