Page 2 of 17 FirstFirst 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 LastLast
Results 21 to 40 of 332

Thread: G5 2.7GHz

  1. #21
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    Today I added two more Gurus 1GB RAM sticks (CL2.5) for a total of 6GB RAM. Added a UL3D-066 to a 100MHz slot. Installed the necessary ATTO software. Also cloned the OEM drive to a 300GB SATA MaxLineIII-16MB cache drive on the mobo bus via Tiger Disk Utility Restore. Before rebooting I repaired permissions on the new drive and did the Repair Disk which found one thing that I think would have prevented booting the new drive. Then selected the new drive and am using it as my primary drive and the OEM WD is my backup.

    Ran the Apple Hardware Test Quick Test, this time the same two drives but 6GB RAM which took 15m:24s compared to the previous Extended Test 1h:40m:03s, then compared the RAM sticks info and the new Gurus ones are exactly the same. k

  2. #22
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    I give up on improving Xbench and Cinebench results since 10.4.1 yields slower results for me with either. So I posted the 406.50 score to the Xbench G5 DB. Wait for 10.4.2? k

  3. #23
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    Comparing CINEBENCH 2003 v1 G5 (beta) previously tested with CINEBENCH 2003 v1 G5 on G5 2.7GHz DP, with Radeon X800 XT and Apple 23" Cinema HD display at 1920x1200, Millions.

    Two different configurations:
    C1: v1G5 (beta), 10.4
    C2: v1G5, 10.4.1

    CINEBENCH 2003 v1
    ************************************************** **
    Tester : kyum
    Processor : G5-2.7GHzDP
    MHz : 2700
    Number of CPUs : 2
    Operating System : C1-10.4, C2-10.4.1

    Graphics Card : ATI X800XT ME
    Resolution : <1920x1200>
    Color Depth :
    ************************************************** **
    Rendering (Single CPU): C1-385 CB-CPU, C2-386 CB-CPU
    Rendering (Multiple CPU): C1-709 CB-CPU, C2-701 CB-CPU

    Shading (CINEMA 4D): C1-361 CB-GFX, C2-364 CB-GFX
    Shading (OpenGL Software Lighting) : C1-1071 CB-GFX, C2-1071 CB-GFX
    Shading (OpenGL Hardware Lighting) : C1-1809 CB-GFX, C2-1833 CB-GFX
    ************************************************** **
    k

  4. #24
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    Some SCSI Burly results:
    G5-2.7GHzDP, 4xX15.3 18GB, UL4D 133MHz slot, 10.4.1, SoftRAID 3.1.3 with 64KB stripe units.

    QuickBenchª 2.1 Test Results File
    ©2003 Intech Software Corp.
    Test file created on Thursday, May 19, 2005 at 9:39:39 AM
    Test Volume name: 4xX15.3SRAID

    Extended Test Size: 20 MB Read: 247.865 MB/sec Write: 274.130 MB/sec
    Extended Test Size: 30 MB Read: 255.747 MB/sec Write: 286.763 MB/sec
    Extended Test Size: 40 MB Read: 269.492 MB/sec Write: 284.149 MB/sec
    Extended Test Size: 50 MB Read: 280.626 MB/sec Write: 282.390 MB/sec
    Extended Test Size: 60 MB Read: 279.521 MB/sec Write: 275.762 MB/sec
    Extended Test Size: 70 MB Read: 285.492 MB/sec Write: 279.890 MB/sec
    Extended Test Size: 80 MB Read: 283.978 MB/sec Write: 284.241 MB/sec
    Extended Test Size: 90 MB Read: 286.832 MB/sec Write: 280.280 MB/sec
    Extended Test Size: 100 MB Read: 286.209 MB/sec Write: 286.876 MB/sec
    -----------------------
    ZoneBench results:

    Test Date: Thursday, May 19, 2005 (10:26:41 AM)
    Test Volume Name: 4xX15.3SRAID
    Test Volume Size: 68.327 Gigabytes
    Transfer Size: 100 MBytes
    Zone Count: 7
    Iteration Count: 3

    Measurement Zone: 1
    First Measured Sector: 0
    Last Measured Sector: 204799
    Read Speeds Write Speeds
    Maximum: 282.007101 275.317510
    Average: 280.252956 260.299193
    Minimum: 277.549570 248.795828
    Iteration 1: 281.202196 248.795828
    Iteration 2: 277.549570 256.784240
    Iteration 3: 282.007101 275.317510

    Measurement Zone: 2
    First Measured Sector: 23881984
    Last Measured Sector: 24086783
    Read Speeds Write Speeds
    Maximum: 271.809299 269.020416
    Average: 271.702240 268.827293
    Minimum: 271.607755 268.490973
    Iteration 1: 271.607755 268.970489
    Iteration 2: 271.689665 269.020416
    Iteration 3: 271.809299 268.490973

    Measurement Zone: 3
    First Measured Sector: 47763968
    Last Measured Sector: 47968767
    Read Speeds Write Speeds
    Maximum: 266.096865 264.490789
    Average: 265.062657 262.652145
    Minimum: 264.323003 260.428196
    Iteration 1: 264.323003 263.037451
    Iteration 2: 264.768103 264.490789
    Iteration 3: 266.096865 260.428196

    Measurement Zone: 4
    First Measured Sector: 71645952
    Last Measured Sector: 71850751
    Read Speeds Write Speeds
    Maximum: 249.624939 252.781225
    Average: 248.942752 251.805259
    Minimum: 248.196234 250.511043
    Iteration 1: 248.196234 252.123510
    Iteration 2: 249.624939 252.781225
    Iteration 3: 249.007084 250.511043

    Measurement Zone: 5
    First Measured Sector: 95527936
    Last Measured Sector: 95732735
    Read Speeds Write Speeds
    Maximum: 226.032630 228.096987
    Average: 225.970158 227.242951
    Minimum: 225.905486 225.664356
    Iteration 1: 225.972359 225.664356
    Iteration 2: 225.905486 228.096987
    Iteration 3: 226.032630 227.967510

    Measurement Zone: 6
    First Measured Sector: 119409920
    Last Measured Sector: 119614719
    Read Speeds Write Speeds
    Maximum: 201.473984 200.396384
    Average: 200.925398 199.671597
    Minimum: 199.985601 199.106014
    Iteration 1: 201.316611 199.512392
    Iteration 2: 201.473984 199.106014
    Iteration 3: 199.985601 200.396384

    Measurement Zone: 7
    First Measured Sector: 143291904
    Last Measured Sector: 143496703
    Read Speeds Write Speeds
    Maximum: 184.361023 183.149848
    Average: 184.186133 181.859985
    Minimum: 183.839751 179.925043
    Iteration 1: 183.839751 183.149848
    Iteration 2: 184.361023 182.505065
    Iteration 3: 184.357624 179.925043

    k

  5. #25
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    My other SCSI Burly:
    4xX15.2 18GB, UL3D 100MHz slot, 10.4.1, SoftRAID 3.1.3 with 64KB stripe units.

    QuickBenchª 2.1 Test Results File
    ©2003 Intech Software Corp.
    Test file created on Friday, May 20, 2005 at 11:21:05 AM
    Test Volume name: 4xX15.2SRAID

    Extended Test Size: 20 MB Read: 213.294 MB/sec Write: 236.375 MB/sec
    Extended Test Size: 30 MB Read: 217.027 MB/sec Write: 229.900 MB/sec
    Extended Test Size: 40 MB Read: 222.879 MB/sec Write: 232.045 MB/sec
    Extended Test Size: 50 MB Read: 224.472 MB/sec Write: 230.031 MB/sec
    Extended Test Size: 60 MB Read: 227.363 MB/sec Write: 231.757 MB/sec
    Extended Test Size: 70 MB Read: 226.112 MB/sec Write: 231.578 MB/sec
    Extended Test Size: 80 MB Read: 227.208 MB/sec Write: 231.788 MB/sec
    Extended Test Size: 90 MB Read: 226.952 MB/sec Write: 231.859 MB/sec
    Extended Test Size: 100 MB Read: 227.763 MB/sec Write: 231.607 MB/sec

    k

  6. #26
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    Now both SCSI Burlys striped, 8x X15s, SoftRAID 3.1.3 with 64KB stripe units. UL4D 133MHz slot and UL3D-066 100MHz slot. 4 channels. 10.4.1.

    QuickBenchª 2.1 Test Results File
    ©2003 Intech Software Corp.
    Test file created on Friday, May 20, 2005 at 11:55:47 AM
    Test Volume name: 8xX15ul4133-ul3100G5

    Extended Test Size: 20 MB Read: 402.018 MB/sec Write: 463.993 MB/sec
    Extended Test Size: 30 MB Read: 425.677 MB/sec Write: 455.276 MB/sec
    Extended Test Size: 40 MB Read: 441.218 MB/sec Write: 478.291 MB/sec
    Extended Test Size: 50 MB Read: 435.816 MB/sec Write: 474.855 MB/sec
    Extended Test Size: 60 MB Read: 439.499 MB/sec Write: 474.807 MB/sec
    Extended Test Size: 70 MB Read: 444.232 MB/sec Write: 471.815 MB/sec
    Extended Test Size: 80 MB Read: 445.211 MB/sec Write: 470.610 MB/sec
    Extended Test Size: 90 MB Read: 450.150 MB/sec Write: 466.087 MB/sec
    Extended Test Size: 100 MB Read: 454.152 MB/sec Write: 467.827 MB/sec

    k

  7. #27
    Join Date
    Nov 2004
    Location
    Germany
    Posts
    2,352

    Default

    Last Friday the G5 died.
    One CPU was bad.

    Got a new one today from the dealer.

    No more fan cycling by now.

    BTW great results Kaye

    Regards

    Nicolas
    Custom Configurations! Rad Hacks and Mods!

  8. #28
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    Ouch Nicolas. I have been using mine a lot. TZ has talked me into an APC BACK-UPS RS 1500VA, Part #: BR1500. My son has a dealer discount with CDW and is checking a price for me today. Online prices look good too. Well below the $300 retail at APC.

    One of my SCSI Burlys is an 8-bay and I already have the 4 more X15.2 drives. Working on that setup. Just need some more parts to arrive to install them and try it out. Wish I had X15.4 drives or those new Atlas 15k drives. But can't have everything. k

  9. #29
    Join Date
    Nov 2004
    Location
    Germany
    Posts
    2,352

    Thumbs up 1500VA good decision

    I decided to buy the BRI 1500 too, instead of the 1000VA.

    BTW the BRI 1500 will be recognized after installing the PowerChute software.

    A friend of mine is using it for a G3 Beige server, a 15" TFT and a 8x HW RAID. We did a test and it has enough power to run this setup 15 Minutes from battery!

    The G5 was connected to an 2300VA UPC wich was left from the Octane and had an battery replacement two month ago.

    Guess what, the same memory wich was not working in the first 2.7GHz is working now in the new one.

    Regards

    Nicolas
    Custom Configurations! Rad Hacks and Mods!

  10. #30
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb

    Nicolas,

    Kaye has been beating his SCSI drive config into the ground or into submission (whichever comes first), to break 500MB/sec, but somehow, Ultra320 drives "prefer" Ultra320 controllers. Or so it seems. X15.3s don't like the UL3D-66 as much as they like the UL4D.

    If you are talking about six 15K II's, I think the UL3D would work with one drive per channel only.

    I cloned my system to a RAID. Quite nice. First time I used the Apple RAID default 32k stripe size. took a long time to clone, about 40 minutes. changed it over to 64k and cut the time in half. Weird? Also, cloning to X15.2 was slower than the Atlas 15K (which is comparable to X15.3 it seems). So when you do RAID your boot drive, or whatever, use the Apple disk utility Raid "options" to change the default.

    Crucial has 2GB DDR400 for only $240 now. Or, there is the premium "advanced mulit-bit ECC" for only $6K for 4GB (2x2). But it looks like the days of 16GB RAM G5 are at hand.

  11. #31
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    TZ,

    I'm stoked again. Check your email for the long answer, but short answer while I'm waiting for some parts is the ATTO Config Tool (for the UL4D anyway, I have to check the UL3D-66 with drives connected) can be set by SCSI ID for U320 (my X15.3s) or U160 (my X15.2s). Emailed ATTO Tech at 6am this morning and received a response at 8:30am about setting the sync by SCSI ID. You can even have a mixture of drives on the UL4D on the same channel or the other channel, some U320, some U160, and optimize the RAID by syncing each ID. k

  12. #32
    Join Date
    Nov 2004
    Location
    Germany
    Posts
    2,352

    Default

    There are two firmware / software switches / BIOS setups for the Seagate drives.

    One is optimized for seek/access times (Server).
    The other is optimized for throughput (Desktop).

    The firmware can be flashed using a PC only (Linux???).

    Some firmware versions are faster with ATTO cards.

    Regards

    Nicolas
    Custom Configurations! Rad Hacks and Mods!

  13. #33
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Default PowerMac 2.7GHz, Xeon 3.6GHz "just about the same"

    Review: 2.7GHz Mac, 3.6GHz PC "just about the same"

    A Digital Video Editing review compares Apple's Power Mac G5 dual 2.7GHz to a dual Xeon 3.6GHz computer from Dell. The review concludes that the Power Mac is only slightly slower than the fastest PC available, and costs less, too.

    On our battery of real-world After Effects benchmarks, the Mac won four of the tests, while the Xeon box won the other four. When you add up the total seconds required to complete six of our original benchmarks (numbered 1-6 here), the Mac needed 432 seconds to complete them, while the PC required 419. Looking further down the table, even with its juiced-up processors and that fancy NVIDIA graphics card, the Mac still couldn’t beat the PC in our CineBench tests.

    "You get speed that’s statistically just about the same as a PC, Bluetooth connectivity, FireWire 800 on the motherboard , and then there’s that striking industrial design and cachet that only goes with owning a Mac." To conclude, the reviewer writes, "I’d say that even though the Mac is still slower than the fastest PC in our benchmarks, for professional content creation it would be hard for me to pass it up."

    The Power Mac G5 receives a 9.5 rating. The article also looks as the 23-inch Cinema HD display. "The monitor has performed flawlessly and receives my highest recommendation, 10 out of 10 stars.


    Digital Video Editing Review

  14. #34
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    The best I could do with these old Cheetah X15.2 18GB U160 drives in combination with the X15.3 18GB U320 drives. I tried even more X15.2 drives which slowed the thruput down. Need more U320 drives, even later than the X15.3 model.

    G5 2.7GHz DP, 10.4.1, 4x X15.2 UL4D 2-channel 100MHz slot, 4x X15.3 UL4D 2-channel 133MHz slot, SoftRAID 3.1.3 striped, 64KB stripe units.

    QuickBench» 2.1 Test Results File
    ©2003 Intech Software Corp.
    Test file created on Saturday, May 28, 2005 at 5:19:34 AM
    Test Volume name: 8xX15SoftRAID64su2xUL4D

    Extended Test Size: 20 MB Read: 418.213 MB/sec Write: 515.809 MB/sec
    Extended Test Size: 30 MB Read: 428.268 MB/sec Write: 481.193 MB/sec
    Extended Test Size: 40 MB Read: 426.202 MB/sec Write: 478.279 MB/sec
    Extended Test Size: 50 MB Read: 438.950 MB/sec Write: 475.631 MB/sec
    Extended Test Size: 60 MB Read: 447.854 MB/sec Write: 474.552 MB/sec
    Extended Test Size: 70 MB Read: 456.740 MB/sec Write: 475.204 MB/sec
    Extended Test Size: 80 MB Read: 465.637 MB/sec Write: 473.684 MB/sec
    Extended Test Size: 90 MB Read: 465.517 MB/sec Write: 472.835 MB/sec
    Extended Test Size: 100 MB Read: 463.274 MB/sec Write: 472.158 MB/sec
    k

  15. #35
    Join Date
    Jan 2002
    Location
    NW Montana
    Posts
    8,197

    Default

    Looks to be screaming to me. - Randy

  16. #36
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    A primer on ATTO Configuration Tools. Maybe I'm the last to know. Wouldn't be the first time.

    An example screenshot of the ATTO Configuration Tool. Under Device listing on the left, the cards are listed in order of increasing slot number. So the UL3D-66 was in slot-3 and the UL4D in slot-4. This test of 12 X15s was not good.

    Selecting a channel, in this case Channel 2 of the UL4D under NVRAM on the right, I made no changes to the box labeled Common NVRAM Configuration. You need only confirm that QAS is enabled. Quick Arbitration and Selection (QAS): This reduces the overhead of control release on the SCSI bus from one device to another. This improvement reduces command overhead and maximizes bus utilization. QAS allows for increased overall system performance by providing a method for arbitration to occur without intervening BUS FREE phases. QAS can only be enabled if information unit transfers are enabled.

    The box below it was the cause of several emails back and forth with ATTO Tech. They were quickly responsive and detailed in their answers. Specified Target is the SCSI ID number, in this case the drive with ID4. Be sure that the 3 Enabled items in the pic are enabled.The UL4D and UL3D are the same until you get to Sync Rate (MB/s). The UL4D offers these options
    320 DT IU
    320 DT
    160 DT IU
    160 DT
    and lower

    The UL3D offers
    160 DT and lower

    DT stands for Double Transition, doubling the data rate without doubling the clock rate. Earlier versions of SCSI than U160 used ST.

    IU stands for Information Unit. Information units is a form of SCSI packetization or encapsulation that is designed to reduce overhead and speed data transfers. The packets speed flow of many items that formerly had to be sent back and forth individually. There are more benefits but my head hurts. IU is only valid with DT and is not a choice on the UL3D, only DT.

    Now when would I want to tune these settings by ID on a UL4D? Probably never unless you have a scanner connected to a channel. And then only to slow it down for that device. I was told that tuning older drives like the X15.2 U160 might, repeat might, show a slight increase in data thruput if U320 and U160 drives were mixed on the same channel. However the UL4D (and UL3D) is smart and does poll the drives and automatically set them for best thruput. But of course the best a UL3D can offer is 160 DT which hinders the X15.3 drive which can do U320 DT IU.

    So with the UL4D, normally use Defaults and then Commit for each channel. Then a reboot is required.

    A late model U160 drive might have more capabilities such as QAS and IU which would benefit from the UL4 series. Otherwise for a G5 and UL4D, a U320 drive is a much better choice. k

  17. #37
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default Widgets

    Anyone trying Dashboard Widgets? Probably mucking up 10.4.1 but some of them are helpful or interesting. I have about 30 so far. A few pics of some:


    k

  18. #38
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb

    Kaye,

    that is a nice little widget - SysStat.

    I haven't tried any so I guess I better start downloading

  19. #39
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    TZ,

    Launch Dashboard. I have it on the dock. After it launches, there is a circled X by the dock icon. Click that which opens the Widgets at the bottom to choose what you want on the Dashboard screen.

    When you plop (drag) one of those widgets onto the Dashboard screen by dragging it from the bottom widget icon list, you get the Core Video rippling water effect. Neat effect. k

  20. #40
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb

    Has anyone looked at Barefeats graphic test of three G5s with X800 XT?
    http://www.barefeats.com/threeway.html

    Weird.

    Maybe the faster bus is the problem of all things? Maybe Apple hit some kind of wall? Doesn't make sense.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •