GA-880GA-UD3H graphics somewhat disappointing?

[KEP]

Okay, got Windows 7 x64 running, /w 8gb RAM, on this system now.  And, I dunno, I guess I expected to see a bit better score reported by the "Windows Experience Index" on the built-in mobo graphics.  I pretty much expected I'd need/want to add an expansion card for that, but still sorta thought that given the spec I might not feel it compelling at least right away.  Anyway, these are the numbers I got...

Graphics	ATI Radeon HD 4250	4.5
Gaming graphics	3323 MB Total available graphics memory	5.4

Is that pretty much what to have expected from this board?  (The Aero score was particularly surprising.)  I guess the follow-up question is, if I were to try speeding that up is this Crossfire thing the way to go?  (I'm not a gamer.)

Thanks...

[absic]

Hi,
First off, don't get to hung up on the Windows Experience Rating it really doesn't give you much help. Secondly, the figures you have for the GPU seem about right for an on-board graphics solution.

The thing to consider is what do you want from your PC? If it is working fine and doing all of the things that you want it to do is there any point in throwing money at it? Or, do you just want some BIG figures to give you bragging rights with your mates?

Adding a dedicated Graphic Card would improve your WE score but, will you actually gain anything from this except a hole in your pocket where the money used to be? If you do add a GPU you would actually do better to disable the On-board Graphics as the WE rating is based on the lowest score and, even if you put it in crossfire, that score won't improve.

[collidingstar]

I got the same scores..

What really got me was the score on the Memory..

But I ended up getting a HIS 5670 and turned off the build in graphics..

Not just for better graphics performance but I didn't want to loose half a gig for the shared memory too..

Anyway back to my point.. as soon as I did this my Memory score jumped..

Now I know this is more of an Index then an actual speed/performance test.. but still..

Anyway I got all 7 + with the only exception being my drives..  which I would need to upgrade to a solid state before that will move.

[absic]

Anyway I got all 7 + with the only exception being my drives..  which I would need to upgrade to a solid state before that will move.

That's the same scenario for me. Even WD SATA3 HDD's didn't improve the WE score above 5.9 Funny thing is, even in a RAID0 array the HDD score didn't go above 5.9 in the WE which I really did think would be different especially as I am getting above 200MB/S Read and over 160MB/s Write speeds with the RAID0 Array.

[KEP]

First off, don't get to hung up on the Windows Experience Rating it really doesn't give you much help. Secondly, the figures you have for the GPU seem about right for an on-board graphics solution.

Yeah, I know it's just an index, but over time I've been able to correlate a good bit of "what to expect" performance-wise into certain ranges.

And, please note I did use a question mark in my Subject.  I wasn't sure whether this was low/med/high for this particular board.  I guess I kinda figured that given the 512mb dedicated, it might've been a tad higher.  But what I was really after was whether this was as good as I should expect with this board.

The thing to consider is what do you want from your PC? If it is working fine and doing all of the things that you want it to do is there any point in throwing money at it? Or, do you just want some BIG figures to give you bragging rights with your mates?

Well, the latter is always fun, sure.  But it's not the main point, by any means.  I have worked extensively on machines that scored 4.6/4.7 on the Graphics index, and definitely found their performance to be lacking (especially with multi-monitor setups).  Windows don't glide along as they're dragged, videos hiccup, and so on.  I'd certainly like to avoid that kind of thing with this new box, and just thought I'd give the mobo a fair shake before plunking down even more money on another card.

Adding a dedicated Graphic Card would improve your WE score but, will you actually gain anything from this except a hole in your pocket where the money used to be? If you do add a GPU you would actually do better to disable the On-board Graphics as the WE rating is based on the lowest score and, even if you put it in crossfire, that score won't improve.

I saw in a later post that you added one in, and it brought you up over 7.  Did you feel the difference with "everyday" activities? 

I really don't know much about Crossfire.  I guess it seems like "a gamer thing"?  So I don't know where I'd get the most bang for the buck.  Whether to look for (this will be a purchase of opportunity if/when it happens) a totally standalone card, or one that works in conjunction with the main board.

To be sure, it's not about the score as much as it is the experience.  I'm a value shopper all the way.

[absic]

A dedicated Graphics Card will give you better results, as you would expect and yes, I did notice an improvement over using an on-board Graphics chip.

You can buy one of the 5 series ATI Graphics Cards quite cheap these days and most of these will give you good results and DirectX 11 too. A 5450 can be picked up for under £40 so they don't really break the bank either.

[KEP]

I got the same scores..

What really got me was the score on the Memory..

But I ended up getting a HIS 5670 and turned off the build in graphics..

Not just for better graphics performance but I didn't want to loose half a gig for the shared memory too..

Anyway back to my point.. as soon as I did this my Memory score jumped..

Now I know this is more of an Index then an actual speed/performance test.. but still..

Anyway I got all 7 + with the only exception being my drives..  which I would need to upgrade to a solid state before that will move.

That's interesting!  So the shared memory caused normal RAM operations to be slower?  I have 8gb of RAM, so I didn't mind the sharing yet, but if that's throttling everything else I definitely need to reconsider.

Not sure what it is with 5.9s on disk.  I get the same with RAID1 (fast disks, but still SATA2), which certainly ought to be less than the RAID0/SATA3 setup absic mentioned.  I gotta run some actual benchmarks on that, to see where it's really shaking out.

[KEP]

A dedicated Graphics Card will give you better results, as you would expect and yes, I did notice an improvement over using an on-board Graphics chip.

You can buy one of the 5 series ATI Graphics Cards quite cheap these days and most of these will give you good results and DirectX 11 too. A 5450 can be picked up for under £40 so they don't really break the bank either.

Yeah, I just put an XFX Radeon 5450 into my kids' machine, for actually under USD30.  Looks like it's $5 more now...

So, I take it you're not all that interested in the Crossfire thing, then?  Why is that?  Is it because it is mainly a gamer thing, or some other flaw?

[Dark Mantis]

That's the problem the WEI  is more of a nuisance than it was meant to be and half the time I think it just guesses the rating from the details of the hardware not actual testing. It is definitely not accurate by any stretch of the imagination.

[absic]

I don't have any problem with Crossfire and I actually think it works better than NVidia's SLI.

The thing is I don't recommend using crossfire with on-board graphics as it will always be pulled down to the lowest common denominator, the speed at which the On-board chip runs. This is also the reason that the WE rating can vary so much when you alter just one thing. If you put a dedicated GPU into your system, you will be releasing the RAM to do number crunching without recourse to the graphics chip. Hence the numbers will show an increase even though your RAM is the same.

Crossfire with 2 cards of the same spec is fine, I actually have a 4850X2 (which is 2 cards on one board) so I have the option to run in Crossfire mode anytime I want and, as I have another 4850 (that I snuck past my wife) I can actually run in triple Crossfire Mode which is great for Crysis and other games. Most of the time though, I don't run in Crossfire mode as I don't do much gaming.  If you run Crossfire with 2 cards of different specs it will always pull the speed down to the lower spec'd card.

[Wonderwrench]

That's the problem the WEI  is more of a nuisance than it was meant to be and half the time I think it just guesses the rating from the details of the hardware not actual testing. It is definitely not accurate by any stretch of the imagination.

Sure its accurate! It accurately measures nothing! Seriously I think it is measuring performance in each given area but with some preset limits that make little sense. Take the disk scores. It's  impossible for a magnetic drive to score over 5.9. A run of the mill WD Blue drive gets a 5.9. Two WD Raptors in RAID 0 get a 5.9 WTH. In the end I think people myself included make to much of a fuss over benchmark scores. The most we need to know, is it working as designed and does it feel right. If so your done! I do save a screen shot of the WEI results after a fresh build so I can compare the scores at a later date. Might come in handy?

I also run Passmark PerformanceTest on each new build and save the results. See screen shot with my personal builds over the last 12 years. Its amazing how far we have come. Notice the 2d power of the ATI 9700. Seems all newer generations cards from both ATI and Nvidia suck at 2d. Also note the huge loss of 2d performance between the C2D and C2Q. This was caused by the move from XP to W7. Some results show as incomplete because the test was not available in PerformanceTest at that time. Some hardware is also not ID'd, not sure why though. Kinda cool Passmark keeps backward compatibility with older versions so you can see how the old stuff stacks up to today's stuff.

http://i156.photobucket.com/albums/t25/wonderwrench/Misc/PerfRes.png
Note click on the image to zoom in if needed.

Bill

[Dark Mantis]

Sure its accurate! It accurately measures nothing!

Absolutely

I did find looking back through your history of ratings very interesting as you say there are some scores there that you wouldn't expect and some that were very obvious.

[collidingstar]

That's interesting!  So the shared memory caused normal RAM operations to be slower?  I have 8gb of RAM, so I didn't mind the sharing yet, but if that's throttling everything else I definitely need to reconsider.

Not sure what it is with 5.9s on disk.  I get the same with RAID1 (fast disks, but still SATA2), which certainly ought to be less than the RAID0/SATA3 setup absic mentioned.  I gotta run some actual benchmarks on that, to see where it's really shaking out.

Not sure if it throttled it.. But the Index changed..

My main reason was to not loose the 1/2 a gig..
I really didn't notice too many issues with the graphics but I didn't push it too much either..

[KEP]

Anyway I got all 7 + with the only exception being my drives..  which I would need to upgrade to a solid state before that will move.

That's the same scenario for me. Even WD SATA3 HDD's didn't improve the WE score above 5.9 Funny thing is, even in a RAID0 array the HDD score didn't go above 5.9 in the WE which I really did think would be different especially as I am getting above 200MB/S Read and over 160MB/s Write speeds with the RAID0 Array.

Well, believe it or not, I nudged the needle on that one!  Darn near doubled the throughput on my disks, and bumped the disk score up from 5.9 to an ever-so-apparently-boastful 6.1! Attached are the RAID1 and RAID10 benchmarks.  Nice of Windows to notice, huh?

[Dark Mantis]

Congratulations! Definitely a significant increase.   I didn't think your scores were bad for RAID1 anyway but they are certainly healthier now.