Jump to content

ORTHOS Discussion Thread!


raju

Recommended Posts

HI Guys,

 

As you all know I to am an Orthos fan for testing system stability. Having used a great deal of boards over the last few months on this new Conroe platform and seeing a few erratic results on the way, talking to people within the industry etc;

 

I am asking the question to anyone who knows the intricate workings or code of the Orthos utility.

 

This post is not to question the validity of Orhtos for stability on newer platforms, but to ask if it is always universally compatible with all newer release processors and chipsets.

 

My own take has been that it is, and any discrepancy to date has been due to an immature bios/hardware setup.

 

However I am aware that in the past to accommodate newer chipsets and cpu's updates have been made.

 

I use Orthos as a tool for my own database too (based heavily on the OCDB here), but I do run my thinking past a larger community who relies on the software heavily too.

 

If there are any amendments to be made to the tool, the drive will come from the larger community.

 

My concern is that if there is an issue and we have failed to pick it up to date, much newer hardware will be judged poorly unfairly. Granted the Infinity stood alone in the Orthos dept after some stiff work from DFI.

 

I would expect at least some educated posts here and no flaming or anti orthos stuff.

 

Is there need for a periodic update, are there any clued up coders in the house that know?

 

I have no vested interests in an Orthos denouncement, least of all a flame war or dissection of my/our own stability methods. That's all been covered in the past, it does not need a resurface.

 

My question is how do we know an update is needed, has there been any known issues rectified in the past?

 

Right now I would love to commission someone to put together a whole new utility containing the full monty of tests in one neat package. It would require a lot of time money and work and a lot of input from experts too. Something I could never do on my own..

 

Counter argument is that Orhtos is here free and available, but I know we're never closed to the changes that hardware brings upon coding requirements.

 

In theory Orthos is just math, so there should be no problem, but I don't know that for sure as I don't know how the code detects an error per-se at the software level, and whether or not that is affected by chipset architecture.

 

Please think through your responses before making them, it's an important issue for us all, I believe. I stand on the fence tbh, I am open to the discussion and educated reasons one way or the other.

 

The floor is all yours, please keep it civil, I have no desire whatsoever to bring all OUR(I am a firm member here no doubts) testing into disrepute.

 

If the admin/mods wants to close the thread and take it private between us I am completely happy with that and offer an apology if I have made a mistake in any public suggestion. It's a friendly and educated debate on my part, and the results will help stabilize a lot of our future testing and knowledge..

 

regards

Raja

Share this post


Link to post
Share on other sites

  • Replies 46
  • Created
  • Last Reply

Top Posters In This Topic

Kudos to raju for formulating an excellent post.

 

Like he said, no flaming or flame-baiting posts will be tolerated. This is a technical post first and foremost.

 

Keep the posts to the technical merits only. Opinions are welcome only to move the thread forward.

 

ExRoadie

Share this post


Link to post
Share on other sites

Here's a though I had. If someone that can consistently get errors with Orthos can run two instances of Prime95. It would be interesting to see if the errors are repeatable with Prime.

Share this post


Link to post
Share on other sites

I will be working with the new prime version this coming week and make comparisons like praz suggested. ATM there's just so much variance across different boards. I have been told that the old winchester cores needed a prime update before they would run prime effectively, I am left wondering if that should be the case everytime a new chipsetRcpu is released?

 

Maybe we should try and question one o the coders directly and see what kind of reponse we get?

 

regards

raja

Share this post


Link to post
Share on other sites

Maybe we should try and question one o the coders directly and see what kind of reponse we get?

Carl (soundx98) did just that. He sent an email to Johnny Lee. I'm sure we'll here something as soon as Carl receives a reply.

Share this post


Link to post
Share on other sites

since orthos is nothing but a GUI for Prime95 (granted it is made by someone other than the Prime95 people) then what normally happens is Mersienne or whoever makes the real Prime95 will update the program for any problems found within newer processors (they did for the original Winchester A64 cpu I think).

 

Then Johnny Lee will usually take the Prime95 source code (freely available) and update his programs.

Share this post


Link to post
Share on other sites

That's why I was questioning about running Prime. Everyone seems to be using Orthos. Was curious if something was happening to the core program when Johnny was adapting it to his interface.

Share this post


Link to post
Share on other sites

Sheesh you guys are real movers, hopefully we can get some conclusive comments from one of the coders about how updates may be necessary before a new chipset or cpu is considered prime ready. I think that worldwide as the whole stability scene has progressed we owe it to ourselves and the manufacturers to make sure that our testing has not been left to any chance at all. Company reputations are at stake, so fair judging is a must.

 

I have always viewed it as math, but the fact that we are playing with newer chipsets now gives me slight concern. Primarily because I have access to a lot of forums and some repect from a lot of people (mutual), I'd at least like to think my results and the db we have here and back home is completley beyond question each time we get something new.

 

thx for the help guys, I posed this question on 3 sites and so far the best reponses have come from here, I have one other person working with me at another forum too, I am happy that there are some real selfless people left in this heavily marketed industry, who wanna do things the right way.

 

regards

raja

Share this post


Link to post
Share on other sites

That's why I was questioning about running Prime. Everyone seems to be using Orthos. Was curious if something was happening to the core program when Johnny was adapting it to his interface.

 

sometimes adding bits and pieces onto a program can have unexpected results (like bugs). Johnny Lee is aware of such and in the past has had to 'refine' SP2004/Orthos a bit.

 

The latest versions that we are all using these days are essentially 'bug free' on all the latest cpu's up to the Core2. As far as I can tell, the Core2 doesn't have any issues either, but with the new 4-way cores coming and having different architectures and memory controller functions (how it splits the RAM use between all the cores instead of a true SMP setup where each cpu has it's own physical RAM), we won't know until those cpu's really hit mainstream.

 

I've yet to find a bug in Prime95 nor Orthos nor SP2004 up to this point (other than my own buggy memory/voltage settings haha). This in no way means it's foolproof ;)

Share this post


Link to post
Share on other sites

ps: we move & shake because stability is the most important thing we preach around here...if we can't have a reliable, honest program to test overall stability, then we are lost as an overclock is useless if it only is stable until you open a program or a game will only run 19 minutes before crashing etc.

Share this post


Link to post
Share on other sites

I myself have yet to find any definate issues with the 975 chipset.

 

I think the issue here is that we never clear oursleves completley with coders before we begin testing on a new chipset or cpu (me anyway). It's just a case of making sure we are absolutley right before we drop results. I had never really shown the concern before tbh, but knowing now that there have been issues in the past, I guess it'd be wise to have some form of clearance first.

 

Might be a bit of a pain, but it'll save from issues in the long run.

 

In the meantime there is a version of prime now that recognises hyperthreading etc. Over the next week I will cross compare with Orthos.

 

Just awaiting a form of confermation now from Johnny Lee or one of the prime coders to see if they need to look into or revise anthing for newer chipsets too..

 

 

regards

raja

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now

×
×
  • Create New...