--------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200049 Date: 03/20/98 From: JONATHAN DE BOYNE POLLARD Time: 11:30am \/To: GEORGE IVANOV (Read 4 times) Subj: 75MHz Motherboards GI> I have to admit that I am a newbie as far as MoBos are concerned and GI> that was the main reason to slip into believing that a few vendors GI> (advertised in ZD PC Magazine) could supply me with 73MHz or even GI> 83MHz Mobo, based on Intel's 430TX chipset, with Cyrix 686 166MX GI> installed on it. They say it is possible (would you believe it?) and GI> are advertising it - openly!!! A lot of motherboard vendors now provide the ability to use a 75MHz or 83Mhz system clock. Not all will actually *work* at that speed, however, and usually (if one goes to the technical support pages on their web sites) one finds that the advice from the motherboard manufacturers is "anything above 66MHz with an Intel chipset and you are strictly on your own, son". This is not to say that speeds above 66MHz are unsupported, full stop. FIC, for example, supports the 75MHz speed on those motherboards in its 1st Mainboard range that are based on the VIA chipset, and gives all sorts of useful information relating to it, such as what PCI cards will operate satisfactorily at 37.5MHz. Other motherboard manufacturers are happy to talk about 75MHz for the SiS, Aladdin, and VIA chipsets. It is only with the Intel chipset that they all start to be cagey. There's a lot of pressure on the motherboard manufacturers from Intel not to support speeds over 66MHz. It is reported that Shuttle won't advertise its 75/83MHz HOT motherboard, based on the SiS chipset if memory serves, on its web site along with all of its other motherboards because of pressure from Intel. Of course, when Intel comes out with its 440BX chipset, capable of supporting up to 100MHz, expect a lot of revisionism in the press, Intel trying to claim that it has supported faster system bus speeds all along, and "why don't you buy our new 440BX chipset?". We have the likes of SiS and VIA to thank for this, of course. They've been supporting 75MHz and 83MHz operation of their 5571 and Apollo VP chipsets for well over a year now, and the public's favourable reaction to this, to the extent that, as you see, "Intel only" motherboard manufacturers are now openly encouraging people to overclock Intel chipsets *despite* pressure from Intel to the contrary, in order to compete, has no doubt put pressure on Intel to finally improve the bus speed of *its* chipsets. But Intel is still playing catch-up in this area. VIA released its MVP3 chipset a month or so ago, which is capable of 100MHz operation. By all reports, 100MHz motherboards using this chipset are now in testing, and we should see them hit the streets in around two months' time. Of course, one needs a CPU capable of running with a 100MHz bus as well, and none of the current crop are officially capable of this. But again, rumour has it that AMD is working on a 100MHz K6, and the fact that the overclocking crowd (I don't think that overclocking is that good an idea myself) reports good results from running the 66MHz AMD K6s (i.e. the 200 and the 233) at higher bus speeds bodes well for this. I'll certainly be following the developments with interest over the next few months. JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3) --------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200050 Date: 03/20/98 From: JONATHAN DE BOYNE POLLARD Time: 12:18pm \/To: GEORGE IVANOV (Read 4 times) Subj: 75MHz Motherboards JDBP>> I went through the whole process again just recently, and finally JDBP>> settled on a FIC PA 2012. GI> It probably works at 83MHz right? I don't overclock. I have a Cyrix 6x86MX 233, which uses a 75MHz bus, ough. GI> Did you manage to find user friendly UK vendor or you ordered it GI> from the States? How much does it costs? I went back to the information that I looked up last year, and contacted the vendor that I had located in the United State. I was given the runaround by some salespeople who didn't really speak English very well, and eventually, after having FAXed my order to them and heard nothing, was told that they don't ship internationally. Understandably, I was quite annoyed with this, especially given the fact that only three days prior to that they had actually quoted the price to me for shipping to the U.K.. So, like a responsible consumer, I exercised my market forces muscle, gave up on Leapfrog Lab, and looked up System 2000 in Portsmouth once again, who last year hadn't been told by FIC about the PA 2007. This year things were better. Apparently the FIC motherboards are selling quite well. System 2000 was very helpful, very friendly, and took my order for a FIC PA 2012, plus Cyrix 6x86MX 233, plus heatsink and fan without fuss. The whole lot came to 208.48+V.A.T. (105 for the FIC PA 2012 with 1MB cache, 98.48 for the CPU, and 9.90 for the heatsink+fan), and arrived, after a slight delay in the back order, just over a week later. The fan turned out to be a ball-bearing fan (which is better than a sleeve-bearing fan, I gather), and the motherboard was preconfigured for the 6x86MX. I checked the configuration myself, but I didn't actually have to alter a thing. I went for the 6x86MX 233, rather than the 200 that I had originally planned on, because this price was only 17 more than the cost, including the $75 shipping and the $19 credit card fee that were quoted, of buying the FIC PA 2012 + 6x86MX 200 from Leapfrog Lab in the U.S.. System 2000 can be reached by 'phone on +44 1705 293000, if you or anyone reading are interested, or FAXed on +44 1705 293029. JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3) --------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200051 Date: 03/20/98 From: JONATHAN DE BOYNE POLLARD Time: 01:03pm \/To: GEORGE IVANOV (Read 4 times) Subj: FIC PA 2012 JDBP>> By coincidence, it is due to arrive today. GI> Lucky you! At home (in my country) I have an old 386 computer (which GI> originally was 286, but I changed the Mobo) and I intend to replace GI> the Mobo with the one you've suggested, plus buying a new EIDE or GI> SCSI HDD and replacing the memory with a fast one. What do you think GI> I should be aware of? Several things: The FIC PA 2012 is an ATX motherboard. I had to buy a new case (an Acme full tower from NexNix), because my existing tower case could only hold AT motherboards. But, on the other hand, the 1st Mainboard range from FIC includes what are essentially the same motherboards in both ATX and AT sizes (although I recommend checking the small print: the L2 cache size options can vary between the AT and ATX flavours), so if a new case is an insurmoutable problem then have a look at the specifications for the PA 2007. The FIC PA 2012 only accepts DIMMs. I only had SIMMs, so I had to buy a 32MB DIMM. This wasn't so bad, though, since they were only 70ns FPM SIMMs (my old motherboard ran at 50MHz) and couldn't have been used in a 75MHz system anyway. A 32MB Synchronous DRAM DIMM wasn't as phenomenally expensive as it had been when I costed everything a year ago, either. I gave my old SIMMs to my girlfriend, to be put them in a 4-way SIMMSwapper and stuck into her Gateway 2000. She now has, instead of an 486DX2/66 with 8MB, which she had all but stopped using because it was so slow, a Gainberry 5x86DX4/100 with 24MB. The FIC PA 2012 is an socket 7 AGP motherboard. Given what I had heard about AGP, I actually ordered a PCI graphics card. I was sent the AGP card by mistake, though. I installed it anyway and it has worked flawlessly so far. I can report that the Matrox Millenium drivers for OS/2 are quite capable of driving the AGP Matrox Millenium, and I have a third PCI slot to play with that I wasn't expecting to have available. The only downside is that I have discovered that I really *do* need a new monitor, since my existing one it turns out cannot handle anything more than 640*480. (-: The FIC PA 2012 has PS/2 connectors for both mouse and keyboard. I, of course, had an AT keyboard. This is perhaps a trivial point, since an AT->PS/2 keyboard converter was only 9.50, but it did spur me to buying a PS/2 mouse as well. I looked at all of the mice on offer, with all of their little wheels ("For scrolling through the Internet!"), extra buttons, and other gimmicks, some of which cost over 60, and finally settled on a Fellowes SoftTouch mouse costing 13. All right, so maybe I did fall for *one* gimmick: the SoftTouch velvet coating. But given that at 13 it was nearly the cheapest mouse in the shop (only the cheap and nasty bog-standard mice were cheaper), and the fact that I *do* like the feel of the thing, I don't feel very guilty. I have to report that it's been a pleasant experience so far. The two aged ISA cards that I was able to transfer over in the interim whilst I upgrade to PCI have just worked, even though the system is running at 75MHz. The LS-120 has worked. There was one heart stopping moment when I first booted the thing up and all it did was beep continuously -- then I remembered to put the RAM in. (-: The AWARD BIOS is up to date and has had no trouble recognising the Cyrix 6x86MX 233, and also supports the Phoenix INT 13h extensions. The APM support in the motherboard and its interaction with OS/2 Warp 4 and FixPack 6 has been *very* surprising --- more of which, though, in another message. JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3) --------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200052 Date: 03/21/98 From: JONATHAN DE BOYNE POLLARD Time: 12:28pm \/To: WILL HONEA (Read 4 times) Subj: 75MHz Motherboards WH> George, let me jump in with one potential problem source that JDBP WH> might miss since I've got a couple of those WH> home-grown-constant-upgrade-specials around. If things start to get WH> flakey with the Pentium class mother board, consider trying a larger WH> power supply as one of your options. I have had to update several WH> older 200 watt supplies, commonly used in 286/386 machines, to WH> 250-300 watts or more to settle some systems. I bought a new case as well, so I also had a new PSU. My old case had a 230W PSU, and my new case I believe has a 240W PSU. WH> Average power capacity was fine, but the transient response WH> wasn't good enough. The new boards use regulators to step down the WH> 5 volts to 2.9 - 3.5 volts (depending on the CPU) and some just WH> use linear regulators hence the current spikes kill the 5 volt WH> supply to the rest of the system got too noisy. The FIC PA 2012 uses a switching voltage regulator. I'm told that that is good, although I have yet to find an explanation as to why. JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3) --------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200053 Date: 03/21/98 From: JONATHAN DE BOYNE POLLARD Time: 12:33pm \/To: GEORGE IVANOV (Read 4 times) Subj: 75MHz Motherboards GI> Type: Super P5MMA98 GI> Specs: [...] GI> Intel 430TX PCIset GI> 8MB to 256MB EDO, SDRAM or FPM GI> GI> Vendor: Choice Peripherals, [...] GI> GI> Some other vendor specifying exactly the same type MoBo even claims it GI> will work on 83MHz GI> GI> Do you believe these guys? Only inasmuch as the SuperMicro motherboards do have jumpers to set the clock to 75MHz and 83MHz, yes. One has to bear in mind that "75MHz support" sometimes means just that: the motherboard clock generator can be set to produce a 75MHz system clock. It says nothing about whether your PCI cards will work at 37.5MHz or 41.5MHz. It also doesn't say whether the *chipset* will work at 75MHz or 83MHz. I suspect that a lot of this, especially in the case of motherboards with the Intel 430HX/VX/TX chipsets, is "checkbox marketing". The vendors sensed that they were losing out to the SiS and VIA chipset motherboards, which *officially* support speeds over 66MHz, and so produced Intel chipset motherboards that could have the "75MHz" and "83MHz" checkboxes ticked in the marketing literature too. The unfortunate truth is that according to several reviews, this sort of oneupmanship results in some motherboards allowing one to set the clock rate to speeds that the motherboard doesn't actually work at. "Caveat utor" applies, and a good read of things like the Motherboard Guide at Tom's Hardware (http://www.tomshardware.com/) is essential. I notice in passing that they omitted to mention that the 430TX chipset won't cache more than the first 64MB of that 256MB of RAM. GI> ACPI/PC98 features: - Boot support for USB keyboard (?) AFAIAA this just means that someone with a USB keyboard won't have to plug in a PS/2 or AT keyboard in order to access the BIOS setup screen. JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3) --------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200054 Date: 03/21/98 From: JONATHAN DE BOYNE POLLARD Time: 12:06pm \/To: GEORGE IVANOV (Read 4 times) Subj: Bidirectional parallel p LN>> This PAR1284.SYS device driver can be found on the IBM OS/2 Device LN>> Driver Pak On-Line at: LN>> http://service.software.ibm.com/os2ddpak/html/ LN>> CE7558BA7305FF150625636F007AAA3D.html GI> When I try to access this (I don't have Warp 4, so I have to type all GI> these numbers myself ) it says "File Not Found". When one is accessing a web site which holds its web pages in a database (and this one obviously does) it's probably better to start at the root, http://service.software.ibm.com/os2ddpak/html/ in this case, and follow the links. For one thing, the database keys to the web pages on such sites tend to vary as the database is updated. It's a lot less typing that way (assuming one doesn't use cut and paste from message reader into web browser) at least. JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3) --------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200055 Date: 03/21/98 From: JONATHAN DE BOYNE POLLARD Time: 12:15pm \/To: GEORGE FLIGER (Read 4 times) Subj: Matrox Millenium II JdBP>> I suppose that it doesn't help you to know that my new Matrox JdBP>> Millenium II has worked fine so far (I haven't installed the JdBP>> drivers yet, so I'm only using it in VGA mode at the moment.). GF> I've got my Matrox Millenium II set at 1280x1024x16.7M and it works GF> great! Even at this setting it blows away my old Diamond Stealth 64 GF> 2200XL I had set at 1024x768x65k. Absolutely no comparison in speed. I've installed the drivers since writing that message, and everything is working fine here too. The only downside is that I have discovered that my monitor (which is a VAR rebadged one, so I have no idea what monitor it actually is) cannot handle the sync rates for anything over 640*480, so I really *do* have to buy a new monitor. Oh well, at least I can have 640*480 in 16 million colours in the meantime. (-: The speed is impressive, though. When running the Matrox Millenium using the standard VGA drivers there was considerable improvement over my old Paradise VGA card. Switching between text to graphics mode was noticably faster, for instance. The big speed increase came with using the Matrox display drivers, however. Scrolling of text in a Take Command for OS/2 window when doing DIR /S C:\ was pretty much the same as it had been on my old VGA card using the standard VGA drivers (I could still just about read the directories as they scrolled by, and the hard disc would only). I installed the Matrox display drivers and the text is now a blur. It is now obvious that the main thing slowing down DIR /S C:\ is the hard disc (I'm using my old one, which is a Seagate ST1480 using PIO mode 0.). JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3) --------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200056 Date: 03/23/98 From: JONATHAN DE BOYNE POLLARD Time: 05:02pm \/To: JONATHAN MICHAELS (Read 4 times) Subj: motherboards and os/2 JM> i am being given a supermicro JM> p6sne complete with an intel pentium pro 180 mhz (and fan) and 128 mb JM> of full parity (synchronius) simms. i hope to add to this an adaptec JM> aha2940uw and a matrox millenium II with about 16 mb of video ram. JM> JM> my question is, quite simple, is this combination enough horsepower to JM> run os/2 warp v4.0 with the whole tcp/ip and other networking stuff JM> loaded ? You must be thinking of some other operating system if you are under the impression that it isn't. (-: I've quite happily run OS/2 Warp 4 with TCP/IP and NetBEUI networking on a 75MHz Pentium with 32MB of RAM, and up until recently was running OS/2 Warp 4 sans networking on a 486DX50 with 20MB of RAM. JM> currently i run warp v4 on a i486dx33 with 16 mb and an adaptec JM> aha2542b and it is slugish to say the least I'd say that the major hindrance to performance there is your 33MHz motherboard. JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3) --------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200057 Date: 03/23/98 From: JONATHAN DE BOYNE POLLARD Time: 05:04pm \/To: JONATHAN MICHAELS (Read 4 times) Subj: motherboards and os/2 JM> on another slightly different track, this motherboard (p6sne) has a JM> ps/2 mouse port, and i think a ps/2 keyboard port as well, is this JM> going to be a problem for os/2, [...] ? No. I must admit that I'm surprised by the question. OS/2 has supported PS/2 mice since at least OS/2 version 2.0, and has probably supported PS/2 mice for as long as there have been PS/2 mice to support. As for the keyboard, there is really very little difference between an AT keyboard and a PS/2 keyboard apart from the shape of the connector. From a software standpoint they are largely indistinguishable -- to support the one is to support the other. JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3) --------------- FIDO MESSAGE AREA==> TOPIC: 260 OS/2 HARDWARE Ref: F4200058 Date: 03/23/98 From: JONATHAN DE BOYNE POLLARD Time: 12:49am \/To: HARRY OLDENHUIS (Read 4 times) Subj: Installing Warp 4 SB>> I just caught this thread. I'm about to install an EIDE drive in my SB>> system but still want to boot from my SCSI (adaptec 2940). How SB>> should I go about this? By not specifying any EIDE stuff in the SB>> CMOS? Will I still be able to access the drive? HO> You have to specifi the hard drive setting in cmos or it wont load Untrue. The BIOS is just *one* source of information that IBM1S506.ADD uses to determine where one's ATA hard disc drives are. Another source of information is the command line that is specified for it in CONFIG.SYS. So even if an ATA drive is not specified in the NVRAM setup information used by the main BIOS it can still be used within OS/2, as long as one tells IBM1S506.ADD explicitly, via command-line options in CONFIG.SYS, where the drive can be found. HO> The IDE is faster to load than a scsi drive so it will boot before HO> a scsi as the scsi it is not ready I have no idea what you mean by "faster to load" here. I assume that you are referring to the time that it takes for the drive to spin up at power on. This has absolutely *no* bearing on the boot order of the drives whatsoever. The boot order is determined by the BIOS, and the unit numbers (80, 81, 82, and so forth) that it assigns to the various SCSI and ATA drives. This latter is, in turn, determined by whether a SCSI BIOS extension is present, and whether it chooses to place its SCSI hard discs before or after the ATA hard discs in the BIOS unit number ordering. JdeBP --- FleetStreet 1.19 NR * Origin: JdeBP's point, using Squish (2:440/4.3)