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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: wanna_bmw who wrote (163874)	4/16/2002 10:40:18 AM
From: Tony Viola	Read Replies (3) \| Respond to of 186894

WBMW, sorry about that. From alan81 on TMF: An analysis of the latest benchmark data from Tom's hardware and a prediction for performance for the remainder of the year. www6.tomshardware.com I entered the data for the 22 benchmarks Tom used into an excel spreadsheet and began crunching the data into some simple percent improvement numbers. This can be very misleading in that it hides much of the information, but it does serve as a useful approximation of where we might be going, and what such an average might look like at some point in the future. Based on the benchmarks that Tom used I found the following: The P4 data: The 2.4Ghz P4 improves performance over the 2.2Ghz product by about 6.45%, and with a speed increase of 9%, this means that about 70% of the clock speed increase shows up as improved benchmark performance. A comparison of the 2.4Ghz P4 with the old 400Mhz bus to the new 533Mhz bus shows an average improvement of about 5.5%. A comparison of the 2.6 Ghz to the 2.4Ghz product with the 133Mhz bus shows a 7.5% performance improvement for an 8.3% increase in clock rate, which is a performance scaling to clock rate ratio of 90%! The Athlon data: Athlon XP improves performance over the old Thunderbird by 4.9% on these benchmarks vs. less than 3% on the last set of benchmarks from Tom that I analyzed. This leads me to believe that these benchmark programs use SSE more than previous benchmark programs. When we compare low speed Athlon in the model 1600+ to 1700+ range we see the performance improve by 3.25% for a model # increase of 6.25% for a performance to model # scaling rate of 52%. When I compared the model 2000 to model 2100+ I saw a performance increase of 2.4% for a model number increase of 5%, which is a scaling factor of only 48%. Using these numbers and scaling factors I predict the following performance comparisons as we move through this year: Time AMD improvement Intel improvement Intel leadership Q1 2100+ 2.2Ghz 6.10% Q2/ now 2100+ 0 2.4Ghz 6.5% 12.94% Q2 / may 2200+ 2.3% 2.4G 533B 5.5% 16.85% Q3 2400+ 4.4% 2.53GHz 4.9% 18.19% Q4 2800+ 8.0% 2.8 GHz 9.6% 21.65% Q4 3400+ 10.3% 3 Ghz 6.4% 18.60% Note that according to these numbers Intel will improve performance over this time frame by about 46% while AMD only improves by 30%. In simple terms the AMD number comes from .48*(3400-2100)/2100=30%. Note that according to this data if Intel simply sticks with the 2.8Ghz plan for Q4 and misses the 3.0Ghz goal, they will still lead hammer performance by about 10%. It is important to note that the Intel data is almost a slam dunk as the extrapolation is only from 2.6Ghz to 3.0GHz or about 15% while the AMD extrapolation is over 60% which could lead to significant error. If you look back a year ago it was thought that three things would bring the P4 performance into line with expectations. (1) Clock rate scaling: The P4 was introduced in Nov. 2000 at a frequency of 1.5Ghz, and hit 2.4Ghz in Apr. 2002. The Athlon was at 1.2Ghz in Oct. 2000 and scaled to 1733Mhz in mar. 2002. This means the P4 clock rate has scaled by 60% in the same amount of time it took the Athlon to scale by 45%. The P4 does have the advantage of 0.13u technology today, while the Athlon is still on 0.18u, however it is believed that the AMD 0.18u technology makes use of many of the 0.13u transistor benefits. This is evidenced by the tremendous 250% clock rate scaling AMD did on the 0.18u process. They are forecasting a much smaller clock rate scaling moving forward onto the 0.13u process. (2) Benchmark optimization: The improved performance of the Athlon XP over the thunderbird indicates that the benchmarks are taking advantage of SSE operations, which the P4 excels at. Code recompiles have also helped with this effort. (3) P4 core improvements: The Northwood added in 512K of L2 cache which appears to have improved performance by about 5%, which is equivalent to about 200 model unit #'s in AMD speak. The new 533Mhz bus appears to have a similar amount of improvement. We see the potential for SMT (hyper threading) on the horizon to further improve performance at a given clock rate. Intel has gone from a significant performance deficit a year ago to having a superior product today. In addition, it appears things are coming up very well in the desktop performance picture for Intel over the course of this year. It will be interesting to view the reviews as each new product is announced to see how well it stacks up against these predictions. boards.fool.com Tony