Effluo (20th August 2011)
what you can do with a good quality efficient 400 watt PSU....
Code:Processor AMD Phenom II X6 1090T Black Edition (3.2GHz / 125w TDP) Motherboard Asus M4A89GTD PRO/USB3 RAM 8GB (2x4GB) DDR3 1,333MHz Low Voltage DIMMs SSDs 2 x Crucial RealSSD C300 256GB SATA III SSDs HDDs 1 x Samsung SpinPoint F3 (HD103SJ) 1TB HDD 2 x Western Digital Caviar Black (WD1002FAEX) 1TB HDDs RAID Card LSI MegaRAID SAS 9261-8i 6Gb/s PCIe Card w/ 512MB memory ODD Asus DRW-24B1ST/BLK/B/AS PSU Corsair CMPSU-400CX Fans 1 x Scythe 120mm / 1 x Scythe 100mm
As you can see from the table above, the maximum power this system will draw is 290 watts. Most people don't run Prime95 and Furmark so power consumption will typically be significantly lower.
Effluo (20th August 2011)
no mention of a dedicated graphics card
Post you are linking to is using onboard graphics which are generally rubbish and also use your RAM so OP would lose RAM if they were to use onboard options
If you were buying dedicated graphics cards they wouldn't be idleing away all the time so am guessing it would draw power / wattage the second you pressed the power button
Power and thermal section above
minimum one is 128 watts + 290 watts = 418
If we use the 237 from the chart ( prime95 )
237 + 128 = 365 - again still quite high on a 400 watt PSU and not much room for upgrade or headway to do much
Maybe a 400 watt PSU for a home brewed NAS where graphics card you could use onboard, gigabit ethernet, 8gb of ram and whatever raid card and sata hard drives you wanted ??
The 128 watts is the idleing wattage with Aero
Last edited by mac_shinobi; 20th August 2011 at 12:00 AM.
I thought he unsubscribed from the thread, was just replying to arthurs post but then why not go with a 350 watt psu :P
page 5) which you are adding to the power consumption of the PC above running Prime95.
From Guru3D's review of the Radeon HD 6870...
Using these figures for the Phenom II X6 system above we get an idle of 125 watts (106 + 19) and a load of 368 watts (237 + 131). As per the review, a 500 watt PSU is recommended for a 6870.We'll be calculating the GPU power consumption here, not the total PC power consumption.
Measured power consumption
System in IDLE = 166w
System wattage with GPU in FULL Stress = 278w
Difference (GPU load) = 112w
Add average IDLE wattage ~19w
Subjective obtained GPU power consumption = ~131 watts
A 500, not 400 but a 500 watt psu
Thanking you :-)
Ignore my math above :P
Ellyou or w.e can carry on using 400 watt psu's. Dont make any diff to me
Some people just don't want to be saved xD
Your entire argument in this tread has been faulty anyway as the OP is not an 'average' user and has stated that they will be using it for virtualisation. This forum has many people with more than a clue about hardware and software so the conensus here is worth considering. At any rate I think that we have established that this whole line of debate is pointless in this particular thread given the OPs requirements.
mac_shinobi (20th August 2011)
The thing is that I've never once said that you should choose a PSU which will give you no overhead. I've in fact said that you do need a decent amount of overhead throughout this thread. So please can people stop putting words into my mouth, it's happened around 3 or 4 times now... Reread my posts, I'm not suggesting anything mad here.
Also @@synack claiming that my whole line of argument is faulty... wut?
Please reread before trying to justify that statement, we've been talking about a wide range of things on this thread and if I ever mentioned the regular user then it would have just been in reference to what someone else had said. The only non-average thing about the op in terms of need for PC hardware is that she would apparently benefit from having more than 4GB of RAM. Most of this last bit of conversation has been about PSU's and anything I've said on that has been very valid and relevant.
Also, I do appreciate that people do have their own experiences and I'm completely open to that, because I love learning more about these things and I yearn for any constructive input. BUT some posters here are just trying to defend what themselves in their own set ways whether right or wrong. Repeatedly showing that they have put no effort into actually researching the thing and then augmenting any benchmarks they have used just to try and justify their own decision. Laughable, then some other guy comes in showing another how he misread the benchmarks and he still tries to twist it to agree with himself, while completely ignoring what I've said throughout the thread.
Not to mention trying to make out that I've said things when I haven't.
A lot of IT people do this, they talk about things they know very little about while thinking they know a lot, because they know more than the common man and they can usually get away with bluffing 90% of the time.
After the kind of attitude that's been displayed in this thread by some people, then the unfortunate conclusion that I come to will be that this guy no matter how clear the evidence or logic suggests will never back down from their position for the sake of not backing down, some kind of man pride. :/
I'm really not interested though
As I've said already, I don't have time to readdress what I've already said...
Have a nice Saturday o/
arguement is moot as per synack and agree with synacks comment - you stick to how you do it and I will do it my way - agree to disagree ( think that is how the saying goes )
Just leave it at that
Going to quote a fellow EG member here in saying the following :
1. 400 Watt PSU's don't have as many power lead outputs
2. They tend not to be modular
3. Reminds me of the tyre thread on here a while back ref using used tyres vs new when its the only thing between you and the tar mac , same here ref its the only thing between you and a fried computer and the only thing between having over head and later on down the line being able to upgrade to that new graphics card that ATI / Nvidia have released that requires an extra power lead, oh hold on that 400 watt power supply can't give you enough power and hold on what lead can I use ? Oh yeah gotta buy another power supply , when instead of buying a reasonable power supply to start with you now have to throw more money at another power supply and that's without buying that new graphics card or new whatever.
I am just saying ( as everyone else on here knows ) if you spend your money wisely to start with ref getting a power supply that does give you the extra over head ( yes you don't require it straight away ) but at least it is there and available and you can forget about it.
At the end of the day I have not built my own PC for a few years now and am sure the OP is extremely capable of deciding which path / route he wants to go down as far as PSU , wattage requirements and the rest
@ Mod - can this debate be deleted from this thread as pointless having it as part of the OP's thread as am sure the OP wanted answers to his original question not a debate because of a difference in opinion
Last edited by mac_shinobi; 20th August 2011 at 12:28 PM.
ArsTechnica has a excellent article on choosing the correct power supply. As they say, one size does not fit all.
Power supplies run from diminutive things running tiny low-profile systems the size of a few paperback books to massive kilowatt-plus units powering extreme multi-GPU setups. Clearly, one size does not fit all. Aside from just total output capacity (size) to consider, form factor, efficiency, and noise are all important factors, not to mention cost.
Power output: Size the power output of your PSU appropriately to your system's requirements. Plan for expansion if needed, but be realistic. Keeping things appropriately sized will keep you in the sweet spot of the power supply's efficiency curve, which is a good thing, even if curves today are fairly flat.
Consider your system both at idle, typical load, and maximum load, and size accordingly. Check individual component draw on specific power supply rails (+3.3v, +5v, and however many +12v rails are involved) and make sure the PSU can deliver power at the correct rails. This has shifted in the past from +3.3v and +5v focused to +12v focused, so those old power supplies, despite their ample total power output, may actually not work on a newer system.
Efficiency: Converting typical household power—delivered as alternating current to the direct current consumed by your computer's internal components—is imperfect. Old power supplies tend to be less efficient than newer ones, although finding lousy new power supplies based on older standards is still easy. A new power supply with a modern design will be more efficient, which means less electricity wasted as heat, and a lower electricity bill. The old first-generation ATX PSU may be 65 percent efficient compared to a modern 80PLUS certified unit that is 88 percent or more efficient. Good quality power supplies also have very efficient active power factor correction (active PFC) rather than older passive PFC or non-PFC designs, which helps in some areas (although at least in the USA, most users don't have to worry about PFC affecting their electricity bills).
Connectors and cable length: Cheap power supplies tend to come with fewer connectors and shorter cables. Make sure your power supply has the right connectors for your components and that the cables are long enough. If your case mounts the power supply in a nonstandard location, double check the cable length. You may need especially long cables or especially short cables in such cases. Those of you who have tried to cram 20 inches of cable into a case the size of a shoebox know our pain, as do those who picked up the high-end components only to find the CPU power cable was an inch too short (!). If your power supply of choice is short on connectors, splitters and adapters may be used, but be careful not to overload the power supply—pay attention to the maximum current available on the required power supply rails and the power draw of the components being installed.
A final side note on efficiency, size, and manufacturer's recommendations: actual power draw on most end-users' systems is much lower than manufacturer-specified minimum recommendations tend to be. This is because most end-users don't run their systems at 100 percent load all of the time, and manufacturers tend to be conservative. After all, the manufacturer of your CPU, motherboard, or video card has no idea if you are using a quality name-brand power supply that can actually deliver the power it's rated to give, or if you are using a generic el cheapo unit that will explode when taxed beyond a small fraction of its load. They tend to try to accommodate the el cheapos out of the simple necessity that they really do not know what quality power supply the end-user may have.
Back on topic.
poor reliability, then there is the performance degradation problem which affects all SandForce SF1200-based SSDs (the Vertex 2E is one of them) and thirdly, if you bought an SSD from Intel or OWC you would get a five year warranty instead of three (link, link).Lastly, the AMD offer is showing its age. The on-chip memory controller on the Phenom IIs struggles to benefit fully from DDR3-1333 and 1600 memory, as bandwidth usage hits a ceiling far too quickly. Performance in the mono-threaded test also shows the limits of an architecture whose successor, Bulldozer, is eagerly awaited and should become available in the second quarter. (Source)
I was about to buy a new machine this month (but that got delayed due to new job), and comparing the AMD to Intel offerings, the Intel trounce AMD in pretty much every way at the moment. The Bulldozer chips are supposed to even that out though, so if you're really going to go for AMD, wait until they're released and see how they compare in reviews before buying, else go for an i5-2500 (or the 2500K, as it is on sale at the moment at http://www.aria.co.uk/SuperSpecials/Other+products/Intel+Core+i5-2500K+3.30GHz+%28Sandybridge%29+Socket+LGA1155+Pro cessor+-+Retail+?productId=43216]aria[/url], as it is is a better chip, and can handle faster memory.
There are currently 1 users browsing this thread. (0 members and 1 guests)