• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Barton 2500+ Results vs. the Grand Daddy of 2100+ T-bred B's

chizow

Diamond Member
Just got my 2500+ Retail from Googlegear, looks to be a decent stepping and date, but I was hoping for something week 10 or later, here it is:

AXDA2500DKV4D 9596378250523
AQUCA 0307 XPEW

Hmmm...retail PIB are no fun, reminds me how much I hate whomever invented blister packs.

Anyways, gonna give it a go on my Rev 2.0 A7N8X-DLX and see how far I can push the FSB. I'm shooting for at least 2300MHz @ 200MHz FSB, but I'll try push the FSB if I successfully hotflash my 1.04's BIOS. Then I'll set up my oldie but goodie 2100+ on the 1.04 and see if I can run some head to head benches at 200MHz FSB and equal clockspeeds.

Chiz

 
Lookin' forward to your results. With the Barton's starting to come down to a very reasonable price I'm kicking around buying one to play with myself, hence, if the results are promising then it'll be helpful in my decision making process.
 
Here's another vote - I'd love to see the results. Trying to decide if Barton 2500 @$138 or T-Bred 2100 @$90 for my A7N8X-DLX that I recently ordered........ go chiz go... 🙂
 
Originally posted by: Duvie
I'm kicking around buying one to play with myself

Don't do that I heard you will go blind!!!!😉🙂:beer:
Could you pick a bigger font? I can hardly see what you typed
😱:Q😉
 
Hmmm....preliminary results not so hot.

The CPU needs A LOT of juice to run stable at 2.3GHz. I'm running at 1.825V (1.875V w/ Asus overvolt) to get it stable...so far. Anything over 2100MHz and less than 1.775V and prime fails the first test!!!

Its definitely not going to be faster than my T-bred B in clockspeed, the only determination will be if the L2 can make up for the lower clockspeed. I was running 11.5x207MHz = 2380MHz with my T-bred B. Right now I'm at 11.5x202MHz = 2323MHz. I'm not even completely sure it will be stable here. I had to tweak a lot of settings just for it to get past the first test in Prime and to finish a loop of UT2K3 demo and 3DMark2K3.

I'm not going to give up on it yet...temps are high now; I'll let the AS3 settle in tonite and I'll loop Prime and 3DMark overnight and see where it goes. I may also reapply the AS3, as the core is at 49C under load.

General windows, websurfing, loading, menus, .dlls etc. do seem snappier as others have noted, but it just doesn't feel stable yet......
Also noticed that UT2K3 demo seemed to run smoother overall. Gonna need to test further and get the 2100+ back up and running.

Chiz
 
Originally posted by: chizow
Hmmm....preliminary results not so hot.

The CPU needs A LOT of juice to run stable at 2.3GHz. I'm running at 1.825V (1.875V w/ Asus overvolt) to get it stable...so far. Anything over 2100MHz and less than 1.775V and prime fails the first test!!!

Its definitely not going to be faster than my T-bred B in clockspeed, the only determination will be if the L2 can make up for the lower clockspeed. I was running 11.5x207MHz = 2380MHz with my T-bred B. Right now I'm at 11.5x202MHz = 2323MHz. I'm not even completely sure it will be stable here. I had to tweak a lot of settings just for it to get past the first test in Prime and to finish a loop of UT2K3 demo and 3DMark2K3.

I'm not going to give up on it yet...temps are high now; I'll let the AS3 settle in tonite and I'll loop Prime and 3DMark overnight and see where it goes. I may also reapply the AS3, as the core is at 49C under load.

General windows, websurfing, loading, menus, .dlls etc. do seem snappier as others have noted, but it just doesn't feel stable yet......
Also noticed that UT2K3 demo seemed to run smoother overall. Gonna need to test further and get the 2100+ back up and running.

Chiz

My first efforts at OCing a Barton 2500+ all Primed out in short order with anything above 2200mhz. After about 48 hours of running, the same settings that had problems ran fine, passing prime95 without errors for overnight runs. Now I'm at a point where 2300mhz is that magical total mhz number for stability, regardless of fsb/multiplier options. I am currently at 11.5 x 200 @ 1.80v (or maybe it is 1.775) with memory at sync with timings of 6-2-2-2. I can POST and run windows up to 12x205 (maybe higher) but it primes out or eventually locks up even with up to 1.85v (my personal vcore upper limit). Lower FSB and relaxed memory timings have similar results.

BTW, after playing around more, I am not convinced that the extra cache on the Barton equals more than about 50mhz performance. My 2100+ at 2360 and this Barton 2500+ at 2300 seem very similar in performance and get very similar benchmarks in Sandra, for example. Obviously real performance will vary from application to application.

Good luck chiz.

 
Thanks for the encouragement guys :beer:😀

Well, there was a rough point last night where I couldn't even get it to POST at 2300MHz, so I had to drop it down to 2230MHz. Ran Prime95 all night and woke up this morning with it still running and notched it up to 2300MHz again before I left for work.

Cowdog, thanks for the feedback, good info bro. Seems I'm seeing the same thing you did. Getting this chip to decent clocks was a PITA at first, but its already more stable after 1 night of torture. There are still some odd issues relating to power draw and the BIOS. Exiting the BIOS seems to be much slower than with my 2100+, which to me has been a sign of instability in the past. Also, it seems at high FSB and high clocks weird things occur like floppy errors and checksum errors. Only way to prevent these errors is going into the BIOS and disabling boot up floppy seek and boot from other device, which is another sign of instability. These errors don't happen with a high FSB but a moderate total clockspeed.

I've also got a ton of stuff running off my PSU, so I may try disconnecting some of them to see if it helps stability. Even though Prime95 and other torture tests/benches will run without issue, the rig just doesn't feel rock solid stable yet.

One weird thing I noticed was that the sound was cutting out on PowerVCR II after 20 seconds or so; Going into settings or recording would reset the driver and the sound would come back on, but cut out again after @ 20 seconds again. I'll try reinstalling it and see if it goes away, but its not something that happened with my 2100+.

I've yet to upgrade to the latest BIOS from Asus' site; I'll be doing that after I hotflash the BIOS for my non-functional Rev 1.04. I'm not sure there's any difference between the BIOS my rev. 2.0 shipped with and the one on Asus' site tho (same BIOS revision codes), although the dates are different by over a month.

I'll keep playing with it this weekend and run some benchmarks vs. the 2100+ to see which is the better performer once I find the max stable clock of the Barton. Right now, as the published reviews indicate, the Barton seems smoother in desktop/Windows apps and in games even though its running ~150MHz slower than my T-bred B and scores slightly lower on a few gaming/graphics benchmarks.

If I can clock it to 2323 again and get it running stable there, I think it'll be a keeper, but heat and voltage requirements are definitely the limiting factors with this chip.

Chiz
 
as I posted in another thread:

with non-hardcore cooling (vantec aeroflow c7040, and the thermal compound that came with it) and the voltage at 1.700 I got 180x12.5 to be very stable. core temp was 47-48 under load (prime95) and everything felt rock solid. this is with A7N8X rev 1.04 and the 1003 BIOS. Processor is from week 8. I cant do higher FSBs since my memory won't cooperate - running 2x512MB TwinMOS PC2700 CAS2.5 (running at 2-2-3-7).

I've stepped down to 180x11 though since noise is a big issue (using computer for music production) and I can now run the CPU fan at 50% using SpeedFan even under extreme load without passing 50 degrees CPU temp.
 
Nice....I'd be happy with my Barton if I could get it to run 2250MHz. Anyways, got home from work and found Prime had errored out after 6.5 hours so back to the drawing board. I'm gonna put a Zalman passive on the South Bridge as I'm pretty sure that's the reason for the sound cutting out from my A2. The MCP-T runs hot on the Rev 2.0 also; the first time I checked it, it was cool to the touch. I'm also gonna try a different BIOS as this shipping version seems a bit flakey above 200MHz.

Chiz
 
I have a 2500+.
For the 1st 6 days it wouldn't OC at all really, anything over a 185MHz fsb and it would hard lock.
After that though it took off.
The limit to the cpu seems to be around 2250 MHz, then it gets flacky regardless of voltage.
At 2200 Mhz however it runs great and solid as a rock and at a 1.75V.

My CPU
Wcpuid 2204MHz
200 Mhz SiSoft memory scores. 3033/2828

The system is really snappy, coupled with some WD SE 80gb drives the system flies.

Hey while your already looking , this is my new HTPC mod.
HTPC from a standard Apex 500W DVD player.
 
Originally posted by: chizow
Nice....I'd be happy with my Barton if I could get it to run 2250MHz. Anyways, got home from work and found Prime had errored out after 6.5 hours so back to the drawing board. I'm gonna put a Zalman passive on the South Bridge as I'm pretty sure that's the reason for the sound cutting out from my A2. The MCP-T runs hot on the Rev 2.0 also; the first time I checked it, it was cool to the touch. I'm also gonna try a different BIOS as this shipping version seems a bit flakey above 200MHz.

Chiz


Just wondering.. how you know it errored out after 6.5 hours? Mine errors out but i have no idea when....
 
Originally posted by: The_Lurker
Just wondering.. how you know it errored out after 6.5 hours? Mine errors out but i have no idea when....
I run it in a window on my 2nd monitor while doing other things to stress it a bit. You know b/c it says "Expecting XYZ, returned ZYX" or something to that extent. It then tells you how long it has been running. If you are crashing out of Prime or it is rebooting your computer, that is VERY bad and is on the "Houston, We have a problem" end of the stability spectrum.

Anyways, I haven't had much time this weekend to do any more testing or comparisons; I totally blew off getting my other A7N8X up and running (I have til Saturday before my bro gets back from vacation and comes looking for his rig) this weekend. I basically set it to 11 x 206MHz = 2266MHz @ 1.85V as that was the highest I could reliably run w/out a reboot or prime error. I'm done tweaking now, as there is a reason for me to stop tweaking and start fragging....got into the PlanetSide beta on Saturday and haven't stopped playing it since I got home on Sunday. The game is schweeeeeeeet!!!! I wasn't sure if I was going to keep the Barton, but now I don't care...even the 30 minutes it would take to swap back to my old 2100+ is too much downtime to bother. 😀

Chiz
 
Just curious to know how a 2.266 Barton compares to a 2.4GHz T-bred in gaming tests. Have a 2100+ @ 2.41Ghz currently, and been thinking about the bartons since the price drop.
 
Originally posted by: eBauer
Just curious to know how a 2.266 Barton compares to a 2.4GHz T-bred in gaming tests. Have a 2100+ @ 2.41Ghz currently, and been thinking about the bartons since the price drop.
From my preliminary tests, it would be very close and game specific. Running a 2380MHz T-bred B vs. a Barton at 2232MHz (early stability tests) yielded similar results in UT2K3 benches, with the slight edge to the Barton. The Barton however, seemed to run smoother. In 3DMark2K3, the T-bred B scored higher in the CPU test. I don't remember the overall score, as they are unimportant in 3DMark2K3 when comparing CPU performance, I just wanted to get a relative idea of how a "gaming" bench rated the CPU performance. I also ran the Splinter Cell benchmark, and the results showed very little difference as the game is highly GPU bottlenecked at the resolutions I run (1280 native LCD). These are based on previous vs. current results; I have my bros new rig set up with my old A7N8X and 2100+, but he has a Ti4200, which wouldn't make a good comparison vs. my 9700pro. I'm not going to be doing more testing though, as I'm simply too busy playing PS Beta 😀 to deal with shifting cards back and forth and potential ATi vs. nVidia driver issues. :|

If you can clock a Barton to the same speed as the T-bred B, it will be faster or at the very least equal in every application. It also feels faster in general Windows and office apps, which means it'll feel faster most of the time you use your computer if you focus on games and general apps. Unless you have some serious exotic cooling and are willing to perform a voltage mod however, I don't think its worth it to get a Barton over a T-bred B, especially since you're clocked at 2.41GHz. This CPU requires more juice and runs hotter than T-breds to run at equal clockspeeds, and doesn't seem to have as high a fault tolerance as the T-bred Bs. I've found it much more difficult to OC my Barton; at 2400MHz+ my T-bred B would remain stable with my fan blasting at max and I could boot and remain stable in most anything other than Prime up to 2500MHz. My Barton will reset or lock at anything beyond 2350MHz.

Chiz
 
It has been noted over @HOCP (article linked to another site) today that the Bartons perform better after a significant burn in period. Maybe yours is still too "tight" Chizow. Hope for more mhz to come!
 
Originally posted by: Tambora
It has been noted over @HOCP (article linked to another site) today that the Bartons perform better after a significant burn in period. Maybe yours is still too "tight" Chizow. Hope for more mhz to come!

I'm hoping that's the case, as I pick up my 2500+ tomorrow... 😉
 
Back
Top