On 2/5/07, Doug Carroll <[EMAIL PROTECTED]> wrote:
HiI'd like to add here neither MIPS nor Ghz are a good indicator of performance. Example on the Intel side Pentium 955 (dual core, 3.3 Ghz) is outperformed by a Pentium E6300 at 1.8 Ghz (again dual core) Why, the New E6300 being a Duo Core Processor has reduced the pipeline from like 32 level down to 14 I think the improvement in performance though at almost have the Freq of the older dual core is due to more efficient pipelining Something IBM learned along time ago, Freq is even more meaningless than MIPS. There is no magical formula here. z9's are much better at some workload, Intel is better at others. I have tried to stay away from this when talking to our Distributed people. the whole mine is better than yours instead we've concentrated on who's better for the job and we're getting somewhere now. they clearly see that web serving and application serving is fantastic on a z9, composing 20million statements a night is better on Intel (though we still do it on a z9, proximity to the data and all) FWIW: One benchmark program put the z9 IFL at 1.7Ghz but cautions take it with a grain of salt William 'Doug' Carroll barton <[EMAIL PROTECTED] ity-software.com> To Sent by: Linux on [email protected] 390 Port cc <[EMAIL PROTECTED] IST.EDU> Subject Re: 1 480 MIP IFL CPU = ? 01/31/2007 12:13 PM Please respond to Linux on 390 Port <[EMAIL PROTECTED] IST.EDU> Instructions and architecture matter. I'd put a z9 closer to 2.5Ghz than 1.7 in comparing linux workloads.... A 2Ghz intel will NEVER outperform a z9. >> >>Well, according to this article " Design methods for attaining IBM >>System z9 processor cycle-time goals" >>(http://www.research.ibm.com/journal/rd/511/mayer.html), the >>cycle time >>of the new z9 processor chip is 1.7 GHz. So, based on these numbers >>alone, a 2GHz. Pentium would outperform a z9 processor chip >>by a factor >>of: 2.0/1.7 = 1.1765.....;-) >> >>Of course, this isn't very useful since we do not take into >>account the >>differences in system architecture (RISC v. CISC, etc.) and design >>(number of processors, memory access, etc.). But at least you can tell >>your management that the processors in your new z9 are as fast as the >>processors in their desktop PCs......;-) >> >>Have a good one. >> >>DJ > > > ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 [attachment "barton.vcf" deleted by William D Carroll/OH/ONE] ----------------------------------------- This transmission may contain information that is privileged, confidential, legally privileged, and/or exempt from disclosure under applicable law. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or use of the information contained herein (including any reliance thereon) is STRICTLY PROHIBITED. Although this transmission and any attachments are believed to be free of any virus or other defect that might affect any computer system into which it is received and opened, it is the responsibility of the recipient to ensure that it is virus free and no responsibility is accepted by JPMorgan Chase & Co., its subsidiaries and affiliates, as applicable, for any loss or damage arising in any way from its use. If you received this transmission in error, please immediately contact the sender and destroy the material in its entirety, whether in electronic or hard copy format. Thank you. ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
This may not apply for everybody but let me give you our results. by the way, Uriel Carrasquilla (from NCCI) will be presenting our results over at SHARE on TUE FEB 13 4:30 PM in Tampa (From A to Z, a Customer Experience). Our application is heavily loaded with I/O's using Oracle. The application itself is Java. I tested AIX on 1.1 GHz and 1.6 GHz CPU's. The results for the P570 1.6 GHz CPU's versus the z9 IFL's is 3:1. The 1.1 GHz was 4:1. The results for memory is 1.7:1, i.e. I needed about twice the amount of memory on AIX compared to z9 IFL's with Linux. We used SLES 9.3 and AIX 5.2 for our testing. If you would like a copy of our paper, please send me a note. ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
