Sign in with
Sign up | Sign in
Your question
Closed

Help identify old P4.

Tags:
Last response: in CPUs
Share
November 9, 2012 7:30:18 AM

...and choose an upgrade :) 

i'm thinking about putting in a new cpu, mostly with better cache to help compile times until i have the money/time to change my main system to a i5 or i7 (or who knows AMD will have something by then)

anyway, i've seem some top-of-the-line-in-2004-p4 in ebay for around $50...

This is the mobo i have: http://www.asus.com/Motherboards/Intel_Socket_775/P5GPL...

and this is the CPU i haveL
  1. hobo$ cat /proc/cpuinfo
  2. processor : 0
  3. vendor_id : GenuineIntel
  4. cpu family : 15
  5. model : 4
  6. model name : Intel(R) Pentium(R) 4 CPU 3.00GHz
  7. stepping : 9
  8. microcode : 0x3
  9. cpu MHz : 3010.900
  10. cache size : 1024 KB
  11. physical id : 0
  12. siblings : 1
  13. core id : 0
  14. cpu cores : 1
  15. apicid : 0
  16. initial apicid : 0
  17. fdiv_bug : no
  18. hlt_bug : no
  19. f00f_bug : no
  20. coma_bug : no
  21. fpu : yes
  22. fpu_exception : yes
  23. cpuid level : 5
  24. wp : yes
  25. flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm constant_tsc up pebs bts pni dtes64 monitor ds_cpl cid cx16 xtpr lahf_lm
  26. bogomips : 6021.80
  27. clflush size : 64
  28. cache_alignment : 128
  29. address sizes : 36 bits physical, 48 bits virtual
  30. power management:
  31.  
  32. # dmidecode 2.11
  33. SMBIOS 2.3 present.
  34.  
  35. Handle 0x0004, DMI type 4, 35 bytes
  36. Processor Information
  37. Socket Designation: LGA 775
  38. Type: Central Processor
  39. Family: Pentium 4
  40. Manufacturer: Intel
  41. ID: 49 0F 00 00 FF FB EB BF
  42. Signature: Type 0, Family 15, Model 4, Stepping 9
  43. Flags:
  44. FPU (Floating-point unit on-chip)
  45. VME (Virtual mode extension)
  46. DE (Debugging extension)
  47. PSE (Page size extension)
  48. TSC (Time stamp counter)
  49. MSR (Model specific registers)
  50. PAE (Physical address extension)
  51. MCE (Machine check exception)
  52. CX8 (CMPXCHG8 instruction supported)
  53. APIC (On-chip APIC hardware supported)
  54. SEP (Fast system call)
  55. MTRR (Memory type range registers)
  56. PGE (Page global enable)
  57. MCA (Machine check architecture)
  58. CMOV (Conditional move instruction supported)
  59. PAT (Page attribute table)
  60. PSE-36 (36-bit page size extension)
  61. CLFSH (CLFLUSH instruction supported)
  62. DS (Debug store)
  63. ACPI (ACPI supported)
  64. MMX (MMX technology supported)
  65. FXSR (FXSAVE and FXSTOR instructions supported)
  66. SSE (Streaming SIMD extensions)
  67. SSE2 (Streaming SIMD extensions 2)
  68. SS (Self-snoop)
  69. HTT (Multi-threading)
  70. TM (Thermal monitor supported)
  71. PBE (Pending break enabled)
  72. Version: Intel(R) Pentium(R) 4 CPU 3.00GHz
  73. Voltage: 1.4 V
  74. External Clock: 200 MHz
  75. Max Speed: 4000 MHz
  76. Current Speed: 3000 MHz
  77. Status: Populated, Enabled
  78. Upgrade: Other
  79. L1 Cache Handle: 0x0005
  80. L2 Cache Handle: 0x0006
  81. L3 Cache Handle: 0x0007
  82. Serial Number: To Be Filled By O.E.M.
  83. Asset Tag: To Be Filled By O.E.M.
  84. Part Number: To Be Filled By O.E.M.


according to wikipedia, i have the top of the line of the non-HT prescotts... even though it does not make sense... there's no match for this:
- 1mb L2 cache
- no HT
- only SSE2

I will ignore the SS3 mentioned for this model and assume i have this one:
http://en.wikipedia.org/wiki/List_of_Intel_Pentium_4_mi...
(family 15, model 4, 3ghz, 1mb L2 cache)
my other guess is the northwood, but they only have 512k cache, and are from another family number from what my chip is reporting.

the best i seems to be able to run on that board is the HT+vtx prescott extreme.
i've found two for $50 each on ebay...

the 662 and the 667
http://www.cpu-world.com/Compare/402/Intel_Pentium_4_66...

...also found some cheaper ones, 661. it's the Cedar Mill (65 nm) Intel Family 15 Model 6
...apparently the exact same performance as the 667 but 2/3 of the power wasted... is that so? (and maybe no vtx, but i don't really need it)

So those are pretty much the upgrade i can make: (assuming i found my cpu correctly)
http://www.cpu-world.com/Compare/747/Intel_Pentium_4_51...

is it even worth the trouble? will probably use it for 6mo to 1yr.

More about : identify

a b à CPUs
November 9, 2012 7:57:47 AM

I would say there isn't too much faster on your chipset. You really have a limited mileage with P4 systems. I think it's worth saving that money and investing more in your next system.
Score
0
a b à CPUs
November 9, 2012 8:06:26 AM

Yup live with it for a bit longer, save the cash,and buy a whole new setup.
Score
0
Related resources
November 9, 2012 10:32:48 AM

Try to get a software , name : CPU-Z ,
to identify it ~

If u unable to identify it ~
Going to get a new PC , better go for core i series , dont buy a core 2 duo , by then , u will realize u waste money on it , as it is no longer is upgradeable ~
Score
0
a c 116 à CPUs
November 9, 2012 11:43:29 AM

vipervoid1 said:
Going to get a new PC , better go for core i series , dont buy a core 2 duo , by then , u will realize u waste money on it , as it is no longer is upgradeable ~

A platform does not need to be upgradable to still be a massive upgrade for whatever the OP currently has. Almost any Core2Duo would already be 3-4X faster than his current PC.

Also, unless you grossly under-buy a new LGA1155 or LGA2011 build, no CPU worth upgrading to is going to come out during the socket's 3-4 years commercial lifespan so although an i5/i7 build may be "upgradable", there won't be anything worth upgrading them to unless you made the mistake of buying a Celeron or Pentium when your usage patterns warranted an i5 or better in the first place. Not many people with i5-2xxx will ever bother upgrading to i5-3xxx... most people don't upgrade CPUs at all between new PCs.

The main thing I would be concerned about with buying an used C2D (after DOA) would be having sufficient RAM included in it if the system uses DDR2 since upgrading that with fresh parts can cost almost as much as a whole brand-new PC.
Score
0
a b à CPUs
November 9, 2012 12:39:15 PM

There is nothing you can really upgrade to I mean it doesn't support Pentium D's or Core 2 so you might as well just relax and save up for something new.
Score
0
November 9, 2012 6:36:32 PM

good comments all around.

about cpu-z and such... i can't run windows on that box. linux only.

about upgrading the mobo, i will probably do it. but for ~$100~£50 i probably can't get a board with the same SATA ports for my setup nor the same high end cooler, nor the same RAM... then prices add up. or i have to use a noise cooler, less RAID redundancy, etc.

i was thinking about the CPU because it's a 15min upgrade that doesn't impact anything else with hidden costs and time sink. as upgrading a MOBO always is... (i will do that later on, just not now)

what i hope is to get my 30~40s compile time to some 15~25s or so. but i can't find a single benchmark for that kind of load on those CPUs... why every damn hardware site only care about the game of the month? :??: 

Score
0
a c 116 à CPUs
November 9, 2012 7:16:23 PM

gcb said:
what i hope is to get my 30~40s compile time to some 15~25s or so. but i can't find a single benchmark for that kind of load on those CPUs... why every damn hardware site only care about the game of the month? :??: 

Compiling stuff is one of the things Intel's Netburst architecture sucked the most at since parsing/compiling code has extremely unpredictable branching and calling behavior and this brings out the worst out of Netburst's heavy pipeline stall penalties from branch mispredicts.

If a P4 can compile a given project in 60 seconds, a similarly clocked Core2 or i3/i5 would likely be done doing the same in less than 20 seconds, even faster if your rebuilds often span multiple object files and you use make -jNN to build them in parallel. Another possible speedup would be less file reloading/swapping due to having more RAM for work, cache and bumping the number of parallel jobs.
Score
0
November 10, 2012 1:29:12 AM

Best answer selected by gcb.
Score
0
!