| Author |
Question about Alphaservers ES40 performances |
MSpec
Member
Posts: 5
Joined: 02.04.12 |
| Posted on April 02 2012 02:47 |
|
|
Hi, guys!
I beg your help!
Do you know what the performances of AlphaServers ES40 6/500, 6/667 and 6/833 in the VUPs measurements, both Int and Float?
Please, help me and clarify these numbers for me!
Thank you in advance!
Edited by MSpec on April 02 2012 03:01 |
|
|
| Author |
RE: Question about Alphaservers ES40 performances |
Bruce Claremont
Member
Posts: 623
Joined: 07.01.10 |
| Posted on April 02 2012 03:49 |
|
|
Start here: http://h18002.www1.hp.com/alphaserver/performance/perf_by_name.html
Edited by Bruce Claremont on April 02 2012 03:49 |
|
|
| Author |
RE: Question about Alphaservers ES40 performances |
MSpec
Member
Posts: 5
Joined: 02.04.12 |
| Posted on April 02 2012 04:06 |
|
|
Thanks you!
I've seen this material. How can I transform TPS in VUPs Int and VUPs Float?
Edited by MSpec on April 02 2012 04:08 |
|
|
| Author |
RE: Question about Alphaservers ES40 performances |
abrsvc
Member
Posts: 108
Joined: 12.03.10 |
| Posted on April 03 2012 01:04 |
|
|
I'm not sure of your goal as VUPs are a VAX only measurement relative to the VAX 11-780 machine. If you insist on translating the TPS numbers ot VUPs, take a known VAX system with its reported VUPs value as well as its reported TPS value and do the math. There is no direct TPS to VUPs value. VUPs depend upon the workload and vary a bit from machine to machine. Please note that all of these measurements use artificial workloads and are for comparison use only and won't necessarily reflect actual performance for your application or environment.
Dan |
|
|
| Author |
RE: Question about Alphaservers ES40 performances |
malmberg
Moderator
Posts: 530
Joined: 15.04.08 |
| Posted on April 03 2012 02:52 |
|
|
VUP is basically a relative metric on how fast a CPU is, and is only accurate on an unloaded system.
TPS is a measurement that takes in the I/O speed, the CPU speed, the lock manager speed, and can be subject to the types of transactions that are being measured.
A MicroVAX II with RA series disks will have a VUP rating of about 1.
A MicroVAX 3500 with RD series disks will have a VUP ratting of about 3.
Of those two configurations, with the same amount of memory, the MicroVAX II will have about 2 to 4 times the TPS rating of the MicroVAX 3500 depending on the test.
|
|
|
| Author |
RE: Question about Alphaservers ES40 performances |
abrsvc
Member
Posts: 108
Joined: 12.03.10 |
| Posted on April 03 2012 09:50 |
|
|
Actually, the previous posting had the final stats backwards. The 3500 will be the faster machine with the higher VUP rating.
Also, the typical configuration would have RD disks on the MicroVAX II and the RA drives on the 3500.
Dan
Edited by abrsvc on April 03 2012 09:51 |
|
|
| Author |
RE: Question about Alphaservers ES40 performances |
malmberg
Moderator
Posts: 530
Joined: 15.04.08 |
| Posted on April 03 2012 15:09 |
|
|
My estimated stats are accurate.
Back during that time, I was setting up MicroVAX II systems with ESDI disks and Digital was originally shipping VAX 3500 systems with RD54 disk drives.
Pure CPU bound, the MicroVAX 3500 is about 3 x the MicroVAX II, but the ESDI or RA disk drives will operate at about 1/2 Q-BUS bus speed or better, while the RD54 runs significantly slower than that.
A TPS measurement will be constrained by the slowest component in the chain, be it CPU or Disk.
|
|
|
| Author |
RE: Question about Alphaservers ES40 performances |
MSpec
Member
Posts: 5
Joined: 02.04.12 |
| Posted on April 12 2012 02:31 |
|
|
|
Gentlemen, thank you for help! |
|
|
