Summary: Craig investigates video card options under OS/2, looking at three of the popular mainstream models and how well they're being supported for Warp users.

Video Cards

The video card, which is inevitable in each computer system, is responsible to process the special video data received from the CPU into a format that a monitor can understand to make a restored picture on the screen out of it. More or less the monitor screen is still THE output device of a computer system, it's the most important port through which we humans get data transferred from the computer. Therefore the Video Card is one of the most important parts of our computer and we should take very good care of it. For OS/2 users, sometimes it can be painful to find the right video card with good drivers, so keep reading.

Nowadays, video cards are sold as 2D, 3D, and 2D/3D cards. A persons screen is always being refreshed so speed is VERY important. There are pros and cons of each but first we need to know what the difference between 2D and 3D are.

• 2D performance, also called GUI performance. It determines how fast your office applications perform, e.g. how fast you can scroll text/graphics or how fast you can open and close new windows. Since the days of the Matrox Millennium 2D and Diamond Stealth, performance of graphic cards got pretty close to the limit and nowadays the latest graphic cards don't differ much in 2D performance anymore, most of them are pretty fast, faster than the old standard of the Diamond cards. In OS/2, most everything is 2D, so it's important to get a good 2D card.
• 3D performance is the most important topic to distinguish between different graphic cards today. Cards without 3D acceleration will soon disappear from the market and Matrox had to learn this the hard way when releasing the Millennium II with hardly any 3D features. They lost their market leader position in an instant. S3 used to be a big player in the graphic chip market, but the mediocre 3D performance of their chip put them almost completely out of business. Similar things were happening to Cirrus Logic and others. The major player is 3Dfx which Diamond and Sound Blaster both place on their dedicated 3D cards. OpenGL is going to be OS/2's only savior for 3D support. More games are starting to use OpenGL, like Quake 2. Let's hope the trend continues.

Two very important things to remember about video cards is the amount of memory it contains and the chipset. Memory is responsible for the color resolution in combination with the screen resolution in 2D and 3D, which is getting more and more important. The amount of local card memory is also determining the maximum 3D resolution as 3D needs much more local memory than 2D for the same resolution. This is due to the fact that 3D needs a front, a back and a Z-buffer. The front buffer holds what you see, the back buffer holds the next picture while it's being processed and the Z-buffer holds the 3rd dimension value (z-value, as x and y make two dimensions, z holds the third). That is the reason why a card with 4 MB local memory can offer a resolution of 1600x1200 at high color (16 bit) in 2D, because it needs 1600x1200x2 byte = 3.7 MB. However games that are using z-buffer information (and the good ones do, offering you real 3D) can only run at 800x600 x 16 bit color x 16 bit z-buffer, 800x600x6 byte (2 byte color front buffer, 2 byte color back buffer, 2 byte 16 bit z-buffer) = 2.74 MB. 3D at 1024x768 would require 4.5 MB and can't be displayed by a 4 MB 3D card.

The chipset is responsible for all performance aspects of the video card. This is where the speed is set. A lot of people have this idea that if they upgrade their video card from 2 to 4 megs of ram they will see a performance jump in speed. This is not the case, the chipset dictates speed, so when you buy that video card your stuck with it's speed unless the company comes out with a software upgrade (drivers) or you go out and buy a new card. However we shouldn't forget that the bus system (PCI/VL/ISA/EISA/MCA/NuBus) and also the main board and the main board chipset are responsible for how fast the data reaches the video card. AGP, the Advanced Graphic Port, can offer much higher transfer bandwidth than PCI.

The last item to talk about before we look at the reviews is PCI versus AGP. All main boards with slot 1 and 2, and some very new socket 7 boards, have AGP slots. The slot 1 and 2 boards are reliable but you will need to be a little wary about the socket 7 main boards as there have been bugs. The AGP specification is based on the 66 MHz PCI specification rev. 2.1, which isn't in much use currently, since all current PCI cards are still only able to use the 33 MHz PCI bus speed. However, AGP is adding three special extensions via so called 'side band' signals, provided by some special lines added to the PCI specs. These three extensions are:

1. Pipelined memory read/write operations
2. Demultiplexing of address and data on the bus
3. Timing for data transfer rate as if clocked with 133 MHz

Now what does this mean in laymen terms?

First of all AGP offers a much higher throughput over the AGP bus than PCI. PCI, is currently clocked at 33 MHz and can transport 133 megabytes per second at peak rates over its 32 bit data bus (33,000,000 * 4 byte * sec-1). AGP is clocked with 66 MHz, which enables a peak rate of 266 MB/s (66,000,000 * 4 byte * sec-1) at the classic so called 'x1' mode. But by using the 'x2' mode, which transports data on both the rising and the falling edges of the 66 MHz clock, it can transport up to 532 MB/s at peak rate (please note that it is up to the graphic accelerator's vendor if 'x2' mode is supported). With this trick it can achieve data transfer rates like that of a '133 MHz' bus, but without actually being clocked at 133 MHz at all! Now in the real world AGP is able to transfer closer to the hypothetical peak values due to some extra signal wires which enable pipelining and queuing of requests. Once again, watch out for companies saying that their product is the fastest, since there are ways to fudge tests and still look "legitimate". If you have a PCI card right now and it's working for you, then leave it. If you were going to upgrade and your current motherboard can support it, get a AGP card. AGP has only five benefits to PCI but they add up:

• Higher bandwidth than PCI, up to 4 times as high
• No sharing of bandwidth with other components like in the case of PCI
• DIME, direct memory execution of textures
• CPU accesses to system RAM can proceed concurrently with the graphics chip's AGP RAM reads
• Allowing the CPU to write directly to shared system AGP memory when it needs to provide graphics data, such as commands or animated textures. Generally the CPU can access main memory more quickly than it can graphics local memory via AGP, and certainly faster than via the PCI bus.

Obviously it doesn't take a Pentium II to provide the needs for an AGP system. This is why Socket 7 systems with AGP will do just the same as the AGP provided by the 440LX chipset for Pentium II platforms.

Ok, now that we're done with that, let's get into the reviews:

1. Matrox G200 AGP with 8 megs of ram (MGA Ver.)
2. ATI all in wonder PCI with 4 megs of ram
3. Diamond Stealth 2500 Series with 2 Megs of Ram (PCI)

All reviews are on Warp 4, Fixpack 9. I'm using only drivers given by the company, GRADD drivers will be in another review. Sorry right now if I didn't cover your card here but I was looking into the fastest cards with good OS/2 drivers.

I hope this will clear some of the smoke away from OS/2 dealing with Video Cards and their drivers. In the months that come I will be writing on Hard Drives, Ram, CPU's, and everything else. If you want something discussed please E-mail me at craig@os2ezine.com.