Encoding Times with 2 Diff. Systems

**thesmallelder** · 26 Apr 2007, 10:49 AM

I think the processor difference is what equates to a majority of the speed boost in the first system. The E6400 is considerably superior to the X2 4600, which is why the encoding phase was so much quicker on the Intel. Having 2 hard disks certainly helps as well, though.

**Art Vandelay** · 26 Apr 2007, 12:08 PM

Well I wondered when I was watching the fps in HCenc with the E6400. I went to Tomshardware and checked the 2 systems with a benchmard of a DVD5 transcode and they were bascially the same with 800MHz RAM (this test used 533MHz). The AMD is at 2.4GHz and the Intel is at 2.1GHz I think.

I remember the AMD was a higer speed processor, but Tomshardware had them pratically identical with the benchmark within 1 second apart for the transcode of a DVD5. I thought it was more from the extra 1 GB of DDR2 RAM, the encode phase also was helped by the 2nd HDD as jdobbs mentioned the HDD head movement will be less with the 2nd drive and it will save some time. I just had the 2 systems and the E6400 was only in my house for a week because the Media Center pic. quality was terrible (TV Tuner card was real bad) and the DVD-RAM drive burnt very poorly and failed most times the firmware couldn't be updated eventhough it was 1 version behind and HP had it for download on the website, it was a Samsung drive rebranded by HP.

**techreactor** · 26 Apr 2007, 02:31 PM

Your tests are very vague since you have done it on 2 systems with totally different configurations and what do you expect a system with lower configuration should win ???.

If you want to do a more conclusive test, you should pick a single system(out of two you mentioned) and then perform test with only one HDD configuration and then with 2 HDD configuration. That would atleast bring out some conclusive results.

More processing power and RAM definitely speeds up a system and thats no rocket science.

**blutach** · 26 Apr 2007, 04:10 PM

For that 2nd system, your speeds are a bit slow - 14 mins to rebuild is very long (as is 5 mins to prepare). I have an 3800x2 OC'ed by 25% with 1Gb of Corsair (2-3-3-6) and get 8-9 mins tops rebuild and 3 mins prep. With HC, I encode regularly at 33-35 PAL fps (you did about that in NTSC with the 4600).

Not sure that 2nd system is optimised.

Regards

**ipaulo** · 26 Apr 2007, 07:31 PM

Art, so the set-up for system 1 is faster than having the source and the work on drive 1 and output on drive 2?

**Art Vandelay** · 27 Apr 2007, 08:30 AM

@IPAULO

I never checked it the other way with the HDD's, I set it up as the DVD-RB author recomends since 90% of the time RB is in the encode phase and he felt the setup was best for it. I had a post some time ago in the RB forum that talked about a 2 HDD setup with DVDRB, and other people thought it should go C, D, C rather than C,D,D. It's certainly isn't scientific as TR points out I was just trying to see how much diff. the RAM and 2nd HDD with free DVD-RB 98.2 and the dual program running HCenc twice. I was led to believe that the 2 processors are close. Both systems are out of the box HP media centers with probably crap mother boards and not tweeked at all. So there is nothing custom or earth shattering with these systems. I still have the AMD system since the media center works well.

Blu, what would you do to fix the 2nd system since you say Not sure that 2nd system is optimised

Thanks Blu, your input is appreciated.

AV

**Chewy** · 27 Apr 2007, 08:40 AM

I learned a new fact today, the core duo and 533 lame ram do great together

high cas ram(4 or 5) and 533 will kill performance with amd's onboard memory controller

you would have to use cas 4 800 ddr2 to keep up with Blu's, which your mobo doesn't support?

**Art Vandelay** · 27 Apr 2007, 09:49 AM

@Chewy,

Field Value
CPU Properties
CPU Type DualCore AMD Athlon 64 X2 4600+
CPU Alias Windsor-512
CPU Stepping BH-F2
CPUID CPU Name AMD Athlon(tm) 64 X2 Dual Core Processor 4600+
CPUID Revision 00040FB2h

CPU Speed
CPU Clock 1002.1 MHz (original: [ TRIAL VERSION ] MHz)
CPU Multiplier 5.0x
CPU FSB 200.4 MHz (original: 200 MHz)
Memory Bus 200.4 MHz
DRAM:FSB Ratio CPU/5

CPU Cache
L1 Code Cache 64 KB per core (Parity)
L1 Data Cache [ TRIAL VERSION ]
L2 Cache 512 KB per core (On-Die, ECC, Full-Speed)

Motherboard Properties
Motherboard ID 08/02/2006-C51/MCP51-A8M2N-LA-00
Motherboard Name Asus A8M2N-LA (NodusM) (3 PCI, 1 PCI-E x16, 4 DDR2 DIMM, Audio, Video, LAN)

Chipset Properties
Motherboard Chipset nVIDIA GeForce 6150, AMD Hammer
Memory Timings 4-4-4-12 (CL-RCD-RP-RAS)
Command Rate (CR) [ TRIAL VERSION ]

SPD Memory Modules
DIMM1: Hyundai HYMP564U64BP8-C4 512 MB DDR2-533 DDR2 SDRAM (5-4-4-12 @ 266 MHz) (4-4-4-12 @ 266 MHz) (3-3-3-9 @ 200 MHz)
DIMM2: Hyundai HYMP564U64BP8-C4 [ TRIAL VERSION ]

BIOS Properties
System BIOS Date 08/02/06
Video BIOS Date 04/25/06
Award BIOS Type Phoenix - AwardBIOS v6.00PG
Award BIOS Message Copyright 2006 by Hewlett-Packard Company Rev. 3.07
DMI BIOS Version 3.07

Graphics Processor Properties
Video Adapter Asus EN7500LE
GPU Code Name G72 (PCI Express x16 10DE / 01DD, Rev A1)
GPU Clock 551 MHz (original: [ TRIAL VERSION ] MHz)
Memory Clock 405 MHz (original: 400 MHz)

DDR2 5300 667MHz cas4 I believe you said is the best this mobo can handle some time ago. Hot CPU but it will never keep up with a tricked out machine like yours or Blu's. Where soes it say in this info. from Everest that it will suport 667MHz?

**Chewy** · 27 Apr 2007, 10:02 AM

Supports PC 4200 (533 MHz) and PC2 5300 (667 MHz) DDR2 DIMMs

Invalid URL

http://h10025.www1.hp.com/ewfrf/wc/document?docname=c00757531&lc=en&cc=us&dlc=&product=3310300

post that core duo

http://www.newegg.com/Product/Produc...+667+(PC2+5300)

to catch Blu you would need cas 3

**Art Vandelay** · 27 Apr 2007, 10:09 AM

I don't have the core 2 duo any more to run a scan but I think the HP model was m7667c-b or something very similar. I know it only had SATA conections in it for the HDD and the DVD roms, not any of the older stuff.

**Chewy** · 27 Apr 2007, 10:28 AM

I found the memory benchmark latency to correlate with amd64's and encoding/transcoding performance

my ballistix running at 500 cas 2.5 screams at 45 ns

my geil

SPD Memory Modules:
DIMM1: GeIL CL2-3-3 Ultra-400 256 MB PC3200 DDR SDRAM (2.0-3-3-6 @ 200 MHz)
DIMM3: GeIL CL2-3-3 Ultra-400 256 MB PC3200 DDR SDRAM (2.0-3-3-6 @ 200 MHz)

is 85ns but is controlled by an amd xp 3200

what's your latency?

**Art Vandelay** · 27 Apr 2007, 11:42 AM

@Chewy,

how do I find latency?

Device Description
DIMM1: Hyundai HYMP564U64BP8-C4
DIMM2: [ TRIAL VERSION ]

Field Value
Memory Module Properties
Module Name Hyundai HYMP564U64BP8-C4
Serial Number 0000201Bh (455081984)
Manufacture Date Week 38 / 2006
Module Size 512 MB (1 rank, 4 banks)
Module Type Unbuffered DIMM
Memory Type DDR2 SDRAM
Memory Speed DDR2-533 (266 MHz)
Module Width 64 bit
Module Voltage SSTL 1.8
Error Detection Method None
Refresh Rate Reduced (7.8 us)

Memory Timings
@ 266 MHz 5-4-4-12 (CL-RCD-RP-RAS) / 16-28-2-4-2 (RC-RFC-RRD-WR-WTR)
@ 266 MHz 4-4-4-12 (CL-RCD-RP-RAS) / 16-28-2-4-2 (RC-RFC-RRD-WR-WTR)
@ 200 MHz 3-3-3-9 (CL-RCD-RP-RAS) / 12-21-2-3-2 (RC-RFC-RRD-WR-WTR)

Memory Module Manufacturer
Company Name Hynix Semiconductor Inc.
Product Information http://www.hynix.com/eng/02_products/01_dram/index.jsp

**blutach** · 27 Apr 2007, 11:47 AM

latency is the CL figure.

Regards

**Chewy** · 27 Apr 2007, 11:48 AM

under benchmark

there's a little refresh symbol at the top left next to report that runs the test

what's your memory latency Blu?

Encoding Times with 2 Diff. Systems

Encoding Times with 2 Diff. Systems

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment