Sandra reference systems accuracy?

Post here if you just want to chat about... whatever.

Moderator: The Mod Squad

Sandra reference systems accuracy?

Postby useless comment » Wed May 29, 2002 5:38 pm

I was wondering if any has done any real world comparisons to the reference system's ratings for benchmarking in Sandra. What I mean is if you have a P4 2Gig, how do your scores compare to the reference P4 2Gig in the CPU Arithmetic Benchmark. If you have an xp1800+, how does it compare to the reference 1800+ in the CPU Multimedia Benchmark? Or if you just happen to have the same chipset/cpu/ram type setup in the Memory Bandwidth Benchmark, how do you fare against it?

I'm using Sandra Pro 2002 and it has no setups comparable to mine(Duron 1GHz, KT266A, 128MB PC2100). The CPU Arithmetic Bench shows my CPU closely rated to -but with moderately higher scores than- the Athlon 1GHz and a P3 1Ghz. My Duron severely smashes the P4 1.2GHz in the Dhrystone ALU but is slightly egded out on Whetstone FPU.

The results are roughly the same with the Multimedia Bench but the Athlon 1GHz and P3 1GHz do a little better than me in Floating Point. Oddly enough my little Duron completely disrespects the P4 1.2GHz and is not too far off from the P4 1.6.

The Memory Bench is where I'm at a loss because there is no way for me to compare with such wildly different chipset/cpu/memory type setups. As far as the ratings go my system closely resembles the results for the AMD 760 PC1600R CL2 ECC DDR on an XP 1400(but again mine shows modestly better). The only KT266A system used an XP 1800 which severely destroys my will to live(but then I realize I can get an 1800+ for $105 at newegg).

Anyway, if anyone really cares, I'm just looking for real or closer comparisons if anyone has some.
useless comment
Black Belt 1st Degree
Black Belt 1st Degree
 
Posts: 1344
Joined: Wed Mar 14, 2001 1:01 am

Postby Aibohphobia » Wed May 29, 2002 6:50 pm

Mine gets roughly the same, but a little lower, on the CPU arithmetic benchmark with the same CPU. On the rest mine are pretty much equal too, but a bit below because I can't help but move the mouse and press buttons on the keyboard when it tells me not to. I believe why you would be getting the same score as a technically-better Athlon 1.0ghz is that its method for testing matters more on the mhz and not on the L cache which makes the Athlon better.
Athlon xp1800/ via kt266a/ 256mb ddr2100 ram :) I can't get a picture though..
Last edited by Aibohphobia on Thu May 30, 2002 2:11 pm, edited 1 time in total.
Go hang a salami; I'm a lasagna hog!
A man, a plan, a butt tub: anal Panama!
Goddam mad dog!
Aibohphobia
Black Belt
Black Belt
 
Posts: 755
Joined: Fri Dec 21, 2001 1:01 am

Postby useless comment » Wed May 29, 2002 7:04 pm

Another problem is not knowing what the rest of the specs are on the system they tested the CPU with. The results for the Athlon 1Gig could have been obtained on an AMD chipset or a lowly KT133.

Thanks for the response Mr. Taylor. If you get back here would you be opposed to editing in your specs(cpu/chipset/ram type)? If not, that's okay, I'm just wanting to gather the info for future reference.
useless comment
Black Belt 1st Degree
Black Belt 1st Degree
 
Posts: 1344
Joined: Wed Mar 14, 2001 1:01 am

Postby Tolemac » Thu May 30, 2002 2:33 am

Here's my SANDRA numbers for the CPU arithemtic benchmark:

Image

Here's the CPU Multimedia:
Image

And the Memory benchmarks:
Image

That's my 1700+ nicely overclocked and I have 512Mb Crucial PC2100 overclocked as well. This is on a Soltek SL-75DRV4 KT266A motherboard.

I find SANDRA to be well within expected error limits. I've used it on many different systems and I get pretty much what I expect from the benchmarks. Some are a little high, like the FSB for example. Mine is at 146, but it shows as 148 to SANDRA. Everything else is spot on. 8)
Last edited by Tolemac on Thu May 30, 2002 1:02 pm, edited 1 time in total.
Tolemac
Mobo-fu Master
Mobo-fu Master
 
Posts: 14276
Joined: Tue Sep 12, 2000 12:01 am
Location: L-1 Bridge

Postby useless comment » Thu May 30, 2002 12:02 pm

excellent tolemac! that kt266a is a bad@ss, eh?

[edit]Tolemac, I don't know what I was thinking. I think the color through me off.[/edit]

hopefully this could help some other people out as well
Last edited by useless comment on Thu May 30, 2002 2:35 pm, edited 1 time in total.
useless comment
Black Belt 1st Degree
Black Belt 1st Degree
 
Posts: 1344
Joined: Wed Mar 14, 2001 1:01 am

Postby Tolemac » Thu May 30, 2002 1:03 pm

That's 2002 Pro. A very good investment I must say. 8)
Tolemac
Mobo-fu Master
Mobo-fu Master
 
Posts: 14276
Joined: Tue Sep 12, 2000 12:01 am
Location: L-1 Bridge


Return to The Water Cooler

Who is online

Users browsing this forum: No registered users and 1 guest