Nahkri wrote:

kanecvr wrote:

ALi Aladdin V vs MVP3 - MVP3's AGP performance is marginally better and it's less buggy then the ALi chipset witch gives me headaches constantly

For me is the other way around,ali aladdin V is the chipset i prefer,the mvp3 on 2 different boards as soon as i enable dma in control panel in win98se,the system starts randomly locking up,sometimes even before it boots into windows,i tried different drivers no luck.
The ali mb has absolutly no problems of this sort,plus it has support for udma 66 compared with the udma 33 on the mvp3.

One of my Ali boards (Commate) does the same thing - locks up randomy in windows or while booting. Sometimes it will display Gate A20 Error while trying to boot. The other one (EPOX) bsods with most AGP cards. Poor overclocker too. Can't set any om my CPUs 50MHz higher - it won't post. All my MVP3's OC my K6-2's from 450 to 500 without issues or voltage increase. At 2.4v one of them even runs at 550 stable.

kanecvr wrote:

philscomputerlab wrote:

RDRAM wrote:

The Epox is a hell of a motherboard, it has its own VDIM VRM, the EliteGroup's an ordinary board. But in 2015, i wont think much about performance, but RAM size , 1.5gb vs 5 12mb..., Apollo 133a was a great chip with bad reviews..., another story will be if you had an Epox BX7+100 !!

Well the BX7+ is 440BX based and only does AGP2x so I think the Apollo would eat it alive... I have a Soyo 440BX based socket 370 board, and It doesn't support CPUs faster than 750MHz...

There were no 440BX based socket 370 Soyo mobos, probably you had a 440ZX (cutted chip) it was a SY-6IZA a very budget board for Mendozinos http://www.motherboard.cz/mb/soyo/SY-6IZA.htm , very cheap board - chipset to kill K6III
BX7+ or CUBX are much better boards than most Apollo or Solano boards, believe me I've had some homemade test on CUBX vs CUSL2 , both @100 2-2-2 , and the old 440BX had always a decent edge, the chipset has faster inner timings. I dont give much for agp 4x performance,but the really bad thing's BX boards are only AGP@3.3v (not 1.5v) so you're limited. Greatting for Rumania, I still have my DACIA 1410 !

A quick update. I tried another 815 board that I had stashed away. It's an Acer board, from an Acer Veriton system. You might be thinking it has no chance against AOpen or Gigabyte, but it turns out it is faster than both of them!

So not quite sure what's going on, but with all three boards on the tightest memory timings, the Acer pumps out higher frames than the other boards.

A few years ago I remember being impressed by the memory transfer speed that was measured in memtest86 with a VIA Pro133A (694X) board. It appeared on par with the 440BX. VIA's prior chipset, the plain Pro133 (694A I think), had been much slower. I know that's a simplistic measurement, and I never ran any detailed benchmarking between the chipsets, but it gave me the impression that VIA really turned a corner with their SDRAM performance on the 694X.

Game benchmark, 3dmark and everest all are faster on my asus tusl2-c compared to asus tuv4x... The 815 is consistantly 5% faster (not a big deal but it is faster).

With Graphics cards i found the 815 board to be much more stable.... i have 2 of these boards and they are the best s370 imho.

meljor wrote:

Game benchmark, 3dmark and everest all are faster on my asus tusl2-c compared to asus tuv4x... The 815 is consistantly 5% faster (not a big deal but it is faster).

With Graphics cards i found the 815 board to be much more stable.... i have 2 of these boards and they are the best s370 imho.

i have the same opinion, 694 is around 5% slower than the 815 with same timings, give both are optimized. a less optimum 694 can be even slower.

I just hate the memory limit on the i815. I think Intel treated it like a Celeron chipset, designed cynically for people who wouldn't submit to their RDRAM vision. But depending on the intended usage, sometimes 512MB is plenty.
I have a 694X machine set up right now which definitely does have problems with a much later bridged AGP->Express 7600GS card. In some games it gets a bunch of artifacts with that card. It doesn't happen with everything, which makes me think it's triggered by some particular function or whatever. I've seen the card work correctly on another motherboard, but that chipset and bridge don't get along.
It's actually kind of a silly combo, I just wanted to stretch that 694X machine to it's limit and see what things I could make it run, and how some newer games would react to it.
I haven't tried that card on an i815, but I wouldn't be surprised if it works better.

When i815e was reviewed they mostly used 128mb ram and that was a very nice amount of ram back then (anandtech for example). Back when i had my athlon xp 3200+ system i had 512mb for the first time iirc.
Only for vista and up you really need more ram imho (big jump to 2gb or more).

When you run software that actually uses more than 512mb ram you are probably much better off anyway with a system that is faster than a tualatin.

People are complaining about the ram limit of intel socket7 systems (64mb cacheable) while 16mb was a lot back then.... same thing. The normal ram limit of 98se of 512mb? same thing imho... it was an INSANE amount of ram back then and wasn't even affordable more most people.

Those review sites have tended to be skimpy on RAM IMO. They like to run a single gamer enthusiast module with aggressive timing.
I had 256MB in a low end Coppermine laptop back then, and that amount of RAM for it wasn't even expensive. My desktop was probably still at 384MB. Thinking through life events helps me to pin it down, and I think I upgraded the desktop to 512MB in Fall 2002. Granted that was more than most people at the time but it wasn't hugely expensive either.

At that period of time I regarded a PC's upper RAM limit to be one of it's most important specs for long term usefulness. It seemed to me that the RAM upgrade limit was typically what limited the usable life of most machines as they got old. However, I'll admit that obsolescence pattern started to fade as you get beyond 512MB.
Many people were okay with the 512MB limit, but many also complained about it almost since the chipset was new, really. It was a disappointing step back from the 440BX, which ironically only supports less dense RAM but still allows more of it in total. The 815 can be maxed out with a single module, which really smacks of deliberate crippling by Intel.

People are complaining about the ram limit of intel socket7 systems (64mb cacheable) while 16mb was a lot back then.

I don't think 16MB was a lot, it was average when those systems were brand new. Over-running the 64MB cacheable limit was pretty common on those machines as their owners upgraded them.

810 has the same RAM limitation. People didn't really talk much about 810 though because it doesn't have AGP, probably again deliberate gimping to get people on RDRAM.

The interesting thing I noticed with 815 is that it will recognize more than 512MB if you install more, but of course it can't actually use it.

meljor wrote:

When i815e was reviewed they mostly used 128mb ram and that was a very nice amount of ram back then (anandtech for example). Bac […]
Show full quote

When i815e was reviewed they mostly used 128mb ram and that was a very nice amount of ram back then (anandtech for example). Back when i had my athlon xp 3200+ system i had 512mb for the first time iirc.
Only for vista and up you really need more ram imho (big jump to 2gb or more).

When you run software that actually uses more than 512mb ram you are probably much better off anyway with a system that is faster than a tualatin.

People are complaining about the ram limit of intel socket7 systems (64mb cacheable) while 16mb was a lot back then.... same thing. The normal ram limit of 98se of 512mb? same thing imho... it was an INSANE amount of ram back then and wasn't even affordable more most people.

I 100% agree with you on this. Everyone is so keen on "maxing" things out 😀

I personally find that 128MB is more than enough for today's use of Windows 98SE (not multitasking like you did back then).
Only if someone wants to add a lot of startup programs (FTP, ISO managers etc etc) and wants to play late 9x games (about 2000-2001) may actually benefit from 256MB.

Personally, on my 98SE systems, I just put 256MB and forget them.
32MB for DOS and 64MB for 95C.

shamino wrote:

Those review sites have tended to be skimpy on RAM IMO. They like to run a single gamer enthusiast module with aggressive timin […]
Show full quote

Those review sites have tended to be skimpy on RAM IMO. They like to run a single gamer enthusiast module with aggressive timing.
I had 256MB in a low end Coppermine laptop back then, and that amount of RAM for it wasn't even expensive. My desktop was probably still at 384MB. Thinking through life events helps me to pin it down, and I think I upgraded the desktop to 512MB in Fall 2002. Granted that was more than most people at the time but it wasn't hugely expensive either.

At that period of time I regarded a PC's upper RAM limit to be one of it's most important specs for long term usefulness. It seemed to me that the RAM upgrade limit was typically what limited the usable life of most machines as they got old. However, I'll admit that obsolescence pattern started to fade as you get beyond 512MB.
Many people were okay with the 512MB limit, but many also complained about it almost since the chipset was new, really. It was a disappointing step back from the 440BX, which ironically only supports less dense RAM but still allows more of it in total. The 815 can be maxed out with a single module, which really smacks of deliberate crippling by Intel.

People are complaining about the ram limit of intel socket7 systems (64mb cacheable) while 16mb was a lot back then.

I don't think 16MB was a lot, it was average when those systems were brand new. Over-running the 64MB cacheable limit was pretty common on those machines as their owners upgraded them.

I don't think so. I have magazines here from the time they sold new pentium systems. High end mmx systems usually were sold with 16mb. When later upgraded to 32mb you could run all games extremely well (memory wise). When you needed more ram the cpu wasn't fast enough either... the same goes for the k6,p2/p3 and tualatin. It sure is nice that we can max these old systems out for very little money but it doesn't mean they run much faster or use much newer software enjoyable... imo

AFAIK for gaming it made little sense to upgrade the RAM beyond 64MB on a Pentium/PMMX, but for everyday usage, like the Internet, it was actually rather common. My Cyrix (which was not much newer than an MMX) got upgraded to well over 128MB back in the day, and it performed nicely for web browsing and such. If it weren't for me wanting to play Warcraft III it would probably not have been replaced by a Duron.

I remember having 128MB in 2000 (using the 810 chipset...) and when I installed XP, then I quickly went to 256MB, 512MB by 2003... at the end of 2004 512MB was no longer that great, with the likes of Half Life 2, and my next PC had 1GB early in 2005.

anyway, I kind of feel the memory limitation with my G31 boards, Intel limited them to 4GB, and those 775 machines are still quite fast overall and 4GB can be problematic,

Sorry to bump this old thread but it might help people who come across it by searching comparisons between Apollo 133A and i815 on Google (just did).
This hungarian ebay seller is a goldmine because he benchmarks on 3Dmark 2001 every single mobo with a 4200ti and, most of the times, same amount of ram and comparable CPUs.
Digging up for a similar 815 platform, it appears this chipset should be where the OP expected : slightly faster than the Apollo 133 (non-A). Something definitely went wrong.

Still haven't found the reply to my own question though (133A vs. 815) as he doesn't appear to have ones and forums say everything and its opposite.

Please do not post ebay links on Vogons. We had bad experiences with users that only posted here to secretly bump their ebay rep

I used to have an ECS P6VXAT. In tests it was faster than other boards with this chipset that I have had.

It is a mistake to say that some chipset is faster because it always depends on the specific motherboard.