benchmark tests

Message boards : Number crunching : benchmark tests
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile badaz907

Send message
Joined: 8 Dec 03
Posts: 41
Credit: 873,780
RAC: 15
United States
Message 729475 - Posted: 23 Mar 2008, 15:42:27 UTC


Benchmark results:
Number of CPUs: 1
1964 floating point MIPS (Whetstone) per CPU
3359 integer MIPS (Dhrystone) per CPU

ID: 729475 · Report as offensive
Fred W
Volunteer tester

Send message
Joined: 13 Jun 99
Posts: 2524
Credit: 11,954,210
RAC: 0
United Kingdom
Message 729503 - Posted: 23 Mar 2008, 15:55:45 UTC - in response to Message 729475.  


Benchmark results:
Number of CPUs: 1
1964 floating point MIPS (Whetstone) per CPU
3359 integer MIPS (Dhrystone) per CPU


Ignore them. They don't bear the slightest resemblance to reality.

F.
ID: 729503 · Report as offensive
1mp0£173
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
United States
Message 729528 - Posted: 23 Mar 2008, 16:18:52 UTC - in response to Message 729503.  


Benchmark results:
Number of CPUs: 1
1964 floating point MIPS (Whetstone) per CPU
3359 integer MIPS (Dhrystone) per CPU


Ignore them. They don't bear the slightest resemblance to reality.

F.

That's a little harsh. They mean something, just not as much as we'd like.
ID: 729528 · Report as offensive
ChrisD
Volunteer tester

Send message
Joined: 25 Sep 99
Posts: 158
Credit: 2,496,342
RAC: 0
Denmark
Message 729540 - Posted: 23 Mar 2008, 16:39:12 UTC - in response to Message 729475.  
Last modified: 23 Mar 2008, 16:40:47 UTC


Benchmark results:
Number of CPUs: 1
1964 floating point MIPS (Whetstone) per CPU
3359 integer MIPS (Dhrystone) per CPU


???

You publish a set of benchmarks.. Now what?? What is Your Question?

Do You want us to make a guessing game? (We try to figure out what CPU You are using based on the figures).

This idea is because You have hidden Your Computer..

That must be it :) because without further info and a question, this is all I can think of. sorry.

Now, go and donate $10 to SETI@Home, and get Your little green star. :)

ChrisD
ID: 729540 · Report as offensive
Fred W
Volunteer tester

Send message
Joined: 13 Jun 99
Posts: 2524
Credit: 11,954,210
RAC: 0
United Kingdom
Message 729541 - Posted: 23 Mar 2008, 16:39:18 UTC - in response to Message 729528.  
Last modified: 23 Mar 2008, 16:40:04 UTC


That's a little harsh. They mean something, just not as much as we'd like.

But how much credence can we give them when my Q6600 running XP Home overclocked to 3336MHz reports 3273 FP and 7476 Int whereas another Q6600 running Vista and not overclocked (i.e. 2400Mhz) reports 6862 FP and 10828 Int?

Just didn't see much point encouraging a thread for swapping figures that are meaningless for comparing different machines.

F.
ID: 729541 · Report as offensive
1mp0£173
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
United States
Message 729600 - Posted: 23 Mar 2008, 19:33:35 UTC - in response to Message 729541.  


That's a little harsh. They mean something, just not as much as we'd like.

But how much credence can we give them when my Q6600 running XP Home overclocked to 3336MHz reports 3273 FP and 7476 Int whereas another Q6600 running Vista and not overclocked (i.e. 2400Mhz) reports 6862 FP and 10828 Int?

Just didn't see much point encouraging a thread for swapping figures that are meaningless for comparing different machines.

F.

I can't comment on the specifics without a whole lot more detail on this specific case.

But I can comment on the arguments over benchmarks over the past six decades or so.

A benchmark is a synthetic application designed to predict performance in the real world. It depends on a lot of factors, including how well the benchmark models the real world instruction mix.

If you're writing benchmarks in a high-level language, it's harder to produce a portable benchmark because compilers optimize. A really good optimizing compiler __should__ remove the actual benchmark code completely, running the "benchmark" in zero time.

What it does mean is if you take your machine running at "stock" and get the benchmarks, then increase the clock speed by 20%, you should see a roughly 20% increase in the benchmark numbers -- holding everything else equal, and assuming memory performance isn't measured (because the benchmark fits in the cache).

It doesn't mean very much when comparing apples and oranges (dissimilar systems).
ID: 729600 · Report as offensive
Profile Dr. C.E.T.I.
Avatar

Send message
Joined: 29 Feb 00
Posts: 16019
Credit: 794,685
RAC: 0
United States
Message 729605 - Posted: 23 Mar 2008, 20:00:25 UTC



The Dhrystone Benchmark checks repeated integer operations, and several operating system file handling operations.

Neither of the tests really checks how well/fast a system can access memory, and the SETI@Home Science Application, for example, accesses memory a lot.

Here is an example of memory introducing a delay. A Pentium 4 CPU of any speed, can calculate the sine of an angle in approximately 170 tics of its internal clock. It could have performed 170 regular integer additions in this time.

But if it wanted to do an integer addition on a number somewhere out in memory (say it was working on a table of numbers), the CPU might have to wait for this memory integer to be delivered to the CPU as much as 260 tics. So a badly timed integer+memory operation would take far longer than a sine calculation.

This is where Celeron CPUs can really slow down. Pentium has many more features

to predict when the CPU might be getting memory, and begins getting it long

before the CPU actually calculates with it.

Thus much less delay for most memory operations = faster.


BOINC Wiki . . .

Science Status Page . . .
ID: 729605 · Report as offensive
Iona
Avatar

Send message
Joined: 12 Jul 07
Posts: 790
Credit: 22,438,118
RAC: 0
United Kingdom
Message 729649 - Posted: 23 Mar 2008, 21:22:20 UTC - in response to Message 729541.  


That's a little harsh. They mean something, just not as much as we'd like.

But how much credence can we give them when my Q6600 running XP Home overclocked to 3336MHz reports 3273 FP and 7476 Int whereas another Q6600 running Vista and not overclocked (i.e. 2400Mhz) reports 6862 FP and 10828 Int?

Just didn't see much point encouraging a thread for swapping figures that are meaningless for comparing different machines.

F.



I suppose you're right, without knowing sytem specifics, they do appear meaningless. Here, I use my PC and my partner's PC (when available) to 'crunch' - they're both 3.2Ghz P4 Prescotts, on i875P chipset boards with 2 Gig of RAM. The 'Benchmarks' reckon that my partner's PC is a tad quicker, but my PC 'feels' quicker and always proves to be a bit quicker - especially when running on-line games (WoW) in spite of the older HIS X850XT Graphics card, compared to their newer HIS X1950 Pro....... Perhaps the RAM in my PC is playing a bigger part in this - OCZ Platinum EL, as opposed to 'standard' timed Hynix.




Don't take life too seriously, as you'll never come out of it alive!
ID: 729649 · Report as offensive

Message boards : Number crunching : benchmark tests


 
©2022 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.