ASRock 939Dual-SATA2 - Socket 939 with PCI-E and real AGP and Socket AM2 upgrade path

[nemesismk2]

Originally posted by: niggles

Originally posted by: nemesismk2

3dmark 2005 doesn't really have much to do with game benchmarking because it's a synthetic benchmark which tests dx9 video cards. Which part of your computer were you hoping to benchmark with 3dmark 2005?

The CPU/Mobo/vid card. Why would there be tests specificallty for the CPU in 3dmark05 if it wasn't testing the CPU/bus as well as the vid card?

Only the game tests have any real effect on the final score because 3dmark 2005 is a video card benchmark. Yeah it does have a single cpu test but as far as I know it doesn't add anything to the final score.

 

[sandorski]

Originally posted by: Woody
I used to have a A7N8X NForce2 with a 3200XP and scored around 6000 with my AGP X800Pro in 3DMark05. With this new ASRock board and a 3700 I scored 6212. Actually I was surprised the score went up that much since 3DMark05 almost completely ignores CPU speed and the results are almost entirely graphics card dependent. My guess is that the AGP interface on the ASRock is a little more efficient than my old ASUS NF2 board.

My review on this board:

I built the system over the weekend and played plenty of Battlefield2 on Sunday.

Overall this board is an excellent value. At $69 I can't complain about anything and there were no surprises. The board is exactly what I expected and went together with no difficulties. A couple things lacking like a PSU fan monitor plug and Firewire. Firewire cards are cheap and I think my Audigy2 has one built in. It works with older 20 pin PSUs that have the 4 pin extra plug.

Performance with the 3700 is troublefree and on par with NF4 systems. AGP performance is very good and without doupt the best performing and most advanced AGP motherboard you can buy at any price for Socket 939.

If you have a good AGP graphics card that you don't want to give up and you feel your CPU is dragging you down you can pick up one of these boards along with any Socket 939 Athlon64 including the dual core models. You are basically getting a new CPU with a $69 surcharge that you can install with all the rest of your existing components. It's definately the way to go.

The PCIe upgrade slot means you can then upgrade your graphics card down the road at your leisure. It also is the only motherboard I know of that includes a future CPU expansion slot for next generation AMD CPUs. The board allows for simultaneous use of both AGP and PCIe graphics cards (not in SLI).

I picked up a Maxtor DiamondMax10 SATA for my boot drive and am using the Coolermaster Aquagate mini for CPU cooling. I set up a 240GB RAID0 with a pair of 120GB 7200.8 Barracuda drives I already had from a couple years ago. The SATA RAID0 performance is considerably better on this board over my old ASUS A7N8X and BF2 maps load nearly instantly. I created a special partition on the RAID for games so I can keep them defragged easily....yes I will be in the map faster than you and I'm gonna come steal your chopper! ;-)

I'm a little concerned that the CPU temp is pretty high.....gets up to around 60C under load and idles in the high 40s. I'm wondering if the water pump block isn't properly seated or the thermal paste didn't spread evenly. The temps are actually normall for Venice cores with this cooler (based on reviews I've read on the Coolermaster) and I would assume the San Diego would naturally run hotter but this is the same CPU that a friend of mine used for overclocking his system back in June. He had much lower temps (40s under load) using air cooling in the hot Kansas summer. I'm going to try a bigger fan on the heat exchanger since I'm using a slower, quieter, and prettier one than what came with the Aquagate mini. I really don't want to take the darned thing apart and reinstall the cooler but I don't want to fry the CPU either, it's the most expensive component in my system now. Possibly just my temp sensor is not calibrated too accurately since the air coming out the heat exchager doesn't seem as warm as I would expect for those temps. The case temp stays nicely cool and is well vented.

I would like to hear what others are getting for temps with this board.

X2 3800 running at stock speeds and using stock HS/fan 42c at full load both cores running BOINC projects.

 

[niggles]

Originally posted by: nemesismk2

Originally posted by: niggles

Originally posted by: nemesismk2

3dmark 2005 doesn't really have much to do with game benchmarking because it's a synthetic benchmark which tests dx9 video cards. Which part of your computer were you hoping to benchmark with 3dmark 2005?

The CPU/Mobo/vid card. Why would there be tests specificallty for the CPU in 3dmark05 if it wasn't testing the CPU/bus as well as the vid card?

Only the game tests have any real effect on the final score because 3dmark 2005 is a video card benchmark. Yeah it does have a single cpu test but as far as I know it doesn't add anything to the final score.

Seriously? Why bother waisting our time testing hardware and not applying it to the final score. It is a valid thing to test for, your CPU, RAM and bus are just as likely to affect your overall game performance. This is why I've never questioned the testing.

 

[Fam Money]

Woody, I would definitely reseat that water block. Those temps are pretty high for an A64.

 

[nemesismk2]

Originally posted by: niggles

Originally posted by: nemesismk2

Originally posted by: niggles

Originally posted by: nemesismk2

3dmark 2005 doesn't really have much to do with game benchmarking because it's a synthetic benchmark which tests dx9 video cards. Which part of your computer were you hoping to benchmark with 3dmark 2005?

The CPU/Mobo/vid card. Why would there be tests specificallty for the CPU in 3dmark05 if it wasn't testing the CPU/bus as well as the vid card?

Only the game tests have any real effect on the final score because 3dmark 2005 is a video card benchmark. Yeah it does have a single cpu test but as far as I know it doesn't add anything to the final score.

Seriously? Why bother waisting our time testing hardware and not applying it to the final score. It is a valid thing to test for, your CPU, RAM and bus are just as likely to affect your overall game performance. This is why I've never questioned the testing.

I agree with you and at the very least the cpu score should be added to the final score. However Futuremark do not see it that way because they want people to purchase pcmark 2005 as well for benchmarking their pc.

 

[imported_Woody]

On 3DMark05.....I don't agree with either of you on this. This is a video card benchmarking utility, that's all. It's not intended to be used to measure CPU performance in any way and if it was it would be useless for what it is supposed to be: An application for comparing video cards directly regardless of platform.

This is extremely useful for comparing relative performance of video cards, drivers, and the effects of overclocking video cards. The great thing about it is you can compare your vid card performance directly with reviews or your friends who may be running on entirely different systems. It's not intended to measure or predict real world gaming performance except on identical systems where all other components are equal.

If you want to measure your overall system performance or specific component performance such as memory bandwidth or CPU capability specific to your application there are a slew of benchmarks to choose from designed for that purpose. Keep in mind that while your high end overclocked Athlon64 may offer superior performance in Battlefield2 than a high end overclocked Pentium4 you will probably get stomped by that same Pentium4 when measuring performance in video encoding.

In short, use the right tool for the right job!

On overclocking: This is a hobby. If you can afford to go out and spend $4000 on components and overclock them to some ungodly speed and be the fastest man on the planet then Kudos to you. If you can only afford a $69 ASRock and a Venice 3000 and get it to perform as fast as an FX-55 running on a DFI Lanparty NF4 board than Kudos to you too. It's all good.

Personally I'm more impressed and satisfied with my older AGP X800Pro running at 6200 in '05 than a brand new stock 7800GTX running at whatever. I feel like I've accomplished something cool and good. A little bit of rebel that makes us strive to beat the system and get more performance from something that wasn't meant to be, regardless of how much we spent on it.

On the cold boot issues: Update the BIOS to 1.20 or better. Mine shipped with BIOS 1.20 and there are no cold boot issues. My board has been perfectly stable so far.

Thanks for the tips on my watercooler....I am going to cut the grill out of the back of my case behind the heat exchanger and put the high speed fan that came with it on to see if that makes a difference.....if temps don't drop significantly I will assume that I seated the water block badly and will tear it down and do it over.

Great thread....keep it going!

 

[niggles]

Woody, have you not noticed that 3Dmark tests the CPU? This is why I thought it was taking in to account your CPU etc into your final score.
As to testing only video would you please explain in what scenario would your CPU, bus and memory not affect your gaming experience? It sounds like you're saying that 3Dmark only tests the video card and nothing else. How would it not test everything else, it doesn't work without taking into account everything else.
Please know this is not an attack, I am simply trying to understand how this works and why it the test would not be affected by my bus.

 

[Diasper]

I wonder has anyone tried this combo yet:

Opteron 144
Asrock 939
X800 GTO2

Ultimate performance/price goodness with upgradeability?

 

[imported_Woody]

Originally posted by: niggles
It sounds like you're saying that 3Dmark only tests the video card and nothing else.

Hello? Yes! this is what I said. It does a fine job of this without being affected by the rest of the system significantly.

Like I said. I went from 6060 to 6216 in 3DMark05 with my X800Pro after putting it into a brand new ASRock 939/ A64 3700 combination from a ASUS NForc2 / 3200XP system. My gaming performance has improved significantly due to a wide variety of factors but my 3DMark stayed very much the same.

The 156 point increase could be attributed to a more efficient AGP bus on the new board, improved performance of the Cat 5.10 drivers over the 5.8 drivers I had previously tested with, or more likely, a combination of both. The better CPU, memory controller, HTT, etc. had little or no effect on the score.

So you ask why this is useful since it means nothing. It's not useful, at least not for what you were trying to evaluate. You should not be using this application to determine the overall performance of your rig.

If you would like to compare the absolute performance of your video card to mine, this is the application to use. If you want to measure the effects of a driver modification, overclocking your video card, different resolutions, and various other graphics card settings, this is the application to use.

"The right tool for the right job."

 

[touchmyichi]

Originally posted by: Diasper
I wonder has anyone tried this combo yet:

Opteron 144
Asrock 939
X800 GTO2

Ultimate performance/price goodness with upgradeability?

got yourself a winner there

 

[Diasper]

Originally posted by: touchmyichi

Originally posted by: Diasper
I wonder has anyone tried this combo yet:

Opteron 144
Asrock 939
X800 GTO2

Ultimate performance/price goodness with upgradeability?

got yourself a winner there

I guess the only thing is that only holds so long as the retailers don't price gauge like crazy.

In that case alternative would be:

Venice 3000+
Asrock 939
7800GT


Again the same setup would allow for future upgrading say to dual-core or the next gen (or even one after if they're not all PCI-E 2) of graphics cards (eg 32 pipes & 600mhz+ core etc ...course I could be dreaming...)

 

[nemesismk2]

Originally posted by: Woody

Originally posted by: niggles
It sounds like you're saying that 3Dmark only tests the video card and nothing else.

Hello? Yes! this is what I said. It does a fine job of this without being affected by the rest of the system significantly.

Like I said. I went from 6060 to 6216 in 3DMark05 with my X800Pro after putting it into a brand new ASRock 939/ A64 3700 combination from a ASUS NForc2 / 3200XP system. My gaming performance has improved significantly due to a wide variety of factors but my 3DMark stayed very much the same.

The 156 point increase could be attributed to a more efficient AGP bus on the new board, improved performance of the Cat 5.10 drivers over the 5.8 drivers I had previously tested with, or more likely, a combination of both. The better CPU, memory controller, HTT, etc. had little or no effect on the score.

So you ask why this is useful since it means nothing. It's not useful, at least not for what you were trying to evaluate. You should not be using this application to determine the overall performance of your rig.

If you would like to compare the absolute performance of your video card to mine, this is the application to use. If you want to measure the effects of a driver modification, overclocking your video card, different resolutions, and various other graphics card settings, this is the application to use.

"The right tool for the right job."

3dmark 2003 and 3dmark 2005 are video card benchmarks so do you know why Futuremark added cpu tests to a video card benchmark? I emailed them about it and never did get a reply

 

[BlingBlingArsch]

Is the Assrock Sata2 a troublemaker? I read some threads where ppl had complaints that it wouldnt run from the start and 60 bucks doesnt indicate that its a high quality board? Any idea about the death rate?

 

[imported_Woody]

Originally posted by: nemesismk23dmark 2003 and 3dmark 2005 are video card benchmarks so do you know why Futuremark added cpu tests to a video card benchmark? I emailed them about it and never did get a reply

I don't know. I've used the free version only and wasn't even aware of the CPU test. I would speculate that the test is there specifically to negate the effects of the CPU on the final results since clearly CPU performance has almost no effect on the final score. I assume the technical problem of creating a graphics card benchmark that is capable of negating the effects of the CPU would require the application to be able to measure those effects since clearly CPU performance has a huge impact on most applications and games that utilize the graphics card.

That is just speculation, however. All I can say for certain is the facts that can be directly observed.

Probably Futuremark will reply to the query with a recommendation to use PCMark to benchmark overall system performance.

 

[niggles]

Originally posted by: Woody

Originally posted by: nemesismk23dmark 2003 and 3dmark 2005 are video card benchmarks so do you know why Futuremark added cpu tests to a video card benchmark? I emailed them about it and never did get a reply

I don't know. I've used the free version only and wasn't even aware of the CPU test. I would speculate that the test is there specifically to negate the effects of the CPU on the final results since clearly CPU performance has almost no effect on the final score. I assume the technical problem of creating a graphics card benchmark that is capable of negating the effects of the CPU would require the application to be able to measure those effects since clearly CPU performance has a huge impact on most applications and games that utilize the graphics card.

That is just speculation, however. All I can say for certain is the facts that can be directly observed.

Probably Futuremark will reply to the query with a recommendation to use PCMark to benchmark overall system performance.

Arrrrg!!! I asked you the same question, you simply clipped it out of the quote you posted back to and ignored it. There is clearly a section where the CPU is being tested. It's been tested in the 03 version as well. If it was not an issue on the overall testing I can't see why it would be tested. If it wasn't an issue why are the specifics of the overall rig listed in the project manager.

Nemesis, THANK YOU for actually reading my posts.

Woody, The feeling is that the weight of 3dmark is on the GPU. Everything since has been about why the CPU is tested and how the rig over all could be ignored. I can see that you are trying to be helpful so thank you, I just want to understand what the point of the CPU test. It's in the free version so you'll see it. I think it's the third or fourth test in.

 

[deadmouse]

Hi everyone, i am new to these forums.

What brought me here was the discussion on the Asrock 939 Dual-Sata2 mobo. I have just recently purchased this board, along with an AMD 3200+ Venice. Full system specs listed below. I have been interested in slightly overclocking this board, but so far, i have had no luck getting my settings to take effect. I have a basic understanding of how overclocking is done, but after i make the changes (which i think are right, but not sure) in the bios, after logging into windows, CPU-Z shows my speed to still be 2.0ghz. I hate to jump right into the middle of this thread, but so far, my research has yielded me nothing. Can anyone email me willing to help me out? It would be greatly appreciated, you guys seem to know your stuff around here.

Specs:
3200+ AMD Venice (runs about 36c normally, 46c under load, stock cooling)
Asrock 939 Sata2 (bios updated to v1.30)
1 gig of ram ValueBrand
160 gig hard drive
ATI Radeon X800 GT
Ultra X-Connect PS 400 watt

 

[imported_Woody]

deadmouse: check out this thread in the CPU/Processors forum....

http://forums.anandtech.com/messageview...atid=28&threadid=1497607&enterthread=y

Overclocking the ASRock is pretty much the same as overclocking other 939 boards.

 

[deadmouse]

Thank you Woody. I had browsed the forums for a bit, but couldn't find what i needed. Just wish i had seen this before i asked... Thanks again!

 

[nemesismk2]

Originally posted by: BlingBlingArsch
Is the Assrock Sata2 a troublemaker? I read some threads where ppl had complaints that it wouldnt run from the start and 60 bucks doesnt indicate that its a high quality board? Any idea about the death rate?

I can only comment from my own personal experience and I have not had any serious problems. I have heard some people blame the motherboard for things which are not connected to the motherboard at all. :shocked:

1. A good example is people not doing a fresh installation of windows and then complaining about stability problems, when upgrading your motherboard it's always a good idea to do a fresh installation of windows.

2. Another example is people installing video card drivers before they have installed the uli integrated motherboard drivers. I don't know why people are doing this because you have to install the motherboard drivers before the video card drivers with all motherboards.

 

[iamTux]

I see contradictory information about the compatibility of Opteron (Venus) processors with the 939Dual-SATA2 board. Some people say it works, some say it works until BIOS rev. 1.20, some say it doesn't. Do you have any reliable info on this?

 

[Nesta]

I got one of these coming in the mail coupled with a 3200+

Anyone running Win XP x64??

 

[deadmouse]

Originally posted by: Nesta
I got one of these coming in the mail coupled with a 3200+

Anyone running Win XP x64??

I was going to install Xp 64 but I heard alot of programs have trouble running under this OS. So i just stuck with XP Pro. I have both, maybe I can give it a shot, but I just formatted when i got this new mobo.

 

[imported_Woody]

Originally posted by: Nesta
I got one of these coming in the mail coupled with a 3200+

Anyone running Win XP x64??

Pretty much what deadmouse said. If you have it try it...you can always have dual installs and run XP 32 on a different drive or partition.

 

[FirNaTine]

To BlingBlingArsch: I got this board as soon as it was available form newegg. The only problem I ran into was that I had to bump up the vdimm to stop memory errors on my 2x1gig ocz platinum. Other than that it has been perfectlly stable. My board came with BIOS 1.1 and I updated to 1.2. The only difference being tha availablilty of OC features.

 

[sandorski]

Originally posted by: Nesta
I got one of these coming in the mail coupled with a 3200+

Anyone running Win XP x64??

I'm running WinXP 64. Haven't really used it long enough to know whether I like it or not, but so far so good.