Although I posted this on the hardware problem sub-forum I thought it deserved a wider audience as the scale of the potential problem with the black screen issue is I think rather alarming. I hope I am being unduly pessimistic but the failure levels being reported do give cause for concern. This issue has the potential to affect all owners (or potential owners) as it sugests that Acer will need to take major action, even possibly to the extent of a product recall. If this is at all true then as a user group we have the potential to help Acer sort this out! (I am aware that the "fix" that Macles reported seems to have saved a number of people but I am assuming that the vast majority of AspireOne owners will not be aware of this and in ny case possibly not techically savvy enough to carry this out). ---------- I have been doing some calculations based on the recent poll of users reporting the "black screen" problem: viewtopic.php?f=24&t=1392 So far 20 members of the forum have reported this problem (as of 30-Aug). Of these 12 users reported that they had made no modiciation to the memory of their machine which would confirm this is not memory related. Currently this forum has about 4000 members so this implies a failure rate of 20/4000 =0.5% (5000ppm). This obviously assumes one unit per user, although of course there are many members of the forum that have not (yet) purchased the One. There way of course be a few people that have more than 2 or more. Also it ignores the fact that there may be a few people that joined up expressely to participate in the poll. Failures within a day or two of using the Atom to several months. The nature of the problem would indicate that this is generic and is this not an infant mortaility issue (e.g. defective motherboard). From this we should conclude that the failure rate will be constant with time since all machines are equally at risk. If we assume that machines have typically failed after about 1 month of use with a use level of about 2-3 hours per day this implies a mean time to failure of around 100 hrs (ballpark figure). The MTBF (mean time between failures) will then be very roughly the mean time to failure divided by the failure rate: I.e. 100/0.005 = 20,000 hours. This is dreadfully low. Even hard disks which are notorious for failure have MTBFs of several hundred thousand hours. This is at least on order of magnitude below this. It also implies a return rate of approximately 5% within the one year warranty period and 15% for the 3 year extended warranty. Even if Acer release a Bios that sorts this out it will be difficult to get most people to do this at home as it requires a relatively high degree of computer knowledge. These machines are sold to the general public, the vast majority of whom will not have the requisite knowledge. At this failure level Acer should seriously consider whether a general product recall (assuming that they can fix it through a Bios upgrade) would be the best approach. The expected marked for the SCC (small cheap computer) segment is thought to be around 10m pieces through to the end of 2008. Acer is one of the leading players and one ccould expect there to be maybe 3 million Aspire Ones sold worldwide this year. If we make the extrapolation that 5% of users will need service then this equates to 150,000 units requiring customer service during the warranty period. Even if the problem can be fixed through a simple reprogramming of the Bios it will probably still cost the order of $100 per unit to fix (including shipping costs). Acer's potential exposure to this problem is then potentially the order of $15 million dollars. However because of claims through the extended warranty program, failures for other reasons etc their exposure is likely to be much higher, maybe 20-30 M$. The loss of goodwill etc resulting from this issue is difficult to gauge financially but could be even higher than this. This would represent the loss of sales of both the Aspire One but potentially also other products tarnished by this. I know that there are assumptions in this which may turn out not to be valid and it is probably a worst case scenario. However if I was in Acer management then I would be seriously concerned. A drop in income of 20 million USD is a signicant amount. To put it in persepctive Acer announed a 175 million dollar profit for 1H 2008 so potentially they are looking at a 5% drop in income as a result of this (and possibly high looking at lost sales opportunities etc).
You fail to mention that users who have defective computers are overrepresented in the forums. This and other facts make your statistics worse than unreliable, using a poll from an open forum as a base for calculations of this kind just doesn't work. This does not mean that your conclusion is invalid, just that most figures in your post are completely arbitrary. Also your poll seems to be targeted solely at 110 users, since other models ships with an extra ram module preinstalled.
Please don't shoot the messenger! ;-) Thank you for your comment however as I raised this as I hoped it would stimulate a debate on this subject. I thought I had made it clear that the sample is not necessarily representative and as I stated in the post this is simply to arrive at a ball park figure. There are many reaosns why the true failure rate could be higher or lower. It was not intended to arrive at an exact failure rate. Nevertheless I was personally alarmed when the possible MTBF came out at only say 20,000 hours when I think components such as hard disks are typically around 300,000 - 500,000 hours and other components such as LCD displlays possibly even an order of magnitude or so higher than this. It is unusual for consumer products that so many different customers observe the same failure mode and report their experiences in this way. If around 20 out of the 4000 people who have registered on this forum have done so then this suggests to me that this is a serious issue. Of course the majority of those reporting problems are from 110 owners. But the 150 was only relased very recently! As I did state this is a worst case scenario (and I for one hope that the reliability figures aren't as bad as this). However llooking through this forum and forums for other machines such as the Asus EEE and MSI Wind I have not found any issues that would be such a concern. To users (at least those that are unaware of workrounds etc) this would seem a catastrophic failure. Some of the other failures that have occasionaly been reported such as USB ports not working, faulty touchpads and webcams are annoying but still allow the machine to be used to some degree. Only Acer wil know the true level of field returns but even if they were only a fraction of what I have suggested then I believe that they have a significant problem to deal with! A workaround has been posted on this forum (thanks to Macles) but I am not such whether Acer service is aware of this. In any case most users wouldn't be able to sort this out themselves without some assistence. My hope is that as a user community we can possibly be of some assistance in this matter. One question that I have myself is whether the 150 and 110 are identical in terms of their motherboard etc. so that the 150 would also suffer from this problem. Maybe there is some change that means that the newer machines are immune to this problem. However from what I have read of the Intel technical documents there are a number of issues with the Atom processor that are only resolved by code in the BIOS. Since the processors are the same and the BIOS is the same (v0.3114) then this would suggest that the 150 is just as prone to this issue as the 110. A further assumption made implicitly in this but I didn't make clear enough was whether the failure rate really is constant with time. A different situation would be if there are a small fraction of faulty products failing with an infant mortality issue. These would quickly fail and the failure rate after say a month or so of use would be much lower. My feeling however is that this particular problem is not of the infant mortality type. The memory issue seems to me to be an irrelevance as the issue has been observed both in machines with 512 MB (as installed) and those with expanded memory. Also, probably there is no influence of the operating system (as far as I can see this problem has been reported both for machines running Linux as well as Win/XP). My intention was to try and quantify the scale of this in concrete terms (even as an order of magnitude). I think it raises a number of issues for (potential) users. One example might be whether to purchase Acer's extended warranty as this would offer protection for up to 3 years.
so do we know the cause of the problem? and is there any proven EASY way of solve this by myself if it were to happen?
Please see the threads in the hardware problems sub forum. Macles has posted a method for reinstalling the BIOS from a USB stick which seems to have worked for a number of people. http://macles.blogspot.com/2008/08/acer ... overy.html This can reset the machine but will not cure the problem. However there is some evidence that the 0.3114 BIOS is more prone to this problem than v0.3109. It is possible to downgrade back to the older BIOS although I can't vouch for this as I haven't tried it myself. (Machines shipped recently will probably have the newer BIOS but you can see if you go through the start information at POST).
as I read, all people reported success with downgrade of 3109 so far, which is encouraging. but i agree, its a issue acer should respond responsively and quickly.
Since starting this thread on Saturday a further 6 users have reported on the poll in the hardware problem sub forum having experienced this problem. The number grows and grows! Seems like about 2 a day. It strikes me that if (as we believe) this is a problem associated with a buggy Atom chip and the more recent v0.3114 BIOS chip we will see a massive increase in these numbers over the next month or two. Hopefully soon Acer will have a better BIOS that sorts out the problem whilst allowing deep sleep Presumably Intel will have a revision of the Atom on its way soon although of course this won't help those of us that rushed out to buy one of these machines already.
Atom is used in alot of other, if not all, NetBooks, I hope its not a issue related to that. But I guess its possible, when intel tried to lower the power consumption.... maybe something goes black for lack of it... lol We will see.
I just joined this forum because my Aspire One 150 (xp) went black screen yesterday after four days of moderate use. Haven't even had a chance yet to read the BIOS fix posted here, since I first wanted to register my data point on this problem. Yes, indeed, the problem exists on the 150 model too. Even though my Aspire One is only 5 days old, it probably has the older BIOS. I got this machine from a Best Buy store in the U.S. It was the only one left in stock in the entire state (of Indiana), so it probably shipped from Acer to Best Buy a while ago. Many thanks to those responsible for posting the BIOS fix.
I haven't seen reports in other forums for Netbooks using the Atom (EEE 901, MSI Wind) that this is an issue for them. However it does seem to be related to the processor as the failure mode is symptomatic of the processor being unable to wake up from deep sleep. Intel have made a list of the bugs in the Atom (see their technical docs for details) and many of these require BIOS workrounds. My feeling is that the Atom is relatively buggy and Acer (unlike MSI and Asus) haven't managed to get their act together enough to provide a decent fix via their BIOS.
