This morning we woke adult splendid and early to benchmark some DDR4 memory kits and found myself waking adult not to Folgers in your cup, nonetheless a smell of burnt electrical after loading a XMP profiles on a memory pack and restarting a system. Let me tell we what happened, a best we can.
On Friday we spent a day jacket adult a benchmarks on a Kingston HyperX Predator DDR4 16GB (4x4GB) 3000MHz quad-channel memory pack (part array HX430C15PBK4/16) that runs during 1.5V. This morning we wanted to get another pack of DDR4 memory tested, so a complement was powered down, a CMOS was reset and a G.Skill 16GB (4x4GB) 3000MHz memory pack (F4-3000C15Q-16GRR) was commissioned and powered adult a system. It posted fine, so we went into a UEFI and set it to run during a usually XMP form on a kit. The UEFI changes were saved and a complement restarted. It was during a successive seconds that both a house and a processor would be killed off in a rather unspectacular death. The complement came up, hung for a unequivocally brief time and afterwards powered off with a listened click of a Corsair AX860i energy supply. If you’ve ever listened a shrill click of a Over Current Protection (OCP) shutting down a PSU we know accurately what click we heard. Now when we press energy symbol on a motherboard a complement clicks after being on for a apart second. we unplugged all a cables on a energy supply and did a built-in self-check and it upheld with drifting colors. we substituted out a PSU with a backup Corsair AX860i and a same click was to be heard. After clearing a CMOS, stealing a memory, SSD and video label a complement still wouldn’t post. At that indicate in time we switched to a non-digital energy supply (Corsair AX1200) and it did a same thing nonetheless this time a OCP took a small longer to flog in. There was some listened crackling noises, followed by some fume nearby a CPU VRM heatsink. So, a heart ruinous smell of burnt wiring filled a room and we knew my day wasn’t going to be a good one.
I private a house from a exam dais and started to do a visible review and couldn’t see anything wrong with any of a components on a front or behind of a board. we know where a fume came from, so we private a VRM heatsink and a burnt electrical smell got stronger. There was some blemish new to where one of a mosfets sits on a thermal pad, so clearly it was a disaster of CPU voltage law complement and one of a 8 60 Amp phases (Dr. MOS IOR 3550M mosfets) has seemed to fail.
It isn’t burnt badly during all, nonetheless we can some of a signs of an electrical disaster on a second energy proviso from a bottom.
Looking during a house we can see that a unsuccessful member in doubt is partial of a PQ1004, that is partial of a VCCIN or fundamentally a processor submit voltage. Crap! On these Haswell-E processors, Intel has altered a voltage law on-CPU as partial of a new Fully Integrated Voltage Regulator (FIVR). Previously there were 5 apart submit voltages a motherboard handled: Vcore, Vgpu, VCCSA, VCCIO, and a PLL. On Intel Haswell-E processors all 5 inner energy rails are pulled from a singular VCCIN and a components on ours usually had a chief meltdown.
You can tell things got flattering prohibited as there were indeed solder balls where they weren’t supped to be!
Our misfortune fears were reliable when we pulled out a backup ASUS X99 Deluxe motherboard and put a strange Intel Core-i7-5960X processor in and a complement wouldn’t post. The play debug arrangement showed Q-Code 00, that is a bad sign. We tossed in a backup Intel Core i7-5960X processor and a complement booted adult usually excellent and we are off and means to benchmark again. The bad news is that we managed to kill an ASUS X99 Deluxe motherboard ($398.99 shipped) and an Intel Core i7-5960X processor ($1049.99 Shipped) after regulating it for reduction than dual weeks, that is a bit surprising and because we am pity information about this disaster to a readers of Legit Reviews. It is not an bland occurrence where $1450 in hardware gets put out to pasture.
LR isn’t a usually site that has had a house go adult in fume as Michael Larabel over during Phoronix had an X99 house go adult in fume as well. He was not regulating a same formula of motherboard or energy supply model, nonetheless to see X99 play unwell this early in a diversion is alarming.
I’ve been in hit with ASUS, Intel, Corsair and Kingston and no one is accurately certain what happened to a system. Was regulating 1.50V on a DDR4 memory too much? Is there something wrong with a VRM pattern or did we have a bad member on a motherboard? Was a energy supply faulty? We aren’t sure, nonetheless we are going to be overnighting this house behind to ASUS Taiwan on Monday (9/8/2014) and we have organised for Corsair to put a PSU on their scopes and exam apparatus to make certain it is operative properly.
We’ll let we know what happens if we find anything out in a weeks ahead!
Update 9/9/2014 - ASUS has perceived a X99 Deluxe motherboard that unsuccessful along with a energy supply that was in a section during a time of a failure. An ASUS worker will be holding a PSU to Corsair for contrast on 9/10/2014. The house will be looked during here in a US and afterwards will be shipped to ASUS HQ in Taiwan. In a meantime we have dual ASUS X99 Deluxe motherboards being used by staff members that are still adult and operative scrupulously and we have been told by ASUS that they have replicated out accurate complement and that they tested it overnight and haven’t gifted any failures. We’ll keep we updated and will post a tangible Chroma exam formula from a PSU once we get them from Corsair. It was also schooled in a past 24 hours that a ‘high’ DDR4 voltage (1.35 to 1.5V) that we were regulating on a house shouldn’t have caused any issues given a memory controller in Haswell-E can indeed support DDR3 and DDR4, nonetheless usually DDR4 is being implemented. DDR3 memory kits run during 1.5V, so that roughly eliminates that from being a issue. The good thing about posting publicly about this disaster is that all a companies concerned are holding this severely and are operative overtime to safeguard there isn’t an emanate somewhere. We are unequivocally blissful about that as we have endorsed a board, processor and energy supply and wish to safeguard that a height works if a readers squeeze a tools formed on a recommendations.
Update 9/11/2014 - ASUS HQ in Taiwan has a motherboard and is looking during it now. ASUS and Corsair were incompetent to get a energy supply tested on Tuesday and scheduled a time with Corsair to get a energy supply on a Chroma tester on Friday 9/12/2014. This was a beginning that a Chroma tester was accessible for use during business hours.
Update 9/13/2014 - Corsair along with ASUS tested a span of AX860i energy supply that Legit Reviews was regulating on a exam dais during a time of disaster on Friday a 12th. Both energy reserve upheld a initial Chroma exam passes, nonetheless we schooled something that we formerly did not know. When a Corsair AXi array of energy reserve came out in 2012 they featured a singular rail design. Corsair switched to a churned rail pattern for a energy supply array in 2013 (previously opposite to us). This is apparently a poignant disproportion in a pattern of a energy supply. We also schooled that a progressing singular rail energy supplies did not have OCP enabled by default. One would have to implement the Corsair Link Software package and manually set a OCP boundary manually for that to duty on a progressing models. We were not regulating a Corsair Link Software on a exam bench, so therefore a energy supply could have 90A or some-more potentially regulating down a rail. This competence have exacerbated a repairs to a CPU’s VR circuit if there was a bad member or solder turn corner present.
We looked around during Newegg, Amazon, Scan and other vital online retailers and have a Corsair AX860i as carrying a singular +12V rail. Heck, even Corsair.com has it listed as carrying a singular rail! Hopefully now that this has come to light a folks during Corsair can refurbish their possess site and get retailers to scrupulously list a specifications of this PSU in a listings.
ASUS is still operative with a unsuccessful house and is going to be replicated a setup with a energy reserve in Taiwan. Yes, these dual energy reserve have now been shipped to Taiwan where ASUS HQ will be means to exam a initial rider AX860i Power Supplies as a ones they were contrast with this past week perplexing to replicate a complement nonetheless disaster was finished regulating a new post-2013 energy supplies. The accurate means of a error is not known, nonetheless many is being schooled by everybody and many of it is profitable information that will assistance a community. We’ll news behind with some-more once ASUS is prepared to give us an refurbish that is prepared for open consumption. That will occur after additional contrast is finished on a house over a days ahead.
Update 9/16/2014 – We are still operative with Corsair to find out some-more on a firmware refurbish that was finished on their energy reserve behind in 2013. We have asked for dates and energy supply lot numbers, so users can find out if they have one of a strange ‘old’ AXi array energy reserve that has no OCP by default. We also forked out to Corsair that there is no plead of this in a instruction primer and that many users competence not be wakeful that their flagship PSU has facilities that aren’t enabled unless they do so manually. From a sounds of it Corsair usually updated a firmware and went to a multi-rail configuration. We’ve talked to several people about this emanate and it was misleading if there was a hardware change and that is still being looked into. The bad news is that a firmware is not end-user upgradable. We have asked Corsair what if anything stream business can do given a firmware can’t be upgraded in a field. If we have an AXi array energy supply we rarely advise downloading a Corsair Link program and programming a OCP setting.
Kingston Technology contacted us currently and sensitive us that they will be obscure a voltages on a pre-production DDR4 memory kits that were sent out during 1.5V to 1.35V when they are shipped out to consumers. Kingston never shipped any DDR4 memory kits during 1.5V and won’t be. It doesn’t seem that a memory regulating during 1.5V had anything to do with a failure, nonetheless ASUS is still testing. We haven’t listened from ASUS in a past 48 hours and final we listened they were still looking into things in Taiwan now that a ‘old’ energy reserve nonetheless OCP have arrived. ASUS pronounced they will be giving us an central matter about a disaster when a investigate is finished and we wish that will be someday soon.
Update 9/17/2014 – Replacement ASUS X99 Deluxe motherboard was delivered (11 days from indicate of disaster to deputy house being delivered).
Update 9/18/2014 - Corsair has gotten behind to us with some answers to some questions that we asked progressing this week. It turns out Corsair shipped AX760i/AX860i/AX1200i energy reserve for about 4 months before they altered a firmware on them nonetheless notice. The firmware is not margin upgradeable and Corsair will not be charity exchanges for anyone with an ‘older’ indication that wants to barter out a PSU for one with a latest firmware on it. Corsair also pronounced that by a motherboard makers [ASUS] possess admission, a X99 Deluxe motherboard was a base means for a failures. Corsair also pronounced this that we will directly quote: “Would an OCP-defaulted AXi or a aspirant OCP-enabled PSU have save a CPU? We’re skeptical, nonetheless maybe.” So, right now it looks like a house had a disaster and afterwards when a complement was restarted a PSU nonetheless OCP competence or competence not have taken out a CPU by a play unsuccessful VR circuit. We are still watchful on ASUS to give us an central matter as to what happened to a house and were told that a gale in a segment this week has slowed things down. In a meantime here are some answers to a QA that we gave Corsair that we can take a demeanour at.
- When did Corsair change a firmware on a AXi array of energy supplies?
AX760i/860i doing date 3/15/2013 Lot#:13119560
AX1200i doing date 3/8/2013 Lot# Lot#:13099520
Corsair shipped a AX760i/AX860i/AX1200i for about 4 months before they altered a firmware on them. If we bought one of these models when they initial came out we approaching have one with aged firmware. The Corsair AX860i initial was finished accessible for sale with Amazon on Nov 1st, 2012, so usually a heads adult to early adopters.
- Can we greatfully prominence what all changes with a new firmware?
PSU set to multi-rail (which by clarification is OCP).
- So, we went from a default pattern of one +12V rail with no OCP to a practical multi-rail setup with OCP enabled by default?
- Why was this change not finished public?
We saw no need for an announcement. The PSU pattern and a facilities stayed a same and this isn’t a pattern fault.
- Can finish users with a strange PSU pattern refurbish their firmware during home?
- How can an finish user know what firmware is on his/her PSU? (Can users brand by a method array what PSU they have? )
By a method number. The initial 4 digits are a date code. The initial dual digits are a year and afterwards a successive dual numbers are a week of a year that a energy supply were produced. The picture above shows a Corsair AX860i Power Supply with method array 1249954 that was finished a 49th week of 2012 and would be regulating a strange firmware.
AX760i/860i doing date 3/15/2013 – First Lot array was: 13119560
AX1200i doing date 3/8/2013 – First Lot array was: 13099520
- If users can't ascent a firmware during home, can users sell their PSU for a indication with OCP enabled by default?
- How many Amps does a OCP default to on a AXi series. we listened it is opposite for any PSU?
By default, 40A. This is configurable.
- we was told that Intel Haswell-E processors are regulating adult to 47A when overclocked to 4.4GHz and that it exceeds a OCP on some PSUs. Some motherboard makers are revelation us to stay divided from certain PSU’s. What are your thoughts on this?
When we have a PSU with churned +12V rails, OCP can simply outing if a CPU is overclocked and regulating over load. This is because Corsair PSUs with Link Digital concede a user to invalidate OCP and because all other Corsair PSUs underline a singular +12V rail.
- ASUS designed a VR circuit on their X99 height with 60A components. Corsair came out with a AXi array in 2012 with an tractable OCP that was off by default. Was Corsair foreseeing a conditions in a destiny where finish users could customize a OCP environment depending on what motherboard they were using?
Initially, Corsair was simply following a existent trend of providing energy reserve with a singular +12V rail. Since OCP is many profitable during a initial build theatre of putting together a PC, it finished clarity for a PSU to have a OCP on by default and therefore we motionless to make a change.
Update 9/22/2014 – ASUS sensitive Legit Reviews that they will need 2-5 some-more days before releasing an central matter on a failure.
Update 10/01/2014 – ASUS is wanting some-more time to plead what they have found out with Corsair before creation a open statement. We have been told that it will be another day or dual until they will be means to contend something. It sounds like some issues have been found during contrast and a formula are being looked into by both companies.
Update 10/6/2014 - The dual deputy Corsair AX860i energy reserve were usually perceived today, that usually happens to be one month to a day from a height failure. ASUS sensitive us this morning that Corsair is requesting some-more information on their exam setup and that indispensable some-more time again. We are carefree we’ll get an answer soon, nonetheless it appears that ASUS and Corsair are carrying issues with what a exam information relates to with a failure. It sounds like churned things went wrong and of march no association wants to take some-more censure than they have to.
Update 10/9/2014 - Just perceived word from ASUS that a Corsair engineering organisation perceived a additional information that they wanted about a contrast ASUS did and now ASUS is watchful to hear back. If ASUS gets an answer tomorrow they are carefree that they will be means to yield a minute response on a disaster or failures that they believed happened to out height on Monday (10-13-2009).
Update 10/13/2014 - We talked with ASUS and Corsair over a past several days and there was plead of wanting a few some-more weeks to figure things out. It appears that Corsair is conflicting with a ASUS commentary about a energy supply we were regulating and a contrast of it during ASUS HQ. Legit Reviews has been kept in a dim for a past several weeks and can’t unequivocally do anything other than wait. Right now it appears that something competence have been off with a energy supply. Corsair gave us a Chroma exam reports from their testing, nonetheless we have not been given anything by ASUS. We don’t know what was seen in a latest turn of testing, nonetheless we do know that ASUS tested a PSU on a complement and not usually a PSU exam machine. What did they find out that Corsair disagrees with? It feels like some people are perplexing to drag this out and wish it blows over. Is it working? We are still seeking answers from both ASUS and Corsair.
Update 10/22/2014 - No refurbish from ASUS USA or Taiwan, nonetheless we did find out that someone during ASUS Singapore posted adult a criticism during Hardwarezone about a failure. ASUS@SG settled that a house unsuccessful due to a inadequate pre-production BIOS/UEFI. They explain that a bug was detected by Intel and a repair was finished by ASUS before a initial central UEFI release. Now Intel is being blamed? The same ASUS eCustomer Service Center worker afterwards followed that post adult with a post saying that Intel could have had a bad collection of Intel Haswell-E processors. Very interesting. Legit Reviews still is still watchful on a central answer from Taiwan. It is unsatisfactory to see probable answers to a issues on other sites as ASUS hasn’t pronounced anything about a bad BIOS/UEFI to us given this whole distress began.
Update 10/22/2014 Part 2 - ASUS USA told us they aren’t certain because ASUS@SG would post such comments and told us that is not a primary reason for a house failure. They didn’t repudiate carrying a UEFI/BIOS bug though, so this is starting to get interesting. Could we be removing tighten to anticipating out because a ASUS X99 Deluxe motherboard failed? Hopefully ASUS will give us something given they have member revelation people what happened on forums, nonetheless afterwards a ASUS employees we are operative with are revelation us that is not a truth. Ugh!
Update 10/23/2014 - ASUS expelled UEFI 1004 for a ASUS X99 Deluxe currently and we have been told that this refurbish includes an EC (Embedded Controller) Firmware refurbish that fixes something detected by a house unwell here at Legit Reviews. We don’t have a central response from ASUS yet, nonetheless Legit Reviews highly suggests that all ASUS X99 Deluxe owners refurbish to UEFI build 1004 due to a fixes implemented in it for a approach a house energy is being handled. The build date on this UEFI is 10/16/2014, so it has been around for a week before it was finished public. ASUS also reprogramed a memory tables after receiving new microcode from Intel. That finished a universe of a disproportion on a house when regulating memory kits over 3000MHz with 1T Command Rates. Here is a list of a pivotal changes:
ASUS X99-DELUXE BIOS 1004 Change Log:
1. Update EC FW
2. Fix pile-up giveaway issue
3. Fix Xonar label harmony issue
4. Revise Thunderbolt memory resource
5. Enhance Xeon CPU compatibility
6. Rebuild SteamOS foot option
We have been told to pattern a final answer on a disaster from ASUS on Friday.
Update 10/24/2014 - Legit Reviews was usually sent a disaster research response on a ASUS X99-Deluxe Motherboard. You can review it in a entirety below:
We have dynamic a primary means of disaster for a pre-production ASUS X99 Deluxe we were contrast on Sep 6, 2014 along with a delegate means collected during a review phase. Our initial research of a VRM Phase-4 MOSFET/Driver package disaster is a bad solder indicate that was also benefaction during a VRM Phase-3 plcae ensuing in a disaster we described along with a participation of solder balls. Additional analyses lead us to trust this was a delegate means for a disaster described.
After endless contrast and partnership with heading energy supply manufacturers and a VRM retailer (International Rectifier) it was dynamic that a new VRM pattern on a X99 Deluxe house needs a firmware refurbish to change start-up and shutdown energy loads and sequencing opposite a MOSFET/Driver packages. We dynamic that aloft loads were placed on a VRM Phase-4 package when a processor was sketch reduction than 70 amps of stream during start-up and shutdown sequences formed on a strange firmware. This along with some comparison energy supply OCP/Shutdown anomalies formula in a bad member multiple that incidentally (very) leads to a VCCIN spike and energy swell on a VRM Phase-4 MOSFET/Driver package that could means member failures. We still have not replicated this disaster in a exam labs after thousands of hours of contrast opposite a poignant array of member combinations. However, we trust this was a primary means for a initial house shutdown and a solder indicate issues exacerbated a successive member failures in a demeanour we summarized when regulating a delegate energy supply. The good news is that we have a resolution to this intensity issue.
We are releasing a new EFI (build 1004) currently that addresses this emanate by balancing start-up and shutdown energy loads opposite all VRM Phases when a processor is sketch reduction than 50 amps. This will severely lessen a possibility of a VCCIN spike or energy swell in singular instances formed on endless testing. Our new balancing/sequencing manners will diminution altogether energy potency formula by a few percent formed on processor loading underneath 50 amps nonetheless differently a board’s altogether opening will not change. In further to a new energy rules, EFI recover 1004 facilities a horde of opening improvements with poignant improvements in a area of memory overclocking regulating a 100 tag during speeds adult to and past DDR4-3300.
We rarely suggest that all users of a ASUS X99 Deluxe house download and implement EFI 1004. This EFI is accessible from a support site starting currently – X99-Deluxe EFI 1004. Please follow a correct instructions for updating your EFI. The refurbish discipline are accessible at ASUS USB BIOS Flashback Guide or follow a instructions in a user primer when utilizing USB BIOS Flashback or EZ Flash 2.
ASUS is resolutely committed to ancillary a category leading X99 Deluxe motherboard that facilities unmatched opening and options like a patent-pending OC Socket, 5-way Optimization for one click overclocking and fan control, 3×3 802.11ac, Fan Expert 3 and Crystal Sound 2.
When a ASUS X99-Deluxe motherboard unsuccessful behind in Sep during a complement restart things usually didn’t supplement up. We’ve spent tens of thousands of hours contrast hardware components over a past 12 years and we posted adult a commentary in a eventuality that others in a fan village competence have identical issues. When this post went online it got churned feedback from many in a community. ASUS seeking us to take it down until a means was figured out was expected, nonetheless we were repelled that other hardware reviewers in a essay posted disastrous comments (still shown below) about how we were doing ASUS an injustice. Unfortunately, we weren’t a usually people in a universe to have an ASUS X99-Deluxe motherboard destroy and roughly all a failures reported online were by folks that also were regulating Corsair AX/AXi energy supplies. Something was apparently going wrong with that multiple and no one could figure out what was causing a failures during first.
ASUS went out and bought Corsair AX860i energy reserve to try to replicate a problem, nonetheless it was after detected that a energy reserve being sole currently were opposite than a ones we were using. It turns out Corsair altered a firmware on their energy reserve and a change was rather significant. The firmware altered the default pattern of a energy reserve in this array from one +12V rail with no OCP to a practical multi-rail setup with OCP enabled by default. This apparently didn’t means a house to fail, a bad solder indicate was, nonetheless it could have contributed to a processor floating adult after a solder disaster as OCP wasn’t enabled on a PSU.
Our disaster caused ASUS to accumulate energy supply manufacturers and their VRM retailer (Iternational Rectifier) and they looked during a new VRM pattern on a X99 Deluxe motherboard for weeks. They went by a pattern from tip to bottom and figured out that they did indeed need to change start-up sequences as a loads weren’t even. Guess what VRM Phase has a top load? Yes, a one that unsuccessful on a house during a shut-down/start-up sequence. It appears that a Corsair AX860i energy supply was one of a energy reserve that had these anomalies and between a PSU pattern and a strange VRM balancing/sequencing on a ASUS X99 Deluxe motherboard were usually right to means a disaster we saw in a field. ASUS has expelled UEFI 1004 that altered a approach a balancing/sequencing is finished on a ASUS X99 Deluxe motherboard. They trust this will forestall any disaster from function to others and we are vehement and happy that ASUS was means to come adult with a fix. ASUS has proven that they will go a stretch to make certain that their play are stable. We are also blissful that we published this article. We schooled a good understanding by doing so and it brought a conditions to light that competence not have been otherwise.
- ASUS X99-Deluxe Motherboard owners should immediately refurbish to UEFI 1004. (No other ASUS Intel X99 house is impacted given a ASUS X99 Deluxe had a singular VRM pattern and firmware setup)
- Corsair AXi Power Supply Users Should Use OC Link Software to capacitate OCP on their PSU if it was finished before March 15th, 2013.
- The ASUS X99-Deluxe Motherboard is protected to buy. We got another ASUS X99-Deluxe house adult and regulating a day this one unsuccessful and it has not failed. We had some DDR4 fortitude issues with memory kits regulating 3200MHz, nonetheless ASUS bound that with new memory tables in UEFI 1004. This house is one of a many entirely looked during Intel X99 play on a marketplace currently and that should move comfort to many. We know many of a readers were watchful on this conditions to finish before shopping a $392 house and we don’t censure you.
This concludes a ongoing coverage of a ASUS X99 Deluxe motherboard failure. We’d like to conclude a readers that stood adult behind us on day one and giggle during those that lashed out for posting this article. Legit Reviews was combined to move a law out and to make a fan village better. This essay is an instance of what drives Legit Reviews and we wish a readers conclude a vision. Also we’d like to conclude a many companies that were concerned in elucidate this issue. You could have left us in a dark, nonetheless we didn’t. You devoted a information and spent time and income to come adult with a resolution to a issue. You don’t get support or trust like that with all companies, nonetheless ASUS, Corsair and Intel are attention leaders for a reason.