• 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.

Asus A7V133 detailed impressions (after MSI Pro2A)

L

Leo V

Diamond Member
Some of you have heard about my persistent problems with the MSI K7T Pro2-A motherboard. These problems justified my upgrade to a newer, KT133A motherboard...and I got mine from Asus. If even Asus can't make a solid VIA-based motherboard, then who can?

Before getting my new A7V133 to work, I had to face a few problems.

The first is the fan on the KT133A chipset. I cannot stand fans. I only have a quiet 16dB Adda 2500RPM CPU fan mounted on a copper Hedgehog (to cool the Tbird 1.1 retail), the other quiet Adda fan came inside the QuietPC 300W PSU. I'm using an IBM 75GXP in "quiet mode", inside a QuietPC SilentDrive enclosure. I know that I must have the KT133A running fanless 🙂 After removing the fan, I sanded the KT133A heatsink to perfection. Originally, it uses no thermal substance, and the heatsink surface was even curved! I almost wrecked my right hand, sanding the heatsink from 180-grit all the way to 1800-grit sandpaper! Then I replaced the heatsink, adding a modest amound of thermal compound between it and the KT133A chipset. This solution seems to work just fine, I haven't encountered chipset-related problems since.

Then I encountered a problem because my CPU had a poor thermal compound which prevented proper contact. By the time I realized this, I had factory-formatted my disk, and so had to reinstall everything cleanly.

I attached both my HDD and CDRW onto the Promise ATA controller, disabling the VIA southbridge ATA. This proved useful later. Another significant thing is that I'm using my Midiman soundcard in conjunction with onboard AC97 sound. The latter is much better for games (and has a gameport), the former for music.

The final problem which I've recently solved is stability. After all the pain with random lockups, I wanted nothing short of 100% reliability. Everything looked good at first, until I invented a new level of PC torture. All of these were run simultaneously:

* Huge directory copy over 100mbps LAN (2 simultaneous copy processes)
* CPU burn-in program with error checking
* CoolEdit upsampling to a huge WAV file, CPU and disk-intensive process
* Playing a CD-quality or 88KHz/32-bit (to parasite bandwidth) WAV file over the LAN
* Exact Audio Copy top-quality audio CD ripping/testing to harddisk
* Random IE windows/browsing
* Outlook Express open
* Random explorer windows

Basically, all peripherals were brutally utilized, by multiple apps at that! 🙂 HDD, CDRW, LAN, sound card, memory, video, etc. The lockups didn't fail to show up! I noticed that I'd get freezes whenever sound was playing. After a very long elimination process, I discovered that hard disk access + playing sound = freeze. In other words, a conflict between the Midiman and PromiseATA. Moving the Midiman 2 slots up solved this. I then had to find a suitable location for the nic (it could also cause crashes in the wrong PCI slot).

So my config is:
AGP: Crative GeForce2 MX
PCI: blank, undesirable
PCI: 3com 10/100 nic
PCI: Midiman sound
PCI: blank, undesirable
PCI: blank, undesirable

At last, my torture test could complete successfully, including the CPU burn program which reported no problems. Outlook Express still crashed itself, but that's a Microsoft feature, not hardware.

Despite Asus's suggestion to disable onboard AC97 if you're using a soundcard, I'm using both Midiman and AC97 perfectly well. No stability issues here either. Same with long Quake3 sessions (where AC97 sound is used).

Now, on to config+performance:

I'm using Asus's latest BIOS (used DOS mode "aflash" to upgrade). My 1.1GHz retail copper Tbird is now configured like this:
133MHz FSB, 8x multiplier = 1066MHz CPU and using 1.57Vcore for reduced temperatures. Yes it's a 3% underclock and a 10% undervolt, which reduces CPU heat dissipation from 59watts to a manageable 43watts. The VIO is at 3.26V since I didn't want to raise motherboard temps either.

I did everything with jumpers, because they're persistent across BIOS upgrades; I had trouble software-adjusting voltage anyway.

The performance in some areas really stunned me, as it was a full 25% faster than the MSI Pro2A!! Here are the numbers from Sandra 2001:

Former Pro2A config:
Tbird 1.1GHz retail @default
Two Kingmax 256MB PC133 sticks at 133/CAS3 (the highest the mobo allowed w/interleaving enabled)
IBM 75GXP ATA100 on the VIA ATA100 controller, "fast/loud" settings
Win98SE clean install

Current A7V133 config:
Tbird 1.1GHz retail @1.067GHz, 133MHz FSB, 3% underclocking
Two Kingmax 256MB PC133 sticks at "7ns/PC143" setting (133/cas222/interleaved)
IBM 75GXP ATA100 on the Promise ATA100 controller, "slow/quiet" settings (huge acoustic difference, it's now truly inaudible)
Win98SE clean install

Sandra2001 results, Pro2A in plain and A7V133 in bold:

CPU Benchmark
Integer 3030 2908
FPU 1486 1422
Comments: completely CPU limited, almost identical bang-per-MHz

Multimedia Benchmark
Integer 6051 5920
FPU 7377 7207
Comments: same as above, identical per-MHz scores

Now things get interesting:

Memory Benchmark (MB/sec)
Integer 421 532
FPU 486 609!!!!
Comments: Treading on DDR territory here! 25-26% gain in both tests despite slower CPU speed! In both cases memory ran at 133 with 4-way interleaving! Asus really seem to know their memory tweaks.

Hard Drive Benchmark
Score 18997 23938!!!
Comments: again a ~26% improvement. despite the A7V133 test running in the slower Quiet Mode!! This is of course thanks to the onboard Promise ATA100/RAID controller. From now on, my HDD and CDRW are connected to it and I've disabled VIA's southbridge ATA100.

These are of course theoretical. What's not are my Quake3 scores, which apparently also improved significantly! Using my performance-oriented settings, I used to get ~103FPS demo001 scores, now it's ~136FPS. The difference isn't theoretical, Quake actually feels noticeably smoother for a discriminating FPS gamer like myself 🙂

Raw CPU-dependent calculations haven't improved, of course. But there are significant strides in performance, which alone go far to justify the cost of my upgrade. To me they're a terrific bonus on top of stability.

Final observations, comparison between mobos:

I must say that Asus "feels" like a higher-class motherboard than MSI. Asus's brand premium isn't for nothing. The board came with 2 ATA100 cables, so I'm spoiling my Plextor CDRW with one of them! (It just feels better.) Asus also included an extra USB header with 2 extra ports. The documentation manual is nothing short of excellent quality, with accurate descriptions for everything.

The other characteristics are also pleasant: bootup (it always boots on the first time! No more failures to produce a video signal) feels much faster. The temperature sensors are more accurate than MSI's, reporting painfully realistic CPU readings. DIMM sticks are installed from right to left, which leaves more "airspace" near the CPU. The performance options are outright superior to MSI's, and probably account for much of the speed improvement. Furthermore the Promise controller, which I regarded as a menace, is actually a better choice than VIA's ATA controller.

So I'm quite satisfied with my "forced" upgrade. This does seem like a motherboard I don't mind being stuck with for a relatively long time.
 


<< Outlook Express still crashed itself, but that's a Microsoft feature, not hardware. >>

\

I can just imagine you saying that with a straight face... 😛



Pretty good.
 
😀 Great post Leo V. I'm getting ready to build a new system on the Asus A7V133 MB and this info will be of great help. I haven't built a computer in 3 years so I'm still a rookie at this. I hope you will update us on this A7V133 setup you have.
 
Hey thanks for that 🙂 I especially like the part about Via 686B vs Promise....makes me glad I moved all my hard drives to my promise controller....

I think the A7V133 is going to be my next board as well, thanks 🙂
 
I'm sure that the (266)133 FSB has something to do with the gain in performance, as well 🙂
 


<< Everybody, dump your low class MSI and go for the luxurious Asus now! >>



LOL...no thanks I`m very happy with my MSI board rock solid even with IE5.5 &amp; Win98 😉

🙂
🙂
🙂
 
Mem, I think he was sarcastic. I'll admit that I wasn't the luckiest MSI user. Others had better experiences.

Geekish Thoughts, very likely. But I doubt it accounts for more than 10-15% difference, the rest is probably either memory or the Promise controller.
 
The FSB had a lot to do with the Q3 gains, Promise accounts for the HDD gains, and the memory gains, well its KT133A, and the memory interleaving options.
 
Leo V I know,I sent LXi a PM,notice the wink at the end of my post,that`s me being sarcastic.

🙂
 
Now if wouldn't mind throwing everything back on the 686B and doing a quick comparison for us. Shouldn't take too long. Heh, just kidding 😉

Great work.

Modus
 
I loved my Asus A7V. My wife has the MSI PRO2A and I think I like it as much if not more. I think you just got one of the bad batches. MY wifes is as stable as can be. I like my new Epox better then both of them though. I will definitely miss my A7V though. The past few mb's I had were Asus.
 
I must say that I built 2 systems for friends with the K7T Pro2-A, and they haven't had a single problem yet.
 
Thanks! Here's another update--my CPU is now running on 1.525Volts stable. 1.475V seemed to work, but the Sandra2001 &quot;burn-in&quot; test crashed after ~30 minutes. I wasn't using my &quot;multitasking torture&quot; test for undervolting, because the CPU is the only affected component.

UPDATE: I haven't yet 100% ascertained that 1.525V is error-free, even if stable. Quake3 has once given me an error, dispatching me back to Windoze. I'm investigating further, but might go back to 1.575V.

UPDATE2: Now I have. Sporadic CPU errors are occasionally produced at 1.525V. I'm back at 1.575V where operation is completely normal.
 
Leo, you didn't say anything about temps. what where your temps with the MSI? what are your temps now at 1.575? are you still using the same Hedgehog and fan? is &quot;QuietPC 300W&quot; the name of the PSU?

thanks for a great quiet review. I've also completed my project on how to make a very Quiet keyboard and it works.

Nic
 
Quiet keyboard? Could you tell us? 🙂

&quot;Leo, you didn't say anything about temps. what where your temps with the MSI? what are your temps now at 1.575? are you still using the same Hedgehog and fan? is &quot;QuietPC 300W&quot; the name of the PSU?&quot;

I'm still using the Hedgehog + 2500RPM quiet fan. However, I've just ordered a &quot;new&quot; Alpha PEP66H aluminum heatsink with inset copper heatplate (see link). And some Arctic Silver II.

Running at 1.1GHz @1.65V with this config, MSI would report ~44C and Asus reports ~62C. The heatsink is the same temperature, so I'm confident the A7V133 readings are simply more realistic. Now that I'm running 1.066GHz @1.575V, the Asus reports ~55C which is pretty good.

Link for QuietPC PSU (yes I'm using the 300W one.)
 
My quiet keyboard. I have an Internet Keyboard Pro, but microsoft really skimped on the lube. For the price we paid for it, they really did a half-ass lube job on this one. After three weeks the keys were already stiff and had become loud to the type.

My magic ingredient - Teflon Lube Gel, &quot;Teflon® offers excellent adhesion and coating characteristics. Lubricant provides a thin, dry coating with exceptionally low surface friction. Product lubricates plastic, glass, metal and wood. The non-gumming formula repels water, oil, dirt and other contaminants.&quot;

There are different types of Teflon lube, I'm not home right now so I don't remember the one I have but it's in a very small bottle about 3-4 oz. It's a thick clear gel. I can put pictures of the bottle on my website in a few days.

I took all the keys off, cleaned the base of the interior of the keyboard to remove the dirt and stuff that accumulated over time. I then applied a good coat of Teflon lube gel on the root of the keys themselves, not too much but I didn't skimp like microsoft either. I applied the lube and re-installed the keys. I applied just a little more on the bigger alt and crtl keys. I did all the keys one by one and that was it.

In the days following the lube job, I found that the keys were just a little bit stiffer than before, It just wasn't the same but I think that was to be expected. After about a week the lube settled in and the keys were much softer and very much quiet, the sound level is barely noticeable. It's been three months now since I have lubed the keys and I'm surprised how quiet and soft the keys still are. I'd recommend to anyone who wants a quieter keyboard but also give it about a week or two before making your final judgment on the softness of the keys because the lube has to settle in. Good Luck.
 
Nice!! 😀 I'm glad this board worked out better then your Pro2A. I remember reading that thread, you were not happy! 😉

After seeing this the A7V133 will definitely be on my too buy list. Thanks Leo. 🙂
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top