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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: Dan3 who wrote (3955)	8/8/2000 11:25:45 AM
From: dale_laroy	Read Replies (3) \| Respond to of 275872

Hi Dan3. While 10% is probably fairly accurate for a single processor P-III based system, it is probably low for both Athlon and multiprocessor systems. Unlike most previous processors, the Athlon does not support critical word first presentation of data, therefore the latency is the entire length of the burst, which is significantly impacted by doubling the data rate. I would guess an average of 15% for Athlon in a single processor system. Additionally, multiprocessor systems are affected significantly more than uniprocessor systems by memory bandwidth. Unfortunately, I am not sufficiently familiar with benchmarking of multiprocessor systems to hazzard a guess.

To: Dan3 who wrote (3955)	8/8/2000 2:54:10 PM
From: mapolo	Read Replies (1) \| Respond to of 275872

Hello, Dan ! You may be right with your assumption, 266MHz FSB will boost overall performance by 10%. Within the hierarchy of memories, the HDD is the "caboose" of the system. If the operating system does does not access the HDD while performing a calculation, then the access to the DRam is the main impact on the performance (Linux OS f.e. accesses the HDD very rarely). From an Excel-sheet, which I put up last year, the rough "calculation by thumb" shows an increase of about 90% in performance (FSB 100MHz compared to 266MHz), when only cpu-registers, L2-cache and FSB are invoked. But nevertheless the FSB stays the bottleneck of the system, therefore new architectures like "Infiniband" are under investigation. If I should have overseen something, please feel free to discuss it! CU Mapolo