IGP Power Consumption - 780G, GF8200, and G35

by Gary Key on 4/18/2008 2:00 AM EST


Back to Article

  • wjl - Sunday, April 27, 2008 - link

    If you own a decent mainboard and processor (meaning anything newer than AMD K7 or Intel P4), I think it's mostly the power supply which can help with efficiency.

    Building my AMD X2 3800+ EE into a bigger case with an 80+ certified power supply (and 90+ is coming soon as well, according to DigiTimes) helped reducing the idle power cosumption from 89W (with one hard disk and TV card) to some 74W (now with two hard disks and TV cards).

    New main boards and single platter hard disks may help, but it's normally better to look at the weak points of your current systems IMHO.
  • hian001 - Wednesday, May 22, 2013 - link

    Penetration Testing A standard penetration test is a way for scientists to get an idea of how resistant the soil in a certain area is to the invasion of water. The tests conducted on the sample are usually unaffected by the disturbance of the soil. <a href="http://www.breaksec.com/">Penetration Testing</a> Reply
  • computerfarmer - Wednesday, April 23, 2008 - link

    It is good to see this kind of information. Lower power consumption has its place considering the cost, those that pay the bills will understand. I am looking forward to the 780G roundup, if it includes the 8200, that's good too.
    Keep up the excellent work.
  • bingbong - Tuesday, April 22, 2008 - link

    Ok Thanks to those who said it is available. I will try the vendors again. I have only been looking online because I tried to buy it a couple of times. Actually I am in Taiwan so the availability is usually pretty good.

  • sheh - Monday, April 21, 2008 - link

    This recent power efficiency trend is something I definitely like. At this rate, and with Atom coming, it might be possible to run a decent general-purpose computer on a <200W PSU.

    *cleans up the AT PSU*

  • smilingcrow - Sunday, April 20, 2008 - link

    I’ve tested a couple of systems with different Gigabyte G35 boards with various 65nm C2Ds and idle power consumption averaged around 55W with a spec similar to yours; only 2x1GB of RAM though. SPCR tested the exact same Asus board and managed no worse than 56W at idle although that was with a lesser spec and older less efficient E6400. This makes me wonder how you managed to record 84W at idle with an E2200 when I managed 52.5W with an E2140! Reply
  • wjl - Sunday, April 20, 2008 - link

    Yeah, exactly. My wife's machine (G35, E8200, 2Gig RAM, and a Samsung F1 750) draws some 69W on idle when in Gnome / Debian Lenny - and that is reported to take more than anything M$. The power supply used here is an EarthWatts (Seasonic) 500W 80+, which seems to be ok even at those low levels.

    My own one for comparison: AMD X2 3800+ EE on Nvidia 6150/430, 3Gig, 2x250Gig, 380W EarthWatts 80+ - takes also 69 Watts in its best moments (Gonme / Debian Etch, which doesn't have something like cpufreqd yet), tho here the "average" idle is more like 75W.

  • smilingcrow - Monday, April 21, 2008 - link

    I emailed the author about the G35 power data as it seemed disproportionately high and was skewing the results; he didn’t reply but the G35/E2200 at idle data has now been reduced from 84 to 74W with no reference being made to the change! Reply
  • smilingcrow - Monday, April 21, 2008 - link

    In a previous article by the same author http://www.anandtech.com/cpuchipsets/showdoc.aspx?...">http://www.anandtech.com/cpuchipsets/showdoc.aspx?... he looked at the same Asus G35 & Gigabyte 780G boards but compared them against an Asus GF8200 board and made the following comment:

    “As far as power consumption goes during H.264 playback, the AMD platform averaged 106W, NVIDIA platform at 102W, and the Intel platform averaged 104W - too close to really declare a true winner.”

    The same CPUs were used so some consistency might be expected between that review and the current one, but if the power data for H.264 decoding is compared we get this:

    Old / New / Difference (Watts)

    GF8200 – 102 / 77 / 25
    780G – 106W / 84 / 22
    G35 – 104W / 103 / 1

    Somehow the G35 platform seemed to gain 21 to 24W of power consumption compared to the two AM2+ platforms. Is this purely down to the higher bit-rate of the movie tested in the current review showing up the inefficiency of the G35 platform or is this another anomaly in the G35 power data!
  • deruberhanyok - Saturday, April 19, 2008 - link

    Interesting article but I'd like to echo some thoughts already posted. Seeing the power numbers is great but without any performance context to them (or comparisons being kept to same brand / models as close as possible) it's hard to see exactly where this all fits in.

    Also, low power boxes would be perfectly content running off of something like a Seasonic S12II-330. It could make a difference in overall power usage as well. Silent PC review recently discussed this in their Sparkle 250W 80 plus PSU review.

    As others are, I'm looking forward to seeing the full roundup. Thanks for all the hard work!
  • Darth Farter - Saturday, April 19, 2008 - link

    awesome, over here where it's US$ 30cents/kWH you can understand that it will start to make a difference. Only thing I would like to see is Undervolting tho that like overclocking depends on the mileage. I'm running a G1 brisbane at 2Ghz with 0.975vcore on a 690g for 24/7 download/internet box. I wonder what it costs me/month Reply
  • JarredWalton - Saturday, April 19, 2008 - link

    Given our earlier calculations of $0.10/kWh, tripling the cost of energy means you're looking at savings of up to $30 per year for 24/7 use and a difference of 10W. If you're running a 100W PC 24/7 for a whole year, that PC would cost $262.80 at $0.30/kWh or $87.60 at $0.10/kWh. Reply
  • royalcrown - Saturday, April 19, 2008 - link

    What is going on with the fried mosfets also, we never did get that weekend update ;) ?
  • royalcrown - Saturday, April 19, 2008 - link

    Why don't you have ANAND buy you guys some meters and on EVERY GFX card or PROCESSOR review list the actual wattage used by the systems. This NEEDING of at LEAST a 550 watt ps is BS for those of us that will never use dual cards.

    I just calculated that my new system on FULL load should draw about 280 watts with an 8800gt, so a 400 watt supply with 450 peak is fine for me . I read than Nvidia claims 125 watts on their page and the real draw is a lot less when they use the meter.

    I for one am sick of these companies pushing monster PSU when they AREN'T needed in every case, and sites like Anandtech should give us the scoop instead of plastering ads for 1200 watt psu and not telling readers that we may not need even 550.
  • Zaranthos - Saturday, April 19, 2008 - link

    That's a fact. I'm so sick of seeing insanely large power supplies shoved down peoples throats. I keep upgrading my computer and my 300W power supply keeps running my computer just fine. You'd think that wasn't even possible by most of the reviews/ads/propaganda. I'd like to see tests showing what the minimum power supply requirements are. Reply
  • JarredWalton - Saturday, April 19, 2008 - link

    You mean like our PSU reviews where we repeatedly state that the only way you can even come near the point where a 1000W PSU is required is if you're heavily (i.e. water- or phase-cooling) overclocking your quad-core CPU and running 3-way or 4-way GPUs?

    Most PSUs are at maximum efficiency around the 50% load mark, but even at 30% load the good PSUs are above 83% efficiency. Couple that to the fact that a 600W PSU is generally quieter delivering 150W than a 300W PSU delivering the same wattage, and there are reasons to buy higher-spec PSUs. The biggest reason to buy a higher spec PSU, of course, is that it's very difficult to find good quality PSUs rated under 400W. (Seasonic and the Seasonic-built PSUs are about the only option.)

    All that is totally overlooking the fact that *testing* with a highly-rated 520W PSU is not the same as saying the PSU is required. What's important is consistency, and here we are using the same PSU for all tests. It should have an 80-85% efficiency across the tested power requirements, which is well within the margin of error. If we drop to a 300W Seasonic, power draw might change slightly, but proportionately the results should be nearly identical to what we see in this article.

    Perhaps Gary can chime in here with some comments; I know that he sent me an initial configuration table for this article on Thursday and then changed the PSU and case later that night. The original PSU was a Seasonic unit, so perhaps he ran into some difficulties. Again, not that it really makes a difference.
  • Wirmish - Saturday, April 19, 2008 - link

    Flight Simulator X Test:
    nVidia vs AMD -> 0W to 3W, or ~2%.
    Ok... nVidia win by 2%.

    And "watt" about the FPS during these benchs ?
    Did nVidia 8200 have -2% FPS vs AMD 780G ?

    And if the 780G is faster, can you underclock it, or overclock the 8200 ?
    Try it... just to compare the consumption at the same performance level.
  • Esben - Saturday, April 19, 2008 - link

    Thanks for shedding light on the current IGP situation. It's great to see Nvidia is still competitive in the IGP-business, consumption wise. Now we eagerly await the performance numbers.

    Please keep writing about IGPs and power consumption. I'd find it very interesting if you made an articles about maximizing performance per watt, and how far in performance you can push the IGP.

    An IGP-system is fitting most peoples needs, so the interest is definitely there.
  • jacito - Friday, April 18, 2008 - link

    The artical is very well written, and this is going to sound rather stupid, but what does IGP stand for? Reply
  • JarredWalton - Friday, April 18, 2008 - link

    IGP = Integrated Graphics Processor Reply
  • Johnniewalker - Friday, April 18, 2008 - link

    I would also like to see what would happen if you swapped out the PSU with a lower power one such as the Seasonic S12 II 380w or 430w.

    If we are maxing out at 115w on the Intel G35 Quad Core, it would seem that either of those PSU's would provide more than adequate power. Would they draw less than a larger PSU or not. If they do, is the difference worth limiting yourself to a small PSU? I am not sure how limited I mean as the 430 PSU is both crossfire and sli certified.

  • kb9fcc - Friday, April 18, 2008 - link

    I found it so surprising that the single core CPU LE-1600 should be beat out by it's multi-core sibling the 4850E, that I had to look for an explanation.

    The first thing I noticed, is I could find no data on the 4850E, it didn't even appear at AMD's site. However, I found data for the 4050E and 4450E at Newegg, so I'll just have to assume that the 4850E is just a faster clock speed (in future articles, could you please have more details on the processors used, especially new/unknown/unavailable units?).

    They're all rated at 45w and have the same amount of cache, so that leaves that fact that the LE-1600 is built on 90nm vs. 65nm tech of the 4850E to be the only cause of this effect, though one wonders if there were other enhancements in the stepping that could be responsible too. That begs the question, what if a single core based on 65nm were used? The only I can find are the LE-1200 series, which are Semperons by definition and have less cache.

    Further, regarding the results that the 8200 board used less than the 780G, I question based on not knowing anything about the Biostar TF8200 A2+ at all. Does it have exactly the same features as the Gigabyte MA78GM-S2H? I.e., are there any other chipset differences on the boards that could account for the power draw? (1394, sound, raid, etc.) Also, do other 8200 based boards return the same results? As for that thought, how do other 780G boards fare (Asus, Biostar, ECS)? Why not put the TF8200 up against its Biostar 780G sibling the TFORCE TA780G M2+? I think calling the 8200 the winner is premature without any other data points.

    Final thought. Just how green can you go? By selecting the right CPU, memory, HD, PS, fans, cases, etc., what can we learn from this test that we can use to select greener components next time? Do really need a 2.5GHz 4850E, or will a 2.3GHz 4450E do as well and draw less? For that matter, just how does an LE-1250 fare? Sure it may not work out in a HTPC, but what can it do? Do we really need two optical drives, especially if the LG unit does it all? Can we use a lower wattage, high efficiency PS to good effect? Are there memory modules that are stable at 1.8v that draw fewer watts? Can only 2GB be used? What about 2x2GB instead of 4x1GB? etc., etc. etc.
  • cghebert - Friday, April 18, 2008 - link

    Regarding the differences between 8200 and 780g. The MAJOR one, for me anyway, is that the 8200 boards will be ably to output 8 channel uncompressed sound (for HD DVD and Blu Ray) from their HDMI ports. 780g (and every other HDMI motherboard, with the exception of the Asus Intel G35 board) only puts out the equivalent of 2 channel uncompressed audio (or 5.1 compressed, i.e. AC3).
    This is why I (and many others over at AVS forums) have been eagerly anticipating the 8200 boards.

    To me that is the most distinctive issue between these two chipsets. Beyond that, most of it just comes down to ATI vs. Nvidia video quality/driver issues.

    Much more info on this discussion here:

  • royalcrown - Sunday, April 20, 2008 - link

    You may want to wait and see about the quality issues as certain families of Nvidia chipsets were notorious for data corruption and it appears that that may be an issue with the 700 series according to a blog on here.

    Read all of the postcards from the edge, and maybe wait for some reviews to be safe ;)
  • lightzout - Saturday, April 19, 2008 - link

    Wow, the HDMI output and the potential to run a cooler, quieter box may mean I will finally be able to get my HTPC box into the living room and hooked up to the HDTV. I am as green as anyone can be without going off-grid with solar but the minimal differences in wattage here aren't the only upside. I want to have a silent pc that works great. It nice to see companies headed towards efficiency but improving other areas of the home are more likely to offer a substantial reduction of power usage overall.

    I want a low power, stable platform for MCE2005 that will fit into my Dvine box. Right now I use an underclocked nforce2 with a geforce 6200 which barely manages the workload. Its too noisy and consumes too much power to justify running 24/7 so I miss recording some shows. I think I may have finally waited long enough to upgrade.

    This was a great article and I hope Anandtech continues to follow other efforts to produce greener, quieter components for the "cooler" under-clocking crowd.
  • DanDeighan - Friday, April 18, 2008 - link

    These CPU's ARE available. Have been since the 15th. Mostly smaller vendors (see Froogle), but also available through buy.com or Amazon.com. Newegg is trailing on this one, not available there yet. Reply
  • Locutus465 - Friday, April 18, 2008 - link

    Seems like availability for AMD products is a little uneven these days... For instance I was going to just purchase my entire "new" rig from ZipZoomFly until I realized (thankfully I did some digging before ordering) that they don't offer any B3 Phenoms. Ended up getting the Phenom 9850be from newegg. Not sure why zipzoomfly is just sitting around with out any B3 parts. Reply
  • FireTech - Friday, April 18, 2008 - link

    bingbong may be in Australia where AMD Australia is reported (on various user forums) as stating low sales volumes of previous 45w CPUs here do not justify importing the new 4850e etc CPUs here.
    Not good news if it turns out to be true...
  • bingbong - Friday, April 18, 2008 - link

    Nice to see some good low power options with a decent intergrated graphics set. Awaiting the shoot out to see 780G or 8200 will be the best bet.

    I am kind of pissed that the AMD 4850e chip cannot be found.

    What's the point of reviewing this from February when it seems like AMD are not going to be making the CPU available.

    Come on AMD!!!

    Or don't you want our money!
  • Martimus - Friday, April 18, 2008 - link


    Plus many other places, but not on newegg yet.
  • LuxZg - Friday, April 18, 2008 - link

    First, I have to agree with Black Jacque, again I see reviews of low power systems done with higher power supplies, which is not good. Out of those 100-130W you probably have another 10% wasted power on PSU inefficiencies, which is sometimes as much as you save by choosing "right" chipset/CPU combo.

    Anyway, now that we know which chipsets are good and low power, and that they can run HD video, could you make an article that would show an alternative power source of around 180W? Let's say solar cells? I saw an article like that a while ago on Tom's (found it! http://www.tomshardware.com/reviews/technical-foun...">http://www.tomshardware.com/reviews/tec...dations-... ) and it would be nice to see a similar "investigation" after you finish this IGP comparision series :)
  • Calin - Friday, April 18, 2008 - link

    You are comparing the performance of an AMD single core (LE1600) processor against the dual core E2200 from Intel? Reply
  • Conroe - Saturday, April 19, 2008 - link

    Why not test with a 440? It's not that anyone would recommend a single core now with prices as they are. In my opinion dual core are even more green as you almost double processing power only adding a few more watts to the total. Either way I'm sure G35 still uses more power.
  • Black Jacque - Friday, April 18, 2008 - link

    I note they used a Corsair CMPSU-520HX for the testing. This PSU is inappropriately sized for the wattages reported.

    To get the highest efficiency out of that power supply you should run at about half its rated wattage.
  • strikeback03 - Friday, April 18, 2008 - link

    Might be the smallest PSU they have around. Other than the small Seasonics and the Antec Earthwatts series, are there many power supplies in the 300-400 watt range that are reasonably high quality and more efficient?

    Side note: Love how Newegg's "Play it green" banner shows an Antec TruePower Quattro 1000. Nothing says green like a thousand watt PSU.
  • cghebert - Friday, April 18, 2008 - link

    The ECS version of the 8200 is available now


    But I believe the rest won't be released until later this month due to incomplete drivers from Nvidia at this point.

    Gary, another great teaser article. I'm still sitting on my hands waiting for the full 780g and geforce 8200 roundup. Will any of the tests be done under Windows XP, or will it all be Vista?

  • Martimus - Friday, April 18, 2008 - link

    That's a good question. I know that I plan on using Linux on my HTPC that I am building. (planning on a 780G and 4850e X2, but may change my mind based on your article)

    Also, does the 4850e X2 come with a quiet heatsink, or is it rather loud? I am trying to make the system as quiet as possible.
  • gogos7 - Friday, April 18, 2008 - link

    You used an Asus motherboard in the Intel configuration.

    Asus M/Bs (the last 2 years) are well known for their high power consumption.

    What do you expect?
  • sprockkets - Sunday, April 20, 2008 - link

    They are the only one at the time that have a G35 board available. Reply
  • R3MF - Friday, April 18, 2008 - link

    I am sat on a wad of cash looking to buy a mATX gaming system for a SUGO 03 case, and completely stymied by the lack of decent AM2+ boards capable to REALLY taking 125W load processors.

    Both MSI and Abit have solid-caps 8200 boards advertised which is great, but there is complete silence over when they will be released!

    When can i buy an 8200 board?
  • spinportal - Friday, April 18, 2008 - link

    Hey Gary, shouldn't the last paragraph title be "Final Thoughts" instead of "First Thoughts"? Or did I read the article backwards? :) Reply
  • Visual - Friday, April 18, 2008 - link

    They do this very often - I understand it as "the product is very new, just launching, and a lot more testing is expected; so far, this is what we think", but I've been confused by it too. Reply
  • Visual - Friday, April 18, 2008 - link

    i am way out of the info loop now.
    are there no current nvidia igp chipsets for intel cpus?
  • smn198 - Friday, April 18, 2008 - link

    Thanks for the article. I found it interesting and glad that a better performing IGP doesn't have to mean worse power efficiency.

    I'd like the performance per watt stats I've seen you do before and also it would be good to get an indication of how much a difference in running costs each platform would have over the year having made some assumptions on typical usage.

    As you mentioned you focused on power which is important but there are many more considerations such as the materials and processes involved in making these components and the impact at EOL.

    Hope to see more like this!

Log in

Don't have an account? Sign up now