Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Seagate Technology - Fundamentals -- Ignore unavailable to you. Want to Upgrade?

To: Carl R. who wrote (155)	11/13/1998 3:03:00 AM
From: Frodo Baxter	Read Replies (1) \| Respond to of 1989

This bandwidth thing just won't die! But you do have a couple minor misconceptions that I can't help correcting. >Multitasking does affect bandwidth. If you are multitasking you keep getting your L1 and L2 cache flushed for some other job, meaning you have to go to memory more often. Worse, if you have so many tasks running that you are using virtual memory (i.e. swapping to disk), you have a serious bandwidth constraint. This is incorrect. Let's assume your 300Mhz processor switches tasks every 1ms. So you get a cache flush every 300,000 clock cycles and the attendant delay with slower system memory. Net effect: No noticeable difference. Also, if you're using virtual memory, that's hardly a bandwidth constraint, it's a memory constraint (i.e. having Rambus wouldn't help) >Again, with faster memory and the same CPU the system will be faster, no question about it. Not noticeably faster except for certain unusual applications, but faster nevertheless. I would assume that the difference would be about 2-5% or so. Try turning off L1 and L2 cache, and the difference would be much larger, though. That reminds me. About 3 years ago, when they were transitioning from FPM to EDO, I ordered a P120 from Gateway. The older models had FPM + L2, while the newer ones had EDO, no L2. What's the difference, I asked? Oh, they're about the same, the salesman said. Have you guys done any benchmarks? No, the salesman lied. I ordered the EDO machine. A little bit after I got my box, the inevitable benchmarks started coming out showing 0-1% difference in performance between identically configured machines save the RAM. But of course, my box wasn't identically configured. They scalped the L2 to save a measly $20. Eventually I replaced the motherboard with one with a pipeline burst L2 and immediately got around 20% extra performance. Noticeable, too. So here is clear cut evidence where bandwidth definitely was a bottleneck. Not because of any inherent design flaws, but because THE BOZOS WANTED TO SAVE A FEW BUCKS. That same stupid logic was also at play with the original Celeron. Save for deliberate, foolhardy attempts at cost savings, I stand by my original contention that internal bandwidth development has yet to become a significant inhibitor of faster components.