Dissapointed with many of the latest reviews.

Grminalac

Golden Member
Aug 25, 2000
1,149
1
0
I have been a devout reader of this site for some time; I have browsed it at least twice a week since the whole celeron 300A overclocking phenom back in the day. I have trusted the hard efforts of the writers and viewed their opinion seriously, but lately I have had second thoughts as to the validity and accuracy of many of their reviews. One glaring example is the Apollo Pro 133A roundup. Basically this review, as well as other reviews of similar motherboards is pointless, entertaining yes, but not a very technical write up. The reason being is that memory bandwith on chipsets running the 133A vary wildy based on tweaking, and I don't mean using H.Oda either; just plain old options available in the bios. Take my motherboard for instance, the Gigabyte GA-6VX-4X, the board is a total dog, but turn on the top performance setting in the bios and mystically the memory bandwith jumps from 332 to 383 as measured in Sisoft Sandra. Set the memory to Cas3 and ouch. I'd rather not even say how low the bandwith drops. Sure displaying key bios setting in reviews would take more time, but you need to ask yourself whats more important: showing pertinant usable information once or twice a week or posting garbage once a day. As a wannabe tech myself i would like to be able to duplicate some of your benchmarks, i'd like to see exactly how my motherboard performs against the ones you tested. Unfortunetly I am not given the variables i need to accomplish this. I have noticed others complaining about the heatsink roundup as well as the linux article. Quite frankly I foolhardily purchased a golden orb thanks to your article, only to later read a contradicting article on toms hardware stating that the intel stock heatsink i replaced it with was a good deal better. Don't get me wrong I have my qualms with toms hardware as well and i'm not saying their review was any more accurate.
I don't mean to be insult any of the hardworking individuals of this site, but at the same time i don't want to see some of the last few good sites out there transformed into what I like to refer to as "tabloid" websites. Posting quantity not quality for the sake of ratings. Most would be "good information sites" are evolving from "PBS" into FOX prime time. We see this all the time: PC world type articles that confuse the public and provide "fuzzy numbers" i suppose for some sites lining their pockets and keeping website hits up is a good thing, but I still believe that deep down the the writers at anadtech providing good journalism is most important.
Thanks for reading this and any thoughtful feedback is welcome.
 

GL

Diamond Member
Oct 9, 1999
4,547
0
0
As an owner of a CUV4X, I sure wish all hardware sites (not only this one) would at least use the standard H. Oda tweaks that are floating around...they really do help quite a bit.

-GL
 

Grminalac

Golden Member
Aug 25, 2000
1,149
1
0
Exactly. Those sites are supposedly aimed towards people with technical knowhow, so when doing a review at least set the machine up as a pc enthusiast would. I am not talking about insane overclocking, just tweak it within reason.
Geez if i see another celeron review of a apollo pro 133 with the fsb at 66 and the memory set to cas3. I am gonna scream. especially when they say "Well the memory benchmark is very low on the celeron" for the love of god its the motherboard not the processor providing the bad marks, at least try it out on a BX board or something. We have users out there that believe the celeron inherently causes "slow memory" as i have heard them say, but what they don't realize is the fact that ANY processor running at 66mhz on a Apollo pro 133 at cas 3 is gonna provide DISMAL benchmarks.
Then again I wouldn't buy a celeron.
 

GL

Diamond Member
Oct 9, 1999
4,547
0
0
I have an old Celeron 433 that I had on an Asus P2L97. I recently upgraded to an Asus CUV4X, with 256MB PC133 CAS 2 SDRAM. My FSB is at 80MHz, so I have a Celeron 520MHz. The RAM is at 2/3 so I have it running at 120MHz. With the simple tweaks available at viahardware.com I'm pulling bandwidth benchmark numbers from Sisoft Sandra that are close to a PIII 1GHz for the ALU, and exceed the PIII 1GHz for the FPU. I have noticed a considerable (read: noticeable without benchmarking) improvement in performance. I don't know...but I'm quite happy with the memory performance of my CUV4X...if people would only take the time to follow a few simple steps outlined at viahardware.com...

-GL
 

Grminalac

Golden Member
Aug 25, 2000
1,149
1
0
Yes, using the tweaks, i have my system happily running 152 CAS2 with 402 mb/s bandwith cpu. If the user has enough technical savvy to purchase a motherboard and build the system they certainly have enough ability to use some well documemted tweaks.
The gigabyte motherboard i have actually has a top performance setting which toggles many of the values you find in the guide.
From what I have seen most apollo pro 133A motherboards run similar to each other performance wise, with the same tweaks applied.
 

Grminalac

Golden Member
Aug 25, 2000
1,149
1
0
Just gonna bump it up for a little more feedback. I figured I might as well since I typed so much.
 

Grminalac

Golden Member
Aug 25, 2000
1,149
1
0
Yes, the linux article is a dissapointment as well. It mentions nothing about the memory settings of the apollo pro, that board could vary anywhere between 300-380, that would definetly affect scores in quake arena.