More Socket 7 AGP Boards Becoming Available Now
A few months have gone by since FIC released the first Socket 7 motherboard with the new Advanced Graphics Port AGP. After there first was only VIA’s Socket 7 AGP chipset Apollo VP3, there are now more AGP chipsets for Socket 7 turning up to improve this market with some competition. SiS released their 5591/5592 chipset and ALI will launch their Aladdin V chipset very soon as well. Whilst VIA’s Apollo VP3 chipset was limited to only maximal 75 MHz system bus clock, ALI comes out straight away with AGP and 100 MHz support. VIA just released their answer to the Aladdin V, they announced the Apollo MVP3 chipset last week, which will also support 100 MHz system bus clock. SiS could not tell me if the 5591/5592 will be for 100 MHz or only for less, but it seems as if the 5591/5592 does indeed support 100 MHz system bus clock, since it’s got what it takes for it. The minimal reqirement for successfully running at 100 MHz is a divider option of 1/3, which makes 33 MHz PCI clock and 66 MHz AGP clock out of the 100 MHz system clock. The ALI Aladdin V, the SiS 5591/6692 and also the Apollo MVP3 are all offering this divider.
Waiting for 100 MHz System Bus Clock ?
This very issue is a serious problem when you are playing with the thought of upgrading to a Socket 7 AGP system. It won’t be too long until all Socket 7 boards will support 100 MHz system bus clock, so why buying a board that’s limited to 75 MHz now? I can’t give you a straight answer to this, but you should look at this issue from two different sides. Although 100 MHz boards will be available within the next 4-6 weeks, it will be the first generation of 100 MHz boards with all the problems this bus clock will hold for us. Initially there’ll be a serious problem getting good quality SDRAM to run at this speed, let alone the fact that there won’t be any real 100 MHz CPUs available at that time either. The RAM timing of the first boards will be very conservative to ensure reliability with lower quality SDRAM, so that there may not even be that huge performance increase over the 66, 75 or 83 MHz now. I expect the first 100 MHz board customers to build a large test field for the board manufacturers and maybe you rather wait until there’s second generation and reliable 100 MHz boards for Socket 7. On the other hand it’s true, 100 MHz system bus clock for Socket 7 will enhance speed significantly, much more than what we can expect from 100 MHz in Pentium II systems. If you want as much Socket 7 performance as possible as early as possible then you gotta go for a 100 MHz board as soon as it’s available. However, if you’re a real performance freak, then you should maybe already go for a Socket 7 AGP board of the ones that are reviewed here. Last but not least is there of course the possibility of moving over to a Pentium II system as Intel wants us to. Even for the ones that don’t want to spend that much money there’ll be the L2 cacheless Pentium II soon and cheap Slot 1 boards are already available now. This is certainly also something to consider. AMD will have to offer more than the K6 266 pretty soon and I hope they’ll release the K6 3D latest at CeBIT this year and this K6 3D should run at 100 MHz in the first place.
Although the 100 MHz system bus story is a very interesting one, there is not too much of a point talking about it yet, because there’s no CPU officially available that would support this high bus speed. Hence I’ll concentrate on boards that are currently available and which can run any Socket 7 CPU that you can buy now.
Old VIA AGP Problem Fixed Now !!
Only today I received the latest VIA_GART.VXD driver, rev. 1.7, which now fixes the nasty problem all Apollo VP3 boards had with ATI Rage Pro AGP cards. The testing showed that Rage Pro cards are now running just as well as all other AGP cards. lease don’t forget that the Apollo VP3 is only capable of AGP x1 mode. This was most likely the reason for the incompatibilities with the Rage Pro, which supports x2 mode. If you’ve read my AGP summaries you’ll know that x2 isn’t really an issue until 100 MHz system bus are reality and games take real advantage of very large textures. It’s currently certainly no performance issue.
CPU Issues
A few new AMD CPUs will be released before 100 MHz CPUs will become available, which are the long awaited K6 266 and K6 300, both running at 66 MHz bus speed. These CPUs are manufactured in .25 micron technology, need less power and get much less hot. They also require a smaller voltage than the predecessors, 2.2 V to be exact. Since I’ve got a K6 266 here I think that I am not asking too much of a modern Socket 7 AGP board running this CPU without problems. Unfortunately were two of the five tested boards unable to run this CPU, just because they would supply a falsely too high voltage when adjusted to 2.2 or 2.1 V. The special thing about the new K6 266 is that due to its structure and .25 micron technology it won’t run at a voltage that’s too high. Hence these two boards were pretty useless with this CPU and need to be reviewed.
Testing Procedure
I ran each board with an AMD K6 233 CPU (3.2 V) and then tried benchmarking it with the K6 266 (2.2 V). I also tried overclocking this CPU to 300 MHz, which was successful in two cases. I also did a reliability test with the Cyrix/IBM 6x86MX PR233 (2.5×75 MHz), but I did not benchmark with this CPU. The good news is that each board ran the 6x86MX, without any problems, four at PR233 one only up to PR200, the bad news is that none of the five boards would run at 83 MHz or higher system bus clock reliably, although one board even offers 100 MHz bus clock.
It’s now 1998, so I was using Business Winstone 98 for benchmarking. This benchmark is pretty huge, but it has the disadvantage of generating pretty low result numbers, which range at about a third of the result numbers of Winstone 97. Hence please realize that a difference of only 0.5 in Winstone 98 is like a difference of 1.5 in Winstone 97. Look at the numbers relatively, not absolutely please. The other benchmarks I ran were for the 3D performance of each board. Surprisingly the TMark from Acclaims new game Turok showed significant differences between the boards. The tests were ran with a Diamond Fire GL 1000 Pro AGP card and additionally with a Diamond Monster 3D. Both cards were used for Turok as well as PCPlayer’s 3D benchmark, to show AGP as well as PCI performance.
Benchmarks with AMD K6 233
Benchmarks with AMD K6 266
Benchmarks with AMD K6 300
Tom’s Two Cent
After having a close look at the benchmarks you’ll see that the FIC PA-2012 is not the most interesting candidate at all anymore. It performed worst in pretty much every benchmark, although the distance to the other boards isn’t big in all except one case. I still wonder what the board has against the 3Dfx Voodoo. I ran this test 5 times, but it kept being almost 10% slower than all the other boards.
Soyo’s SY-5ED5 shares one problem with the PA-2012, it doesn’t supply the correct voltage when setting it to 2.1 V. I measured 2.44 V and this is simply too much for the new K6 266. This does not have to be a serious issue for long, because I would expect this problem top be solved pretty fast. However it shouldn’t have happened in the first place and how do you know if you’ve got a board with the correct voltage or if you haven’t? The performance of the SY-5ED5 is hardly much worse than the performance of the others, but it still couldn’t make more than 4th best in the benchmarks.
The surprise candidate to me was the EpoX P55-VP3. It was the only board that ran straight away, without any fiddeling around. The performance of the board is fine and I am sure that I only couldn’t overclock it to 300 MHz because the voltage was with 2.1 V a little bit too low for overclocking. As soon as EpoX could fix the voltage and raise it to 2.2 V I’m sure this will be an easy to use and reliable Socket 7 AGP board.
EpoX P55-VP3:Very easy to configure, but no 83 MHz system bus clock.
DFI’s P5XV3 is the fourth candidate with VIA’s Apollo VP3 chipset, but as well as FIC’s PA-2012 it offers 1 MB of L2 cache and contrary to the PA-2012 it’s able to make use of it. The 1 MB L2 cache is most likely the reason why it won almost all benchmarks. Although the manual wouldn’t mention a word about 2.0, 2.1 or 2.2 V CPU voltage I eventually could find the according jumper setting for the K6 266. The multiplier setting for the K6 266 is also different to the description in the manual … if you put the jumpers as said there you’ll run the K6 at 333 MHz instead. Nevertheless, this board is currently the fastest Socket 7 board I’ve tested and if you can live with the fact that it doesn’t even support 75 MHz (no good for 6x86MX PR233 owners), let alone 83 Mhz, then this is currently the Socket 7 AGP board to go for.
Elitegroup’s P5SD-B is currently the only board that came with a different chipset, the 5591 from SiS. I haven’t been a big fan of SiS chipsets in the past, because they normally had quite a lot of trouble with the RAM timing. However, this board seems to put an end to this rule. When I received this Baby AT board I couldn’t wait testing it, because even on the board it shows the jumper settings for 83/33, 90/30 and 100/33 MHz system/PCI clock. Unfortunately 75 MHz was all that ran reliably, but that could get fixed with a BIOS update. After all this is the first official 100 MHz board I’ve seen. The performance of the P5SD-B isn’t quite as good as the performance of the DFI board, but the P5SD-B only comes with 512kB L2 cache, and for this amount of cache the second place in performance is quite considerable. Of all the five boards I’ve tested it would be the one I would go for, although it may be smart to wait until 83, 90 and 100 MHz bus speed will run reliably.
Elitegroup P5SD-B: If this doesn’t look promising ..
