PDA

View Full Version : xBench 1.0 final released!


vniow
Jan 28, 2003, 03:51 PM
1.0 [2003-01-27]

* first non-beta release
* added "Model" detection
* added CPU Version detection
* made cache reporting more reliable on upgraded machines
* made processor type detection more reliable on upgraded machines
* corrected a crash while gathering cache information
* improved video card detection on a variety of machines
* fixed detection of dual processors
* fixed a problem causing the UI test to display progressively lower results each time it was run
* added progress sheet for results submission
* fixed a potential crash while submitting



http://www.xbench.com/

P-Worm
Jan 28, 2003, 04:30 PM
Very cool. Now I can show how much my dual Ghz kicks butt!

P-Worm

mac15
Jan 28, 2003, 04:48 PM
nice, It grabbed as soon as it was on versiontracker

my imac 700mhz, 512mbs ram scored a 61

ELYXR
Jan 28, 2003, 05:08 PM
My TiBook 400Mhz scored a 43. Nice app.

ryme4reson
Jan 28, 2003, 05:50 PM
98.35 is my score. 1 GIG RAM
Now we can all compare the machines with a standard test. Hopefully the 1.42DP will kick some serious butt

KVH
Jan 28, 2003, 09:02 PM
Scored a 70.52, 1.2GB Ram

Kid Red
Jan 28, 2003, 10:14 PM
111.86 (14 open apps, itunes playing)
116.28 (7 apps open, no itunes)

10.2.3

Kid Red
Jan 28, 2003, 10:15 PM
Curious to see how the 1.25ghz and 1.4ghz do compared to first gen dual gig.

edesignuk
Jan 28, 2003, 10:25 PM
...no apps open...

119.02

vniow
Jan 28, 2003, 10:35 PM
mine (http://ladd.dyndns.org/xbench/merge.xhtml?doc2=6982)

ryme4reson
Jan 28, 2003, 11:49 PM
These scores are interesting in comparing these machines:

700 iMac score= 61
233 MHZ and 37pts difference
933 1GIG RAM score = 98
1,077 MHZ and 21pts difference
DP 1GIG 1GIG RAM score = 119

Seems to me that second processor is a totaly waste of money, and I would rather get 233 MHZ increase on a single proc, then get another proc.

233/37= 6.26MHZ per Point
1077/21 = 51.28MHZ per Point

You do the math...

scmacdaddy
Jan 29, 2003, 02:22 AM
ibook 700 mhz 384 ram
score =54:confused:

digitalgiant
Jan 29, 2003, 04:50 AM
55.78:( Sombody should try this after defraging a hardrive. See if they get a better score. I would but i just did mine.

zoetropeuk
Jan 29, 2003, 08:30 AM
110.8

The scores for DP's should be much higher if both processors were utilised efficiently. If you have CPU monitor opened while running xbench you'll see that neither CPU is being used more than 50% of its capacity. If it does than the other processor suffers.
Even if you combine both processors they are never being used more than 50-55%, this is true for all the apps that I use. This even seems to be true for apps that push the fact that they utilise MP's.

onimac
Jan 29, 2003, 01:50 PM
Dual G4 867-768/ram

Very nice program.

Looks like my dual 867 performs like the other at the same speed.

ps! My first post;)

moby1
Jan 29, 2003, 02:13 PM
New 12" PowerBook arrived! This is my 1st post with it! 1st thing I did? Download Safari! 2nd? Download Xbench, here it is;


Results 74.20
System Info
Xbench Version 1.0
System Version 10.2.3
Physical RAM 256 MB
Model PowerBook6,1
Processor PowerPC G4 @ 867 MHz
Version 7455 (Apollo) v3.3
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 534 MHz
Bus Frequency 134 MHz
Video Card GeForce4 MX
Drive Type FUJITSU MHS2040AT D
CPU Test 93.09
GCD Recursion 91.85 3.59 Mops/sec
Floating Point Basic 93.69 315.80 Mflop/sec
AltiVec Basic 87.99 4.77 Gflop/sec
Floating Point Library 99.57 4.47 Mops/sec
Thread Test 55.37
Computation 53.74 432.97 Kops/sec, 4 threads
Lock Contention 57.10 716.69 Klocks/sec, 4 threads
Memory Test 80.82
System 71.03
Allocate 104.50 35.23 Kalloc/sec
Fill 68.18 390.96 MB/sec
Copy 55.56 277.82 MB/sec
Stream 93.75
Copy 96.50 422.00 MB/sec [altivec]
Scale 96.95 422.98 MB/sec [altivec]
Add 94.25 428.83 MB/sec [altivec]
Triad 87.87 386.10 MB/sec [altivec]
Quartz Graphics Test 87.71
Line 94.06 2.39 Klines/sec [50% alpha]
Rectangle 97.01 6.82 Krects/sec [50% alpha]
Circle 103.21 2.38 Kcircles/sec [50% alpha]
Bezier 87.05 945.82 beziers/sec [50% alpha]
Text 67.17 1.09 Kchars/sec
OpenGL Graphics Test 105.06
Spinning Squares 105.06 73.52 frames/sec
User Interface Test 66.54
Elements 66.54 22.64 refresh/sec
Disk Test 58.09
Sequential 58.06
Uncached Write 52.80 23.03 MB/sec [4K blocks]
Uncached Write 51.33 20.88 MB/sec [256K blocks]
Uncached Read 96.64 15.22 MB/sec [4K blocks]
Uncached Read 49.69 21.44 MB/sec [256K blocks]
Random 58.13
Uncached Write 68.44 1.03 MB/sec [4K blocks]
Uncached Write 51.53 11.82 MB/sec [256K blocks]
Uncached Read 57.15 0.37 MB/sec [4K blocks]
Uncached Read 57.82 11.33 MB/sec [256K blocks]


----------------------------------------------------------

I've only run it once, 256 MB (stock) I'll have 640 in it this weekend.

moby1

blogo
Jan 29, 2003, 03:21 PM
moby1, thats good, just what i needed to prove the barefeats article wrong.

Biggles
Jan 29, 2003, 03:25 PM
i got a 34 with my iMac. ouch

Edit: i tested it with some common apps running (safari, AIM, iTunes, and Mail), and it got a 23. double ouch

Foucault
Jan 29, 2003, 06:42 PM
New Powerbook 867/768 MB/40GB HD...

XBENCH score 79.94
CPU TEST 98.19
Thread Test 56.49
Memory Test 91.98
Quartz Graphics Test 88.05
Disk Test 54.64

Quiet an impressive score... indeed

MacBandit
Jan 29, 2003, 06:53 PM
XBench hasn't and doesn't support Dual Processor reporting. So what you see in the tests is the result of one processor.

If you want more information on XBench look at this thread (http://forums.macrumors.com/showthread.php?s=&threadid=11789) .

Also just to keep everything together please post all your XBench results in the previously mentioned thread.

ryme4reson
Jan 29, 2003, 06:56 PM
XBench hasn't and doesn't support Dual Processor reporting. So what you see in the tests is the result of one processor.

How does a DP score 119 and a 933 score 100 if its only showing 1 processor. I think want you tried to say is it doesnt effectively use the second proc, or it doesnt use the 2nd proc to its maximum.

MacBandit
Jan 29, 2003, 07:02 PM
Originally posted by ryme4reson
XBench hasn't and doesn't support Dual Processor reporting. So what you see in the tests is the result of one processor.

How does a DP score 119 and a 933 score 100 if its only showing 1 processor. I think want you tried to say is it doesnt effectively use the second proc, or it doesnt use the 2nd proc to its maximum.

No not at all. It doesn't use the second processor. Plain and simply the score you are referring to is the system score. The system score is influenced by all factors of the system (i.e. disc drive, vCard, system bus, etc.).

moby1
Jan 29, 2003, 11:19 PM
Originally posted by Eple
moby1, thats good, just what i needed to prove the barefeats article wrong.

I guess it means that you can't take just one review or benchmark (or opinion) as a complete evaluation of a product.

I do like to visit barefeats.com on a regular basis and I'm glad to see the site's back up.

Catfish_Man
Jan 29, 2003, 11:48 PM
http://ladd.dyndns.org/xbench/merge.xhtml?doc1=9145&doc2=6663

Take a look. It says they're running 10.2.5

Also:
1) The 1.25GHz G4s are revision 3.2, my dual 867 is 2.1, as are the 1GHz machines. They're NOT overclocked 867s!
2) the 12" Powerbook is rev 3.3
3) The 12" Powerbook has an underclocked L2 cache

digitalgiant
Jan 30, 2003, 12:42 AM
1 ghz PB,,,,,512 megs of ram and 1 gig of ram. thank you

nospleen
Jan 30, 2003, 08:50 AM
Originally posted by Foucault
New Powerbook 867/768 MB/40GB HD...

XBENCH score 79.94
CPU TEST 98.19
Thread Test 56.49
Memory Test 91.98
Quartz Graphics Test 88.05
Disk Test 54.64

Quiet an impressive score... indeed

I have an 867 PB too, but I only have 512mb of ram. My score was 80.25?:confused:

MacBandit
Jan 30, 2003, 10:04 AM
Originally posted by nospleen


I have an 867 PB too, but I only have 512mb of ram. My score was 80.25?:confused:

What was your hard drive score?

MacBandit
Jan 30, 2003, 10:06 AM
Originally posted by nospleen


I have an 867 PB too, but I only have 512mb of ram. My score was 80.25?:confused:

What was your hard drive score? It could be that his hard drive is closer to capacity then yours and therefore scores lower bringing down his total score. Also maybe he did the test without a fresh restart, without repairing permissions, with back ground apps running, the list goes on and on. This is why I wish people would read the XBench (http://forums.macrumors.com/showthread.php?s=&threadid=11789) thread. The XBench thread explains all of this and also how to get consistent scores.

Omen88
Jan 30, 2003, 03:06 PM
QS G4 733mhz 768MB RAM: 66.00
I tried a renice -20 on the process of xbench but it didn't make a difference.

greetz

Kid Red
Jan 30, 2003, 03:07 PM
Originally posted by digitalgiant
1 ghz PB,,,,,512 megs of ram and 1 gig of ram. thank you

81.56 with 1 gig ram.

ryme4reson
Jan 30, 2003, 03:12 PM
1 ghz PB,,,,,512 megs of ram and 1 gig of ram. thank you


81.56 with 1 gig ram.

I cant believe that the 1GHZ PBook scores an 81, and my 933 scores a 98. That is almost 20 points higher with a slower processor, and same bus. Also, the 1 GIG has a better vCard, as I have a GForceMX 64MB. We also have the same amounts of ram (1GIG). The only difference in my favor is my stock 60 GIG HD that spins at 7200 RPM. That cant be the reason why my machine is so much fast is it? I was trying to trade my 933 for a PBook, but if I going to take such a huge performace hit, then its not worth it.

Someone please enlighten me, as to the big difference. Thanks

Catfish_Man
Jan 30, 2003, 03:36 PM
Originally posted by ryme4reson
1 ghz PB,,,,,512 megs of ram and 1 gig of ram. thank you


81.56 with 1 gig ram.

I cant believe that the 1GHZ PBook scores an 81, and my 933 scores a 98. That is almost 20 points higher with a slower processor, and same bus. Also, the 1 GIG has a better vCard, as I have a GForceMX 64MB. We also have the same amounts of ram (1GIG). The only difference in my favor is my stock 60 GIG HD that spins at 7200 RPM. That cant be the reason why my machine is so much fast is it? I was trying to trade my 933 for a PBook, but if I going to take such a huge performace hit, then its not worth it.

Someone please enlighten me, as to the big difference. Thanks

The final XBench score doesn't measure speed. It measures the average of several categories of speed. Clearly, the laptop hard drive sucks compared to yours. This means that it will lag in application launch, system boot, and virtual memory performance. For non drive dependent tasks, it will win.

ryme4reson
Jan 30, 2003, 03:39 PM
I understand the score is an avg of all tests, but you think the hard drive can bring the score down 20 points?

The person with the 1 GIG PBook, can you post you entire score results? I am very curious about this. Thanks.

MacBandit
Jan 30, 2003, 04:46 PM
Originally posted by ryme4reson
I understand the score is an avg of all tests, but you think the hard drive can bring the score down 20 points?

The person with the 1 GIG PBook, can you post you entire score results? I am very curious about this. Thanks.

Yes the hard drive score could bring the overal score down that much. Also the vCard in the Powerbook even though it is based on a faster chip then your G4MX is still a mobil vCard which in the end is slower then your G4MX.

lmalave
Jan 30, 2003, 04:52 PM
Did anyone else see their scores go way down? I was scoring a 60 with the beta 5, and now I'm only scoring a 45 with the final version. The main culprit seems to be the CPU score, which went from about 90 to about 55. Has anywone else seen big changes like this?

Catfish_Man
Jan 30, 2003, 07:11 PM
Originally posted by Catfish_Man
http://ladd.dyndns.org/xbench/merge.xhtml?doc1=9145&doc2=6663

Take a look. It says they're running 10.2.5

Also:
1) The 1.25GHz G4s are revision 3.2, my dual 867 is 2.1, as are the 1GHz machines. They're NOT overclocked 867s!
2) the 12" Powerbook is rev 3.3
3) The 12" Powerbook has an underclocked L2 cache

Sorry to reply to my own post. I think it's more likely that XBench misreports the cache on rev 3.2 and higher G4s (it misreports the 1.25's cache too). The first two points still hold.

Kid Red
Jan 31, 2003, 09:13 AM
Originally posted by ryme4reson
I understand the score is an avg of all tests, but you think the hard drive can bring the score down 20 points?

The person with the 1 GIG PBook, can you post you entire score results? I am very curious about this. Thanks.

Results 81.56
System Info
Xbench Version 1.0
System Version 10.2.3
Physical RAM 1024 MB
Model PowerBook3,5
Processor PowerPC G4 @ 1.00 GHz
Version 7455 (Apollo) v3.2
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 667 MHz
L3 Cache 1024K @ 134 MHz
Bus Frequency 134 MHz
Video Card ATY,RV250M9
Drive Type FUJITSU MHS2060AT
CPU Test 101.14
GCD Recursion 106.39 4.15 Mops/sec
Floating Point Basic 99.45 335.20 Mflop/sec
AltiVec Basic 94.19 5.11 Gflop/sec
Floating Point Library 105.51 4.74 Mops/sec
Thread Test 58.96
Computation 57.86 466.10 Kops/sec, 4 threads
Lock Contention 60.10 754.42 Klocks/sec, 4 threads
Memory Test 90.91
System 93.09
Allocate 147.14 49.61 Kalloc/sec
Fill 140.61 806.32 MB/sec
Copy 54.59 272.96 MB/sec
Stream 88.83
Copy 88.59 387.41 MB/sec [altivec]
Scale 90.79 396.10 MB/sec [altivec]
Add 88.82 404.12 MB/sec [altivec]
Triad 87.18 383.09 MB/sec [altivec]
Quartz Graphics Test 112.51
Line 112.99 2.88 Klines/sec [50% alpha]
Rectangle 116.00 8.16 Krects/sec [50% alpha]
Circle 125.25 2.89 Kcircles/sec [50% alpha]
Bezier 106.77 1.16 Kbeziers/sec [50% alpha]
Text 103.95 1.69 Kchars/sec
OpenGL Graphics Test 82.39
Spinning Squares 82.39 57.66 frames/sec
User Interface Test 94.01
Elements 94.01 31.99 refresh/sec
Disk Test 61.29
Sequential 62.13
Uncached Write 57.51 25.09 MB/sec [4K blocks]
Uncached Write 56.04 22.80 MB/sec [256K blocks]
Uncached Read 91.90 14.48 MB/sec [4K blocks]
Uncached Read 54.73 23.61 MB/sec [256K blocks]
Random 60.47
Uncached Write 70.01 1.06 MB/sec [4K blocks]
Uncached Write 53.15 12.19 MB/sec [256K blocks]
Uncached Read 58.84 0.38 MB/sec [4K blocks]
Uncached Read 62.29 12.21 MB/sec [256K blocks]

James.Paul
Jan 31, 2003, 11:26 AM
Originally posted by zoetropeuk
110.8

The scores for DP's should be much higher if both processors were utilised efficiently. If you have CPU monitor opened while running xbench you'll see that neither CPU is being used more than 50% of its capacity. If it does than the other processor suffers.
Even if you combine both processors they are never being used more than 50-55%, this is true for all the apps that I use. This even seems to be true for apps that push the fact that they utilise MP's.

I've got a single processor 867 and i've just scored

Results 88.63
System Info
Xbench Version 1.0
System Version 10.2.3
Physical RAM 1024 MB
Model PowerMac3,5
Processor PowerPC G4 @ 867 MHz
Version 7450 (V'ger) v2.1
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 867 MHz
L3 Cache 2048K @ 217 MHz
Bus Frequency 133 MHz
Video Card GeForce2 MX
Drive Type IBM-IC35L060AVER07-0
CPU Test 97.54
GCD Recursion 101.49 3.96 Mops/sec
Floating Point Basic 97.16 327.46 Mflop/sec
AltiVec Basic 91.98 4.99 Gflop/sec
Floating Point Library 100.09 4.49 Mops/sec
Thread Test 56.56
Computation 55.10 443.90 Kops/sec, 4 threads
Lock Contention 58.10 729.34 Klocks/sec, 4 threads
Memory Test 92.10
System 101.59
Allocate 136.21 45.92 Kalloc/sec
Fill 93.03 533.49 MB/sec
Copy 87.42 437.08 MB/sec
Stream 84.23
Copy 84.85 371.04 MB/sec [altivec]
Scale 85.94 374.94 MB/sec [altivec]
Add 83.10 378.12 MB/sec [altivec]
Triad 83.10 365.16 MB/sec [altivec]
Quartz Graphics Test 105.02
Line 114.69 2.92 Klines/sec [50% alpha]
Rectangle 101.96 7.17 Krects/sec [50% alpha]
Circle 112.10 2.58 Kcircles/sec [50% alpha]
Bezier 106.77 1.16 Kbeziers/sec [50% alpha]
Text 92.61 1.51 Kchars/sec
OpenGL Graphics Test 110.74
Spinning Squares 110.74 77.50 frames/sec
User Interface Test 100.97
Elements 100.97 34.36 refresh/sec
Disk Test 85.20
Sequential 92.69
Uncached Write 82.67 36.07 MB/sec [4K blocks]
Uncached Write 85.94 34.96 MB/sec [256K blocks]
Uncached Read 134.87 21.24 MB/sec [4K blocks]
Uncached Read 83.27 35.93 MB/sec [256K blocks]
Random 78.84
Uncached Write 55.31 0.84 MB/sec [4K blocks]
Uncached Write 83.51 19.16 MB/sec [256K blocks]
Uncached Read 96.36 0.63 MB/sec [4K blocks]
Uncached Read 97.05 19.02 MB/sec [256K blocks]


looks like it does look at both processors!!!

mc68k
Jan 31, 2003, 11:32 AM
30.10 on my PM 9600/300

G4/450
768MB RAM
Radeon Mac Edition w/QE
Adaptec 29160N U160
Cheetah 15K HDD

mc68k
Jan 31, 2003, 07:45 PM
ran it a second time w/o QE:

only difference was the quartz test
w/QE over PCI: 30.31
w/o QE: 40.25

ryme4reson
Jan 31, 2003, 08:20 PM
Sounds like my 933 is still faster than the 1GIG, but I noticed that my 933 used DDR for the cache. and the 1GIG doesnt. My 933 is faster in all categories. I know the 1GIG is a PBook, but that is current generation, and my 933 is 2 generations old. I think its a shame that Apples products progess to slow.

Results 96.03
System Info
Xbench Version 1.0
System Version 10.2.3
Physical RAM 1024 MB
Model PowerMac3,5
Processor PowerPC G4 @ 934 MHz
Version 7455 (Apollo) v2.1
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 934 MHz
L3 Cache 2048K @ 234 MHz
Bus Frequency 133 MHz
Video Card GeForce4 MX
Drive Type IBM-IC35L060AVER07-0
CPU Test 106.90
GCD Recursion 109.68 4.28 Mops/sec
Floating Point Basic 105.39 355.18 Mflop/sec
AltiVec Basic 103.73 5.63 Gflop/sec
Floating Point Library 109.02 4.90 Mops/sec
Thread Test 60.88
Computation 59.72 481.10 Kops/sec, 4 threads
Lock Contention 62.09 779.32 Klocks/sec, 4 threads
Memory Test 110.11
System 136.30
Allocate 144.97 48.88 Kalloc/sec
Fill 248.73 1426.34 MB/sec
Copy 90.16 450.81 MB/sec
Stream 92.36
Copy 91.63 400.69 MB/sec [altivec]
Scale 94.17 410.87 MB/sec [altivec]
Add 91.38 415.77 MB/sec [altivec]
Triad 92.30 405.55 MB/sec [altivec]
Quartz Graphics Test 121.17
Line 131.60 3.35 Klines/sec [50% alpha]
Rectangle 119.03 8.37 Krects/sec [50% alpha]
Circle 133.99 3.09 Kcircles/sec [50% alpha]
Bezier 119.69 1.30 Kbeziers/sec [50% alpha]
Text 105.87 1.73 Kchars/sec
OpenGL Graphics Test 125.61
Spinning Squares 125.61 87.90 frames/sec
User Interface Test 114.74
Elements 114.74 39.04 refresh/sec
Disk Test 89.10
Sequential 95.02
Uncached Write 88.04 38.41 MB/sec [4K blocks]
Uncached Write 92.77 37.74 MB/sec [256K blocks]
Uncached Read 130.71 20.59 MB/sec [4K blocks]
Uncached Read 81.23 35.05 MB/sec [256K blocks]
Random 83.88
Uncached Write 62.05 0.94 MB/sec [4K blocks]
Uncached Write 93.11 21.36 MB/sec [256K blocks]
Uncached Read 95.60 0.62 MB/sec [4K blocks]
Uncached Read 96.43 18.90 MB/sec [256K blocks]

Catfish_Man
Jan 31, 2003, 11:06 PM
From the readme file:

"3. Thread Test
tests multiple processors - a test of multiple applications, or a single app doing multithreading
A) Computation - measures 4 worker threads performing integer operations, plus some memory bandwidth
- always faster on MP machines
B) Lock Contention - measures 4 threads quickly acquiring and releasing thread locks
- usually faster on MP machines
- locking performance may be important in certain types of multithreaded code.

The thread tests aren't representative of all types of threaded code, but they do measure some of the factors that affect threaded applications.
"

So dual machines will score higher on this section of the benchmark.

AssassinOfGates
Feb 1, 2003, 01:12 AM
Dual 867, 512 ram, super drive. Stock Geforce 4 mx (32). Running 10.2.3 and some Apache server additionals.

Results 116.26
System Info
Xbench Version 1.0
System Version 10.2.3
Physical RAM 512 MB
Model PowerMac3,6
Processor PowerPC G4x2 @ 867 MHz
Version 7455 (Apollo) v2.1
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 867 MHz
L3 Cache 1024K @ 217 MHz
Bus Frequency 134 MHz
Video Card GeForce4 MX
Drive Type ST360021A
CPU Test 108.60
GCD Recursion 109.01 4.26 Mops/sec
Floating Point Basic 109.48 369.00 Mflop/sec
AltiVec Basic 110.75 6.01 Gflop/sec
Floating Point Library 105.31 4.73 Mops/sec
Thread Test 110.37
Computation 109.66 883.45 Kops/sec, 4 threads
Lock Contention 111.09 1.39 Mlocks/sec, 4 threads
Memory Test 114.61
System 110.91
Allocate 106.48 35.90 Kalloc/sec
Fill 176.01 1009.34 MB/sec
Copy 83.49 417.46 MB/sec
Stream 118.57
Copy 118.06 516.28 MB/sec [altivec]
Scale 120.73 526.75 MB/sec [altivec]
Add 121.23 551.58 MB/sec [altivec]
Triad 114.52 503.19 MB/sec [altivec]
Quartz Graphics Test 114.82
Line 113.37 2.89 Klines/sec [50% alpha]
Rectangle 115.93 8.16 Krects/sec [50% alpha]
Circle 120.78 2.78 Kcircles/sec [50% alpha]
Bezier 109.72 1.19 Kbeziers/sec [50% alpha]
Text 114.83 1.87 Kchars/sec
OpenGL Graphics Test 111.21
Spinning Squares 111.21 77.82 frames/sec
User Interface Test 106.56
Elements 106.56 36.26 refresh/sec
Disk Test 85.36
Sequential 86.36
Uncached Write 80.99 35.33 MB/sec [4K blocks]
Uncached Write 83.57 34.00 MB/sec [256K blocks]
Uncached Read 103.31 16.27 MB/sec [4K blocks]
Uncached Read 81.15 35.01 MB/sec [256K blocks]
Random 84.39
Uncached Write 80.95 1.22 MB/sec [4K blocks]
Uncached Write 75.91 17.41 MB/sec [256K blocks]
Uncached Read 90.70 0.59 MB/sec [4K blocks]
Uncached Read 92.19 18.07 MB/sec [256K blocks]

MacBandit
Feb 1, 2003, 01:15 AM
Originally posted by ryme4reson
Sounds like my 933 is still faster than the 1GIG, but I noticed that my 933 used DDR for the cache. and the 1GIG doesnt.

The current PowerBooks have L2 and L3 cache which are both DDR memory.


Kid Red did you run the test with your PowerBook plugged in? If not your processor was speed stepping bringing down overall scores.

ryme4reson
Feb 1, 2003, 01:19 AM
This is the results for my machine, look at the speed.
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 934 MHz
L3 Cache 2048K @ 234 MHz

here is the 1GIG results
Processor PowerPC G4 @ 1.00 GHz
Version 7455 (Apollo) v3.2
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 667 MHz
L3 Cache 1024K @ 134 MHz

but look at the DP 867, the bus speeds are also faster
Processor PowerPC G4x2 @ 867 MHz
Version 7455 (Apollo) v2.1
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 867 MHz
L3 Cache 1024K @ 217 MHz

MacBandit
Feb 1, 2003, 01:28 AM
Originally posted by ryme4reson
This is the results for my machine, look at the speed.
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 934 MHz
L3 Cache 2048K @ 234 MHz

here is the 1GIG results
Processor PowerPC G4 @ 1.00 GHz
Version 7455 (Apollo) v3.2
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 667 MHz
L3 Cache 1024K @ 134 MHz

but look at the DP 867, the bus speeds are also faster
Processor PowerPC G4x2 @ 867 MHz
Version 7455 (Apollo) v2.1
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 867 MHz
L3 Cache 1024K @ 217 MHz

Just because it's running slower then CPU speed for the L2 cache does not mean it isn't DDR ram. It has to be DDR it's running faster then bus speed. The speed has a lot to do with how much money Apple wanted to spend on it and the power consumption of the processor they are using. It's normal for low power chips to use smaller and slower memory. The faster the memory the higher the power consumption.

You can even check Apples specs for the PowerBooks the L2 and L3 cache are both listed as DDR.

ryme4reson
Feb 1, 2003, 01:33 AM
It has to be DDR it's running faster then bus speed.

First of all here are the speeds for the 1GIG
L2 Cache 256K @ 667 MHz
L3 Cache 1024K @ 134 MHz

It is NOT running faster than bus speed, secondly, I am under the impression that the G4's and G3's are the same processors in the desktop, and laptop, so there should be no difference.

My point is just to show some concern about the 1 GIG's speed. I think the DP 867 and my 933 performed well, but I am not impressed with the 1GIG.

MacBandit
Feb 1, 2003, 01:48 AM
Originally posted by ryme4reson


First of all here are the speeds for the 1GIG
L2 Cache 256K @ 667 MHz
L3 Cache 1024K @ 134 MHz

It is NOT running faster than bus speed, secondly, I am under the impression that the G4's and G3's are the same processors in the desktop, and laptop, so there should be no difference.

My point is just to show some concern about the 1 GIG's speed. I think the DP 867 and my 933 performed well, but I am not impressed with the 1GIG.

I'm sorry but the 7455 in the laptops is a different processor. If you are dissapointed in the speed of a 1Gig because of a test in a PowerBook don't be. It could simply be that the test was ran with the PowerBook running on battery which would put the CPU in speed stepping mode. Also I have a Dual/GHz/DDR. If you want to see a test of it check this out my overall score is 136.

http://forums.macrumors.com/report.php?s=&postid=193639

I will post new tests with XBench 1.0 as soon as I get my system reinstall finished. The true speed of a Dual can not be tested by XBench. The real speed comes when you are running multiple programs, that is when a dual and OSX really shine.

iStream
Feb 1, 2003, 03:54 PM
You say you want to see a Powerbook 1GHz, 1GB RAM... here it is:

Overall: 88.96
CPU Test: 116.08
Thread Test: 64.69
Memory Test: 97.37
Quartz Graphics Test: 121.63
OpenGL Graphics Test: 96.17
User Interface Test: 109.43
Disk Test: 60.30 <--- Sucks

iStream
Feb 1, 2003, 04:02 PM
Ryme4Reson... your 933 scored a 106 on the CPU Test and my 1GHz Powerbook scored a 116 on the CPU Test. The only reason your overall speed is faster is because of the disk test... (mine is a laptop, yours is a desktop.. expect it) so I don't know what you're bitchin about.

ryme4reson
Feb 1, 2003, 06:25 PM
Here is a score from a 1GIG PBook
Results 81.56
System Info
Xbench Version 1.0
System Version 10.2.3
Physical RAM 1024 MB
Model PowerBook3,5
Processor PowerPC G4 @ 1.00 GHz
Version 7455 (Apollo) v3.2
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 667 MHz
L3 Cache 1024K @ 134 MHz
Bus Frequency 134 MHz
Video Card ATY,RV250M9
Drive Type FUJITSU MHS2060AT
CPU Test 101.14

Here the CPU scored 101.14. My 933 scored 106.90. I know that is a small sample size but I won't think the 1GIG would score slower on the CPU test. Its good to hear your CPU scored significantly higher than this CPU.

Catfish_Man
Feb 1, 2003, 09:04 PM
Originally posted by ryme4reson
Here is a score from a 1GIG PBook
Results 81.56
System Info
Xbench Version 1.0
System Version 10.2.3
Physical RAM 1024 MB
Model PowerBook3,5
Processor PowerPC G4 @ 1.00 GHz
Version 7455 (Apollo) v3.2
L1 Cache 32K (instruction), 32K (data)
L2 Cache 256K @ 667 MHz
L3 Cache 1024K @ 134 MHz
Bus Frequency 134 MHz
Video Card ATY,RV250M9
Drive Type FUJITSU MHS2060AT
CPU Test 101.14

Here the CPU scored 101.14. My 933 scored 106.90. I know that is a small sample size but I won't think the 1GIG would score slower on the CPU test. Its good to hear your CPU scored significantly higher than this CPU.

I would imagine that testing differences could EASILY account for a +/- 5 difference. Perhaps even a +/- 10. Trust me, you won't be able to tell the difference if the scores are that close (I've heard that anything under a 30% difference isn't noticable). The slower hard drive will be much much more noticable.

Kid Red
Feb 1, 2003, 11:39 PM
Originally posted by MacBandit


The current PowerBooks have L2 and L3 cache which are both DDR memory.


Kid Red did you run the test with your PowerBook plugged in? If not your processor was speed stepping bringing down overall scores.

Yes, it was plugged in. Was the score low for the pbook gig?

Kid Red
Feb 1, 2003, 11:44 PM
Originally posted by MacBandit


I'm sorry but the 7455 in the laptops is a different processor. If you are dissapointed in the speed of a 1Gig because of a test in a PowerBook don't be. It could simply be that the test was ran with the PowerBook running on battery which would put the CPU in speed stepping mode. Also I have a Dual/GHz/DDR. If you want to see a test of it check this out my overall score is 136.

http://forums.macrumors.com/report.php?s=&postid=193639

I will post new tests with XBench 1.0 as soon as I get my system reinstall finished. The true speed of a Dual can not be tested by XBench. The real speed comes when you are running multiple programs, that is when a dual and OSX really shine.

136? Damn, I onlt got a 116 with my dual gig. I did have a few apps open but not doing anything, does that affect the score that much? You were 20 points higher then mine. Ah, you have the MDD dual gig? That's why.

iStream
Feb 2, 2003, 04:07 PM
1 GHz Powerbook... CPU Test, 116. I've ran the thing 50 times its 116 every time. So whoever got those results something is messed up!

MacBandit
Feb 2, 2003, 05:06 PM
Originally posted by Kid Red


136? Damn, I onlt got a 116 with my dual gig. I did have a few apps open but not doing anything, does that affect the score that much? You were 20 points higher then mine. Ah, you have the MDD dual gig? That's why.

Yes having other open apps will greatley affect the score. To get the best and truest test. Do a permissions repair and then a restart. Be sure you don't have any apps that start up automatically even ones that run in the finder menu bar. If you do have such apps disable there auto startup and restart. Once restarted run XBench before you do anything else. Do not touch the computer for any reason while it runs.

Here is my score running XBench 1.0.

http://forums.macrumors.com/showthread.php?s=&postid=239741#post239741