Our title comes from an email I received today. It is a very good question and one that certainly deserves an answer. Well, let's start off with Anand and Derek attending GDC 2009 this week so they are consumed with meetings and even more meetings. Hopefully we will hear from them shortly, if not, we will send a process server out to find them.

Jarred is busy working on a slew of notebook articles and will have a review up on an interesting unit from Dell in the very near future. I have motherboards and memory kits stacked in such a way that it has created an elaborate maze between test rooms. In fact, without a GPS device for navigation, I sometimes feel like an axe-wielding Jack Torrance wandering the hallways. The majority of these products are still running a variety of tests, some for the fourth or fifth time to be honest, which might explain why I have an axe and ulcer.

Besides a rouge power supply that took delight in slowly damaging some of the latest and greatest products on the test bench before it went up in a blaze of glory, we have had an interesting time with our AMD AM3 motherboards. We tested several BIOS releases, provided our input, and then tested what came back to us. This should have been a quick process based on our agreements with the suppliers and fairly mature chipsets.   As it turns out, the performance differences between BIOS releases have been all over the map. It has been a case of plug one leak and then find two more a day later instead of just fixing a particular set of problems. Although the boards have been extremely solid at stock settings, when pushed they have exhibited more personalities than Sybil.

This is not to slight the products or scare monger away potential customers. We absolutely love the latest AM3 products from MSI, Gigabyte, and ASRock along with their updated AM2+ lineups. Testing started a few weeks ago with our motherboards from ASUS for a series of articles ranging from CrossFireX performance with the ultra hot and chic Phenom II X3 720BE, a DDR2 versus DDR3 memory performance comparison, enhancing platform performance via increases in Northbridge speed, and of course the actual motherboard reviews.

The good news is that the latest beta BIOS' have solved the vast majority of problems, it just took sometime to get to this point. The one interesting aspect is that all of the motherboards at one time or another suffered from similar problems. This usually does not occur and it lead to some interesting discoveries and then solutions on behalf of the BIOS teams. At one point I joked with a particular supplier if all of the BIOS engineers from the various motherboard companies met at the same Internet Cafe in Taipei during evening hours to discuss ways to ensure my total hair loss by May.

It was just strange as each 790FX AM3 board eventually suffered the same problem. Everything from the boards hard locking if HT reference clocks were set over 214 with CrossFireX active to various problems clocking Northbridge speeds up and poor memory performance just to name a few. To everyone's credit, the problems have been identified and are being solved or have been solved. We are at the point now where the last few remaining problems will mainly affect those who utilize Dry Ice or LN2 for overclocking, otherwise, the boards are in or will be in excellent shape within a week as updated BIOS' are released. We will discuss this further in an upcoming article, but a major factor for some of the quirky behavior between BIOS releases had to do with updated AGESA code from AMD during and after the AM3 launch.

All that said, we are at the point now where our results at least make sense and are repeatable. Later today we will publish the 720BE CrossFire article and in a couple of days we will have the DDR2/DDR3 comparison on the AMD platform. In April we will start rolling out the motherboard reviews for the 790FX/790GX AM3 boards along with a really good AM2+ 780G product update from ASRock that impressed us and should be in stock shortly.

In the meantime, we are expecting our first retail Core i7 D0 stepping processors in a few days and will provide an immediate update on a product that has been generating a lot of forum buzz the past few weeks. Speaking of the i7, we will discuss the ASRock X58 motherboard shortly as it is running four GTX-295 cards without a problem and our long term stability tests indicate this is a very solid product from a newcomer in this market. Also on the calendar in April is our second X58 motherboard roundup with a focus on the motherboards priced under $250. Raja will provide results for the more extreme crowd with the latest enthusiast boards from EVGA, Foxconn, and ASUS.

We have a mini-ITX roundup that is almost finished with chipsets from both Intel and AMD plus a few other surprises later on for the HTPC and SFF audience. Right now we have to say the smaller and more nimble companies like Zotac and J&W have really impressed us with their latest products.

With our AM3 memory performance problems behind us, we can finish up testing several DDR3 kits for a large roundup early next month. In fact, our entire focus on the article changed midway through testing as the latest DDR3-1333 kits are providing the best balance of price and performance at this time.

Last but not least, we are knee deep in testing our first workstation products from Supermicro and will have products from HP and others in the future. We really appreciate all of the comments and suggestions from our last blog about the subject matter. Since that time we have hooked up with several suppliers for both hardware and software that is workstation centric. I will go into this in more detail shortly, but we have a variety of ATI FirePro video cards from AMD (before anyone brings it up, we asked NVIDIA to also participate with their latest Quadro series and our requests were declined), Corsair provided an excellent 24GB DDR3 kit, RAID controllers from HighPoint and Areca, enterprise drives from Western Digital, Seagate, and Fujitsu along with software from Autodesk to MainConcept to SolidWorks just to name a few items.

It is going to be a very busy April for us and I only touched on about half of what is going on at this point. I just know it is an exciting time on this end and hopefully our extended coverage in several areas will be exciting for you also. If not, be sure to let us know. In the meantime, I have an article to finish editing and an axe to sharpen for someone if another BIOS shows up today.

Comments Locked

29 Comments

View All Comments

  • Penti - Monday, March 30, 2009 - link

    They are sourcing the same BIOS probably thus the same problems.

    There are only two makers of bios any way, but I guess both could have the same problems as it's a new platform. Normally the manufacturers shouldn't have to do much. Nice to see problems ironed out any way. But I really think it should be the manufacturers job, hopefully Award/Phoenix and AMI learned something too.
  • Penti - Monday, March 30, 2009 - link

    Shouldn't be.
  • ChuckR - Saturday, March 28, 2009 - link

    You mentioned HP being tested. By any chance would this be the new
    HP Pavilion Elite m9600 and m9600t series? I have had good luck with HP and am unable to build my own. It is to bad the specs on pre-builts don't give PSU, Fans, Case, Type of DDR3 for I7, Motherboard, etc.
  • strikeback03 - Sunday, March 29, 2009 - link

    Most likely because those components are subject to change based on who can offer the lowest price at the moment. The motherboard and case might stay the same, but certainly PSU, RAM, and most of all fans are going to be changing based on price.
  • MadBoris - Saturday, March 28, 2009 - link

    One thing for sure is I notice I only need to make monthly visits.

    I remember when it used to be daily visits, but if I come here weekly it seems too often now. I can come monthly and not have much reading to catch up with at all. I don't know how that keeps page hits coming at all.

    Like another poster said, I also see a lot of times how articles/reviews are mentioned then just disappear completely. There was something about Linux almost a year ago, etc.

    Anyway, I used to like coming every several days, but unfortunately you guys don't produce enough anymore which has to hurt page hits.

    I still like the quality and insight so I keep coming back.
  • Gannon - Friday, March 27, 2009 - link

    I'd like to see older generation cards (8800 GTS 640MB, 8800 GTX) always included in reviews of newer generation cards. Many people don't upgrade to the latest and greatest every year. Truth be told the GT200 and 4870 were both underwhelming parts compared to anyone that still owns an 8800 (or their rebadged versions 9800 gtx+, etc, etc)

    Also including benchmarks of a wider selection of older games would be nice, not all of us play Crysis. I'd like to see supreme commander added back into the list, Quake wars and even the older Doom 3.

    I really dislike many review sites "cheating" their users of this kind of historical comparison. Also it would reveal driver cheating much easier if you were to use older games to see if GPU manufacturers are targetting performance for specific games while neglecting others.
  • Minion4Hire - Saturday, March 28, 2009 - link

    I've seen many review sites claim that Supreme Commander is a very poor choice for benchmarking due to performance consistency issues. That could be one of the reasons it's no longer used.
  • Gannon - Sunday, March 29, 2009 - link

    Wherever you heard that it is incorrect... I'm sorry to tell you.

    http://enthusiast.hardocp.com/article.html?art=MTM...">http://enthusiast.hardocp.com/article.html?art=MTM...

    The fact is some games are very intensive and we'd like to know just how little a video card upgrade would have on them.

    I like to see a bit more diversity of applications, instead of a small amount of highly selective games.
  • dlweir - Friday, March 27, 2009 - link

    Gary:

    The reason you are having so much trouble lately with BIOS releases is simply that the ABIT engineers dispersed throughout the industry.
  • Serup - Friday, March 27, 2009 - link

    I indeed appreciate the info in blogs like this - i think it gives me better perspective as to how vendors work when it comes to launching products. I must say, that i'm surprised and disapointed about the way they send half-finished products to test, so you can help then iron out the last bugs.

    This blog-post, the info anand shared in a recent mobo review, about why everything takes so much time to complete, and the article about ssd's and the vertex-drive where he tells about the communication with ocz to try to make them focus on other things than sequential throughput are the reasons as to why i rate this site my #1 "source for hardware analysis and news" as you put it.

    I really think it's a great idea to share whats going on behind the scene, instead of just shead out big chunks of stats that show me that X mobo can oc 0,5% more than Y mobo, and that everything scales perfect in all benchmarks with these 0,5%, and therefore it gets the gold award.

    I think it's bad if everything comes down to who's product can get the best result in some synthetic benchmark, rather than stability, real world performance and value for money.

    (sorry for my spelling - english is my second language)

Log in

Don't have an account? Sign up now