Hello everyone,
I finally completed building, OCing and stress-testing my rig to replace my aging 5 year old one. I couldn't have done it without everyone's help, so to start off, thank you.
There were a few people who were interested in what the performance increase was like, specifically a user named 'shortass' so I did some quick benchmarking for the numbers.
Let's start off with the old rig:
Antec SLK3700AMB
Seasonic S12II 380GB (PSU that came with the case died.)
Asus A7N8X Deluxe Rev 1.04
AMD Athlon XP 2600+ (Thoroughbred) @ STOCK/Auto
2x 256MB Crucial DDR333 @ STOCK
1x 512MB Crucial DDR333 @ STOCK
ATI Radeon 9800 Pro @ STOCK (9500 Pro died unfortunately, grabbed an unused 9800 Pro from a friend.)
Creative SB Audigy MP3+
1x Raptor74
2x WD 120GB
1x WD 500GB
Labtec Pulse 424 (2.1 speakers)
Loved that rig to death, but after 5.5 years it really started to show its age.
Power Consumption:
IDLE: 155W average
LOAD: 200W average
3DMark '06 Score:
926 3DMarks
455 SM2.0
N/A HDR/SM3.0
723 CPU
Power draw is measured from the wall using the P3 Kill-A-Watt meter set to measure watts. I don't have any kind of logging mechanism though, so these figures are from my watching the meter and eyeballing an average power draw that it liked to hover around. It peaked at around 202W.
All 3DMark '06 benchmarks are run at 1280x1024 resolution because that's the highest resolution my Samsung SyncMaster 172x LCD supports. I select as many of the benchmarks as I can, and leave all the other settings at default/optimized.
I couldn't run some of the benchmarks because the 9800 Pro didn't support some of the newer technologies. It also scores abyssmally low with the '06 version because honestly, who in their right mind uses a 9800 Pro in 2006 to play the latest games? The score probably would've been higher if I used '03 or something since that's when the 9800 Pro was around, but that doesn't show a true comparison as it's like playing Half-Life vs playing Crysis.
I skipped PCmark entirely because half the tests wouldn't even run. I could only run less than a dozen tests so I decided it wasn't worth it. For consistency I also avoided PCMark on my new rig, although if you really, really want to know, you could ask and I'll think about running it.
Now, on to the meat of things..
New Rig:
Antec P182
Seasonic S12II 380GB
Gigabyte GA-EP35-DS3P Rev 2.1
Intel Core2Duo E8400 @ 3.6GHz (400x9 @ 1.25000v)
2x 2GB Corsair DDR800 @ STOCK
eVGA e-GeForce 8800GT (stock) @ 700/1750/2000 core/shader/memory
Creative SB Audigy MP3+
1x Raptor150
1x Raptor74
3x WD 500GB
Acoustic Energy Aego M (2.1 speakers)
I bought the stock 8800GT from eVGA to save a few bucks, but it ended up OC'ing to the SSC version pretty easily. At least it seems so for now.
Yes, I re-used a few things, namely the PSU because when my old one died I purchased the Seasonic with the intention of re-using it on my next rig, as well as my SB Audigy because I didn't feel like spending more money on another dedicated sound card. I was always of the opinion that a dedicated sound card would always beat onboard sound, but the Gigabyte board comes with the Realtek ALC889A "HD Audio" which was supposedly pretty good.
My friend and I did some tests to compare the two with RMAA and an external Audiofire 2 to eliminate variables, but that's for another post if people are interested. I'm still gathering the graphs together. Bottom line, the Realtek is only slightly better than the Audigy, because it doesn't suffer from the crazy intermod the Audigy has due to it resampling everything to 48KHz. Unfortunately, the Realtek suffers from a lot of low frequency noise, presumably due to it being around so many other electronics on the motherboard.
Getting back to the rig, most everyone's biggest worry was over the PSU, and for good reason. Looking at the parts list, it seems as it it will chew up power like no tomorrow. However, it seems most people over-recommend the PSU insanely just for the "headroom," and while I understand it, anything over 400w is overkill for a single video card system.
Power Consumption:
IDLE: 122W average
LOAD: 200W average
As you can see, even with these beefy components -and overclocked, at that- it still draws roughly the same amount of power at full load. For both power consumption figures of both rigs, it was run while idle in Windows Desktop for a while for IDLE, then while it was running 3DMark '06 for LOAD. The new rig peaked at 209W. Idle consumption was significantly lower, a good 33W savings is huge as most of the time the PC is on it is spent idling, and using less power while overclocked is even better. Full CPU load with this rig was about 170W, running Prime95/Orthos.
Now then ... wait for it....
3DMark '06 Score:
13128 3dmarks
6036 SM2.0
5629 HDR/SM3.0
3355 CPU
LOL @ these scores. 926 to 13128 is a massive difference, and you could definitely tell the FPS difference when the benchmarks were running. It still dipped to around 30fps for one of the benchmarks though. Oddly, the CPU benchmark was struggling at 1fps. Again, this was at 1280x1024 as I am re-using my Samsung SyncMaster 172x. I don't have another $500-600 right now to upgrade to another decent LCD.
This is definitely a huge improvement, but I have yet to try it out in real world games. I'll report back later with some findings in Team Fortress 2. The old rig barely managed 10-30fps and was incredibly random at 1024x768.
The speakers have nothing to do with the system performance, but my friend nagged me to buy some new speakers. I really didn't think anything could beat the Labtecs as they were fairly good sound for $50. The Aego M's on the other hand were $200, and for that price it had better be freaking amazing, since I didn't want to spend that much.
They arrived, and I had to eat my words, much to the dismay of my starving wallet. The Labtecs I believed were great for the 5 or so years I had them, but its problems was that it had fairly high bass. I turned the sub all the way down to minimum just so I wouldn't wake the neighbors. They also had an exaggerated midrange. The Aegos on the other hand, appeared to have a "hole" in the upper bass (around 100Hz) and the subwoofer's bass was not nearly as strong. This isn't necessarily a con for me as I'm not one of those people who think more bass is better. Accurate sound is better than more bass IMO.
I hook them both up and switch between them with an audio switch I purchased to compare them side by side with several songs. I am in no way anything close to an audiophile, so while my friend (the audiophile) was spouting all this technical babble about why the Aego's were superior, I could only say "Man the Labtecs feel like someone is covering my ears, and the Aego's are like them taking their hands away." To say the least I'm quite happy with the Aego's.
Anyway, sorry for the long post, but I hope it helps someone out there.
I'll edit this later with more information, specifically the heatsinks/fans and of course the idle power consumption of my new rig.
Edit1:
Got the power consumption figures in after C1E/EIST have been enabled. Now idling at 400x6 for 2.4GHz. I think I saved about 5-9W simply lowering the multiplier with EIST. I doubt C1E can do anything since I have Vcore manually set at 1.25000 but who knows.
Will add Heatsink/fan information tomorrow as I'm too tired now and I need to look up past invoices to figure out what I put in the old rig. I know it was a bunch of Panaflo fans though.
Edit2:
Okay I got the rest of the info I needed. I'm going to put it down here so it's easier to see. I don't know if it's better to merge the contents to the rest of the post or not.
This stuff is in regards to cooling and niose.
Old Rig:
CPU Heatsink - Thermalright SK-7
CPU Fan - Panaflo 80mm M1A Medium Output @ 12v
Case Fans - 2x Panaflo 120mm L1A Low Output @ 12v
VGA Heatsink - VGA Silencer on my old 9500 Pro, 9800 Pro I ended with had a Zalman of some kind with a 120mm fan strapped on to its side @ 12v.
Back then I didn't know too much about the cooling business, but the Panaflo's were very quiet back then. The SK-7 was "huge" because it was over the recommended weight AMD gave for Socket A cooling. I was actually a little afraid of mounting it because it might fall off if I moved the case too much. But of course, I ended up getting an even bigger one for my new rig...
The CPU temp when I first built it was around 32C Idle if I recall correctly from BIOS that first time we powered it on, and we were like MAN that's cold. The 120mm Panaflo's were also much thicker than normal (32mm IIRC vs current 25mm) and it just barely fit into the case. There was 1mm of clearance or less between the intake and the HD cage. Side effect was that it pushed so much air, everything was kept very cool. Downside was it also pushed a great deal of dust which eventually degraded its cooling ability somewhat. At the end of 5 years when I was ordering parts for my new rig, it was idling at 50C and going to 65C load as reported by Motherboard Monitor.
Sound-wise, I didn't have any professional noise testing equipment but even with quiet fans, they pushed so much air, most of the time there was a loud whooshing sound. Get behind that and you could hear the fans whirring but that took a bit of effort, or a quiet room.
Now for the comparison..
New Rig:
CPU Heatsink - Scythe Ninja Plus Rev B
CPU Fan - Scythe Slipstream SY1225SL12M 1200RPM
Case Fans -
- Scythe S-FLEX SFF21E 1200RPM - Exhaust
- Nexus 120mm 1000RPM - Lower HD Rack
VGA Heatsink - Accelero S1 Rev 2 w/ Turbo Module @ 5v
I thought the SK-7 was over-spec. My jaws dropped when I actually held the Ninja and felt the sheer size of it. (Now, imagine the Orochi.. *shudder*) I decided to use the stock fan since the Scythe Slipstream was a fairly good fan as well as the other two. I was going to run the Slipstream and the Nexus at 5v while leaving the S-FLEX at 12v but I haven't gotten around to messing with all those cables yet so for the time being, they're just plugged into the motherboard and it's being controlled via that. The Turbo Module is 5v since I modded that cable when installing it.
There may be an issue with my heatsink not being mounted securely enough, or that the E8400 simply isn't reporting correct temps, or a combination of both, but my temps are quite a bit higher than I'd like. I'm going to take some time next week to redo the system a little bit and bolt it down some more. As of currently, the CPU idles at 45C and loads not more than 60C. I'm using Ceramique because I had a lot of it on hand.
The Accelero is simply a beast. Case in point, my overclocked load temps are the same as my stock idle temps. With the stock card, it was measuring 58C idle and 79C load. After I mounted the Accelero, idle was 39C with load of 58C. This was measured via ATITool while it was running the fuzzy cube to heat up the card and scan for artifacts with the overclock.
As far as sound goes, it was immediately apparent that this system was obviously quieter than my previous rig. The fans themselves are quieter and the P182 does an excellent job of dampening some extra sound. Again, I do not have any specialized equipment but my room is very quiet and I can't hear the fans spinning most of the time (except the dead of night.) Most of the sound is again, air whooshing through the case. Some fans may be running faster than I'd like though, as I haven't gotten around to modding them to run at 5v - they're controlled via the motherboard's temp sensors. I do have a couple fanmates that I could attach though, and again I'll find some time later in the week to do so. For now, I will say that the PC is a lot quieter than before. Under normal circumstances it's almost inaudible. If i could slow the fans down a bit I'm pretty sure it would be inaudible.
One final note, I managed to play some Team Fortress 2 today and I have to say the difference was absolutely stunning. Old rig was running at DirectX 8.1 at 1024x768 with minimal visual effects and struggling to get a stable 15fps most of the time.
I just ran TF2 at 1280x1024 @ DirectX 9.1 with full visual effects and 4x MSAA and 2x AF and still got 110-140fps.
I'm sorry this post has become so absolutely huge, but I hope it helps sone people looking to upgrade to see what kidn of performance gains they will see upgrading from 5 year old components. If there's anything else you want me to talk about, feel free to ask.
Thanks again!
I finally completed building, OCing and stress-testing my rig to replace my aging 5 year old one. I couldn't have done it without everyone's help, so to start off, thank you.
There were a few people who were interested in what the performance increase was like, specifically a user named 'shortass' so I did some quick benchmarking for the numbers.
Let's start off with the old rig:
Antec SLK3700AMB
Seasonic S12II 380GB (PSU that came with the case died.)
Asus A7N8X Deluxe Rev 1.04
AMD Athlon XP 2600+ (Thoroughbred) @ STOCK/Auto
2x 256MB Crucial DDR333 @ STOCK
1x 512MB Crucial DDR333 @ STOCK
ATI Radeon 9800 Pro @ STOCK (9500 Pro died unfortunately, grabbed an unused 9800 Pro from a friend.)
Creative SB Audigy MP3+
1x Raptor74
2x WD 120GB
1x WD 500GB
Labtec Pulse 424 (2.1 speakers)
Loved that rig to death, but after 5.5 years it really started to show its age.
Power Consumption:
IDLE: 155W average
LOAD: 200W average
3DMark '06 Score:
926 3DMarks
455 SM2.0
N/A HDR/SM3.0
723 CPU
Power draw is measured from the wall using the P3 Kill-A-Watt meter set to measure watts. I don't have any kind of logging mechanism though, so these figures are from my watching the meter and eyeballing an average power draw that it liked to hover around. It peaked at around 202W.
All 3DMark '06 benchmarks are run at 1280x1024 resolution because that's the highest resolution my Samsung SyncMaster 172x LCD supports. I select as many of the benchmarks as I can, and leave all the other settings at default/optimized.
I couldn't run some of the benchmarks because the 9800 Pro didn't support some of the newer technologies. It also scores abyssmally low with the '06 version because honestly, who in their right mind uses a 9800 Pro in 2006 to play the latest games? The score probably would've been higher if I used '03 or something since that's when the 9800 Pro was around, but that doesn't show a true comparison as it's like playing Half-Life vs playing Crysis.
I skipped PCmark entirely because half the tests wouldn't even run. I could only run less than a dozen tests so I decided it wasn't worth it. For consistency I also avoided PCMark on my new rig, although if you really, really want to know, you could ask and I'll think about running it.
Now, on to the meat of things..
New Rig:
Antec P182
Seasonic S12II 380GB
Gigabyte GA-EP35-DS3P Rev 2.1
Intel Core2Duo E8400 @ 3.6GHz (400x9 @ 1.25000v)
2x 2GB Corsair DDR800 @ STOCK
eVGA e-GeForce 8800GT (stock) @ 700/1750/2000 core/shader/memory
Creative SB Audigy MP3+
1x Raptor150
1x Raptor74
3x WD 500GB
Acoustic Energy Aego M (2.1 speakers)
I bought the stock 8800GT from eVGA to save a few bucks, but it ended up OC'ing to the SSC version pretty easily. At least it seems so for now.
Yes, I re-used a few things, namely the PSU because when my old one died I purchased the Seasonic with the intention of re-using it on my next rig, as well as my SB Audigy because I didn't feel like spending more money on another dedicated sound card. I was always of the opinion that a dedicated sound card would always beat onboard sound, but the Gigabyte board comes with the Realtek ALC889A "HD Audio" which was supposedly pretty good.
My friend and I did some tests to compare the two with RMAA and an external Audiofire 2 to eliminate variables, but that's for another post if people are interested. I'm still gathering the graphs together. Bottom line, the Realtek is only slightly better than the Audigy, because it doesn't suffer from the crazy intermod the Audigy has due to it resampling everything to 48KHz. Unfortunately, the Realtek suffers from a lot of low frequency noise, presumably due to it being around so many other electronics on the motherboard.
Getting back to the rig, most everyone's biggest worry was over the PSU, and for good reason. Looking at the parts list, it seems as it it will chew up power like no tomorrow. However, it seems most people over-recommend the PSU insanely just for the "headroom," and while I understand it, anything over 400w is overkill for a single video card system.
Power Consumption:
IDLE: 122W average
LOAD: 200W average
As you can see, even with these beefy components -and overclocked, at that- it still draws roughly the same amount of power at full load. For both power consumption figures of both rigs, it was run while idle in Windows Desktop for a while for IDLE, then while it was running 3DMark '06 for LOAD. The new rig peaked at 209W. Idle consumption was significantly lower, a good 33W savings is huge as most of the time the PC is on it is spent idling, and using less power while overclocked is even better. Full CPU load with this rig was about 170W, running Prime95/Orthos.
Now then ... wait for it....
3DMark '06 Score:
13128 3dmarks
6036 SM2.0
5629 HDR/SM3.0
3355 CPU
LOL @ these scores. 926 to 13128 is a massive difference, and you could definitely tell the FPS difference when the benchmarks were running. It still dipped to around 30fps for one of the benchmarks though. Oddly, the CPU benchmark was struggling at 1fps. Again, this was at 1280x1024 as I am re-using my Samsung SyncMaster 172x. I don't have another $500-600 right now to upgrade to another decent LCD.
This is definitely a huge improvement, but I have yet to try it out in real world games. I'll report back later with some findings in Team Fortress 2. The old rig barely managed 10-30fps and was incredibly random at 1024x768.
The speakers have nothing to do with the system performance, but my friend nagged me to buy some new speakers. I really didn't think anything could beat the Labtecs as they were fairly good sound for $50. The Aego M's on the other hand were $200, and for that price it had better be freaking amazing, since I didn't want to spend that much.
They arrived, and I had to eat my words, much to the dismay of my starving wallet. The Labtecs I believed were great for the 5 or so years I had them, but its problems was that it had fairly high bass. I turned the sub all the way down to minimum just so I wouldn't wake the neighbors. They also had an exaggerated midrange. The Aegos on the other hand, appeared to have a "hole" in the upper bass (around 100Hz) and the subwoofer's bass was not nearly as strong. This isn't necessarily a con for me as I'm not one of those people who think more bass is better. Accurate sound is better than more bass IMO.
I hook them both up and switch between them with an audio switch I purchased to compare them side by side with several songs. I am in no way anything close to an audiophile, so while my friend (the audiophile) was spouting all this technical babble about why the Aego's were superior, I could only say "Man the Labtecs feel like someone is covering my ears, and the Aego's are like them taking their hands away." To say the least I'm quite happy with the Aego's.
Anyway, sorry for the long post, but I hope it helps someone out there.
I'll edit this later with more information, specifically the heatsinks/fans and of course the idle power consumption of my new rig.
Edit1:
Got the power consumption figures in after C1E/EIST have been enabled. Now idling at 400x6 for 2.4GHz. I think I saved about 5-9W simply lowering the multiplier with EIST. I doubt C1E can do anything since I have Vcore manually set at 1.25000 but who knows.
Will add Heatsink/fan information tomorrow as I'm too tired now and I need to look up past invoices to figure out what I put in the old rig. I know it was a bunch of Panaflo fans though.
Edit2:
Okay I got the rest of the info I needed. I'm going to put it down here so it's easier to see. I don't know if it's better to merge the contents to the rest of the post or not.
This stuff is in regards to cooling and niose.
Old Rig:
CPU Heatsink - Thermalright SK-7
CPU Fan - Panaflo 80mm M1A Medium Output @ 12v
Case Fans - 2x Panaflo 120mm L1A Low Output @ 12v
VGA Heatsink - VGA Silencer on my old 9500 Pro, 9800 Pro I ended with had a Zalman of some kind with a 120mm fan strapped on to its side @ 12v.
Back then I didn't know too much about the cooling business, but the Panaflo's were very quiet back then. The SK-7 was "huge" because it was over the recommended weight AMD gave for Socket A cooling. I was actually a little afraid of mounting it because it might fall off if I moved the case too much. But of course, I ended up getting an even bigger one for my new rig...
The CPU temp when I first built it was around 32C Idle if I recall correctly from BIOS that first time we powered it on, and we were like MAN that's cold. The 120mm Panaflo's were also much thicker than normal (32mm IIRC vs current 25mm) and it just barely fit into the case. There was 1mm of clearance or less between the intake and the HD cage. Side effect was that it pushed so much air, everything was kept very cool. Downside was it also pushed a great deal of dust which eventually degraded its cooling ability somewhat. At the end of 5 years when I was ordering parts for my new rig, it was idling at 50C and going to 65C load as reported by Motherboard Monitor.
Sound-wise, I didn't have any professional noise testing equipment but even with quiet fans, they pushed so much air, most of the time there was a loud whooshing sound. Get behind that and you could hear the fans whirring but that took a bit of effort, or a quiet room.
Now for the comparison..
New Rig:
CPU Heatsink - Scythe Ninja Plus Rev B
CPU Fan - Scythe Slipstream SY1225SL12M 1200RPM
Case Fans -
- Scythe S-FLEX SFF21E 1200RPM - Exhaust
- Nexus 120mm 1000RPM - Lower HD Rack
VGA Heatsink - Accelero S1 Rev 2 w/ Turbo Module @ 5v
I thought the SK-7 was over-spec. My jaws dropped when I actually held the Ninja and felt the sheer size of it. (Now, imagine the Orochi.. *shudder*) I decided to use the stock fan since the Scythe Slipstream was a fairly good fan as well as the other two. I was going to run the Slipstream and the Nexus at 5v while leaving the S-FLEX at 12v but I haven't gotten around to messing with all those cables yet so for the time being, they're just plugged into the motherboard and it's being controlled via that. The Turbo Module is 5v since I modded that cable when installing it.
There may be an issue with my heatsink not being mounted securely enough, or that the E8400 simply isn't reporting correct temps, or a combination of both, but my temps are quite a bit higher than I'd like. I'm going to take some time next week to redo the system a little bit and bolt it down some more. As of currently, the CPU idles at 45C and loads not more than 60C. I'm using Ceramique because I had a lot of it on hand.
The Accelero is simply a beast. Case in point, my overclocked load temps are the same as my stock idle temps. With the stock card, it was measuring 58C idle and 79C load. After I mounted the Accelero, idle was 39C with load of 58C. This was measured via ATITool while it was running the fuzzy cube to heat up the card and scan for artifacts with the overclock.
As far as sound goes, it was immediately apparent that this system was obviously quieter than my previous rig. The fans themselves are quieter and the P182 does an excellent job of dampening some extra sound. Again, I do not have any specialized equipment but my room is very quiet and I can't hear the fans spinning most of the time (except the dead of night.) Most of the sound is again, air whooshing through the case. Some fans may be running faster than I'd like though, as I haven't gotten around to modding them to run at 5v - they're controlled via the motherboard's temp sensors. I do have a couple fanmates that I could attach though, and again I'll find some time later in the week to do so. For now, I will say that the PC is a lot quieter than before. Under normal circumstances it's almost inaudible. If i could slow the fans down a bit I'm pretty sure it would be inaudible.
One final note, I managed to play some Team Fortress 2 today and I have to say the difference was absolutely stunning. Old rig was running at DirectX 8.1 at 1024x768 with minimal visual effects and struggling to get a stable 15fps most of the time.
I just ran TF2 at 1280x1024 @ DirectX 9.1 with full visual effects and 4x MSAA and 2x AF and still got 110-140fps.
I'm sorry this post has become so absolutely huge, but I hope it helps sone people looking to upgrade to see what kidn of performance gains they will see upgrading from 5 year old components. If there's anything else you want me to talk about, feel free to ask.
Thanks again!