Post your Cpumark99 scores

Page 5 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

sm625

Diamond Member
May 6, 2011
8,172
137
106
It's funny how this 15 year old benchmark perfectly showcases exactly where and how AMD screwed up. Computers are still single threaded beasts. If you get the single threaded uA correct, it is of course going to be a simple matter to scale it up to multiple cores.

"IPC increases! IPC increases! IPC increases!" lol what a tool.
 

abbcccus

Member
Feb 10, 2012
62
1
71
I ran it on all of my machines that have Windows installed and could easily be accessed at this point:
i5 2500k (4.5) 692
i7 860 (3.8) 592
i5 750 (3.6) 556
i3 550 (4.0) 612
i3 550 (3.2) 488
E5400 (2.7) 387
E3200 (2.4) 253
i3 2310M (2.2) 303
 

grkM3

Golden Member
Jul 29, 2011
1,407
0
0
Cant wait to see how ivy does clock for clock.

This program scales well with actuall ipc and dosnt care about memory timings or bandwith as I tried 5.2ghz at loose timings and 1800mhz on the ram and also tried same speed with tight timings at 2240mhz and scored within 1 point of each other.

801-802@5.2ghz
 

Kenmitch

Diamond Member
Oct 10, 1999
8,505
2,249
136
Cant wait to see how ivy does clock for clock.

This program scales well with actuall ipc and dosnt care about memory timings or bandwith as I tried 5.2ghz at loose timings and 1800mhz on the ram and also tried same speed with tight timings at 2240mhz and scored within 1 point of each other.

801-802@5.2ghz

I'm still waiting for somebody to break 900 :)

Looks to be next to impossible without LN2 and a 24kt chip are far as I can tell.
 

Miramonti

Lifer
Aug 26, 2000
28,651
100
91
i7-950: 647

(can't remember what my system is tho except it's air cooled, lol.)

Yes I am the one that makes the CPUmark99 post every few years. I've been benching my systems with this since my P5-90 that scored a 5.5. While it's integer only and not multithreaded it still provides a good indication of the strength of a core. And it's fun to compare systems from 20 years ago to today using the same benchmark.

Rig in sig - 640

I don't know who got me started, maybe you :), but I've been tracking this since I got 27 on a system, and then 36 on a pentium3? 450mhz workstation. It has been a very valuable tool for evaluating potential work upgrades, which involves a single-threaded database number-crunching app.

Cant wait to see how ivy does clock for clock.

This program scales well with actuall ipc and dosnt care about memory timings or bandwith as I tried 5.2ghz at loose timings and 1800mhz on the ram and also tried same speed with tight timings at 2240mhz and scored within 1 point of each other.

801-802@5.2ghz

I just read we're looking at late 2013 for Ivy Bridge. :(

Very cool to see some 800's! On my next upgrade, most likely Ivy Bridge, I'm determined to finally use some sort of water cooling this time.
 
Last edited:

Idontcare

Elite Member
Oct 10, 1999
21,118
58
91
Cant wait to see how ivy does clock for clock.

This program scales well with actuall ipc and dosnt care about memory timings or bandwith as I tried 5.2ghz at loose timings and 1800mhz on the ram and also tried same speed with tight timings at 2240mhz and scored within 1 point of each other.

801-802@5.2ghz

Given the age of the bench, I think the cache levels on our 2600K's now is roughly the same order of magnitude of the ram people had in their systems when the bench was released.

In other words, its probably 100% cached and never accesses the ram in the first place.
 

Chiropteran

Diamond Member
Nov 14, 2003
9,811
110
106
Q8200@ 2.8GHz = 400 even. I'm quoting this becuase the only thing I can see my quad being faster is on the bus speed (400mhz quad pumped).

It could just be margin of error. I've seen variance of 100 points between runs, just because of some background process.
 

gibbs007

Member
Dec 8, 2011
91
0
0
8150 @ 4.56Ghz = 507

fits right where I expected it. Pretty much a side grade from my 1055t @ 3.7ghz as expected, assuming the x6's show the same single core frequency scaling. Maybe I'll run the benchmark in a windows VM later on the 1055t.

Can you run FX 8150 with only 4 cores enabled to see if that helps:hmm:?
 

Kenmitch

Diamond Member
Oct 10, 1999
8,505
2,249
136
Did some testing with a 2550k the last day or two :)

5.3ghz

53ghz_2550k.png


5.4ghz

54ghz_2550k.png


Board in Sig struggles at these speeds. Pre-testing on a P8P67 Pro seemed to be much easier. My rig is water cooled so took the quick way out and only swapped in the 2550k instead of the combo.

My 2700k is a good clocker. This 2550k seems to be of similar quality if not slightly better. Couldn't get the 2700k to boot at 53x :D
 
Last edited:

SonicIce

Diamond Member
Apr 12, 2004
4,774
0
76