Quarterdeck MagnaRAM '97

MagnaRAM's performance surprised me. I had expected very little if any acceleration. Instead, I got quite a nice shock. First I tested TurboLoad alone. To test load time acceleration, I loaded a sequence of large applications, and then tested Hexen II framerate to check CPU usage. I loaded applications in this order: Adobe Photoshop 4.0, Microsoft Frontpage 97, Microsoft Excel 97, America Online, Netscape Navigator 4.0. Then, with all these open, i checked the Hexen II framerate. I found a small amount of acceleration when using TurboLoad alone (without RAM compression), but the boost wasn't exceptional. (see the table below)

Next, I tested MagnaRAM Compression alone. When I repeated the same test with just compression, I got impressive results. The load times for Adobe Photoshop and Microsoft Frontpage were actually about one second slower than normal. But after this, they became much faster. Excel and AOL loaded 3 seconds faster each, but Netscape loaded more slowly. MagnaRAM really shows its stuff here. After loading the first two applications, Windows would normally have started swapping to the hard disk. However, with MagnaRAM running, the swapping was less frequent, as well as more efficient. The Hexen II framerate jumped to almost 29 FPS. Obviously MagnaRAM does not hog resources, as the framerate had actually increased by 5 FPS from normal. I repeated the test once more this time combining the compression with TurboLoad, and i found that the performance was a bit less than with the compression alone.

Here we see that a system with only 2MB of Video RAM, when compared to an identically configured system equipped with a 4MB Video Card actually outperforms it in a gaming situation. However, this does not hold true when using above 32MB of System RAM and in high end situations where quite a large amount of RAM is devoted to the Graphics Card.

MagnaRAM does improve performance quite a bit in systems only equipped with 32MB of System RAM, however as you can tell from the chart below, it really begins to shine when the test system boasted 64MB of RAM instead of 32MB. The increase in performance after 64MB of RAM however is minimal.

In addition to applicaiton load times, I ran the Wintune 97 benchmark on the system, but the difference in results was minimal, except for that of Cached speed for drive C (the Windows drive). This is most likely due to the fact that MagnaRAM's Cacheback feature uses the hard drive cache memory when it is idle, and probably took time to return it to the system when it was needed. Compared to these figures, actually using MagnaRAM gives a much better impression. I really felt the improvement in the system performance while running MagnaRAM.

Final Decision

MagnaRAM is an excellent, useful tool for Windows users, but probably only worth the buy if you are really going to use your RAM to its potential.