So what motherboard was used? I took a close look at the intel DX58SO and that thing looks like a piece of crap. I cannot believe people pay that much for a motherboard with the build quality of a $95 Asrock. Don't bother saying 'reliability' or 'it just works' because on newegg and other sites this mobo is racking up the complaints due to its shoddiness like no other else.
Hey, easy on Asrock. Personally, I find that Asrock boards are a great no fuss solution around(especially P55 chipset) when we speak of motherboards. Having built a few P55 and 1366 desktops, I find that they post virtually every time, fuss free. Overclock utilities just work, too. I have found them more stable than boards twice the price. I have yet to receive a DOA Asrock board. Not crazy about their color schemes though. My two cents.
Hi Anand, great site and awesome articles, thank you!
I have a question, at the price range we are looking at for the new CPU, how will a dual socket motherboard, with 2 quad core Xeons (i.e. 8 cores, 16 threads) compare with a single 980X (6 cores, 12 threads). I am not refering to the top of the range XEONS but to have two systems, with more or less the same priced CPU's go up against each other.
I've checked quickly, there are a fair number of XEONS retailing (newegg) for less than $500, which will make it an interesting comparison.
What will be super special, is if the benchmarking on a dual socket board include some gaming benchmarks as well.
Thank you, very nice indeed. So from my understanding of that screenshot, a dual socket, XEON setup with 8c, and 16t will outperform the $999 socket? Cool.
Agree. Now please excuse me, I have never really dealt with server prcoessors, but those specific ones used in the specific test, how much would they retail for?
Also, I am really curious as to how well that motherboard will hold up, with two un-overclocked XEONS, and a single graphics card (top of the range), in terms of gaming and FPS.
I am nearly due for an upgrade, closer to the end of the year. I still have the QX6700 with 2x8800GTX in SLI but am considering a dual socket board. I mainly play games on the PC, and granted, I can still play pretty much any game right now without issues, but I on average upgrade every 3 to 4 years and it is getting towards that time now...
Your opinion? Would you consider either the 980x, or rather a dual socket xeon setup at more or less the same price?
Thanks for the responses it is always interesting reading your views.
Since the 980x is an EE chip, doesn't that mean its unlocked? Whats the deal with not adjusting the multiplier? Wouldn't that yield a better overclock?
Just because you can max the multiplier, doesn't mean it's limitless on its overclock. Otherwise it would max to an infinite Hz.
Sometimes it's better to lower the clock and raise the multiplier, sometimes it's not. Though, todays Intel CPUs have other things going on then simply raising/lowering clocks and adjusting multipliers, just because of the L3 Cache and other nuances.
I agree its not infinite(limitless as you put it). But, why didn't Anand explore the option? That's my point...I do understand it's not his intention to reach overclocking limits though, thats left to other sites. I just feel the topic should have been discussed.
When will we have the ability to just 'overclock' the TurboBoost maximums (for 1 core, 2 cores, 4 cores, etc.)?
Why should any Corei5/i7 chip, with all its smarts, be forced to idle at +4Ghz?
With the power gating and all it might not be the cores drawing that power, could be either uncore stuff or other motherboard components (like the RAM).
This was the first thing that came to mind as well. Your original review placed a lot of the benefit of this chip in the fact that you could use an 18-month old mobo and get great performance. Now, yes at stock and slightly elevated OC-levels you can, I'm just wondering if the new mobo has something that the ones meant for the 920 don't or whether it's just a bios issue that can be fixed/improved with a firmware update.
Why are all these tests being done with the stock cooler? Anyone
spending $1K on this CPU is not going to overclock with the supplied
cooler. Until we have numbers from using a better cooler like a TRUE
or something, it's hard to judge the chip's proper oc potential.
Re video encoding (the review article), more cores are all very well,
but not many apps support more than 4. Indeed, some codec paths
don't even support 2 (eg. MJPEG). Given the cost involved with an
X58/980X setup, it may be more sensible to simply get two separate
Lynfield i7 860 systems, process more than one video file at a time.
Just a shame the price is so high for the 980X, but until AMD has
some kind of performance option, Intel doesn't need to sell for
anything less (sensible business practice from their point of view
I suppose). I'd originally planned to get a 920 system last year
for video encoding, with an eye toward the 6-core upgrade option,
but I'm glad I didn't; with hindsight, Lynfield looks far more
sensible, the 860 inparticular. At the budget end of course, an oc'd
Athlon2 X4 630 is hard to ignore.
Btw, how high does the POV-ray test scale? ie. how many threads can
it spawn? I was wondering how the test would behave on an SGI
UltraViolet system which will take the new 6-core Gulftown, ie. a
single combined shared-memory system with up to 256 x 6-core XEONs.
No need to use two systems. You can process more than one video file on the same system. If you have 8 cores but the program only uses 4, then you can run two instances and it will use all 8. I've done it many times. As long as the program can run more than one instance. Most can be made to.
"Why are all these tests being done with the stock cooler? Anyone
spending $1K on this CPU is not going to overclock with the supplied
cooler. Until we have numbers from using a better cooler like a TRUE
or something, it's hard to judge the chip's proper oc potential."
Oh good grief. Have you ever heard of "base line" overclocking? You establish a base line for future reference (IE: how high can you get without raising voltage or how high can you get on the stock cooler, etc.). From that point you then upgrade the cooler, raise the voltage, and on and on. Overclock much?
Basic computer literacy advice - when typing text in a text area, do not insert line-breaks (means, do not press Enter) just because you think there is no more space left on the current line. Computers are not type-writers, they can wrap text to new lines automatically according to the available space, which will often be different for the different places that the text might be presented on, as well as for the different users with different user agents each with potentially different font settings...
Yeah, I was a bit surprised when I saw the Swedish hardware site Nordichardware get it up to 4.4GHz without much difficulty as soon as I had finished reading Anandtechs review. :)
But dont despair Anand, you guys are fantastic at other things. :)
So if you wanted an apples to apples comparison, the data is there for a stock i7 975 vs stock i7 980X. That's your comparison of like for like and gives you an idea how much faster the 980X is. You can always extrapolate the i7 975 to get a representation of it's performance overclocked against the 980X overclocked. I don't see why the authors would want to re-run all those tests on a 975 to cover that scenario.
just stable enough to run the benchmarks? or total system stability? always wondered about numbers sites show with oc's...theres a huge dif between being able to boot and show a screen and a system thats considered oc stable. with the numbers being thrown out can you run prime small fft stable for at least 3 hours? just curious :/
Yeah, given that this processor will in most cases be used for more serious stuff than gaming, I'd like to know how many hours of stress testing can it take before producing an error.
ye, not to mention it really depends what you stress test with. Found out hard way few years ago that usual stress test can run over 24hrs no problem and then I ran optimized Seti@Home (using (S)SSEx) and got in a BSOD in few minutes resulting in need to push up voltage few steps higher to keep it rock stable.
Anand, can you quiz your Intel guy on the release of a mainstream version? AMD will have at least 3 Thuban 6 cores, Intel have a range of Xeon 32nm 6 cores, it doesnt make sense that this is the only desktop version. It would be a bad business decision and Intel are not stupid.!!
Yeah, as if the initial ~79W was not enough at idle lol.
I suppose it is pretty cool that you *can* do that when you wish, but personally, I'd rather see Intel ( and AMD for that matter ) put out more efficient CPUs. Then while we're at it, I would like to see nvidia, and AMD do the same things with their GPUs . . .
That's total system power consumption, including loss at power supply, for a six core I7 Extreme with a Radeon 5870. 79W -- about one light bulb. Are you really criticizing power consumption here?
1 ancient light bulb. Afaik production of those is even forbidden in EU, thay are THAT efficient. Modern version of the same power would eat 17-20 Watt.
Eating 80 Watt when doing nothing? There is, obviously, huge room for improvements.
I have an electric heater that uses 1500 Watts for most of the winter. The day they come out with a 1500W CPU is the day I start caring what the power consumption is (as a bonus, I'd have something to replace that heater with).
Doing nothing? You think that when a computer is idle it does "nothing"? Does your car not use any gas while "idling"? Turn your computer off, then it will be doing "nothing" and will use 0 watts.
Actually that's not entirely true either. Just being plugged in a pc (turned off) and monitor (in standby) are probably drawing anywhere from 7-35 watts depending on the pc, pc power supply switch and the type of monitor.
The only TRUE way to prevent power draw is to completely unplug everything.
That's why I hate going back to EU. You worry about stuff like that. You have a government that tells what's right and what's wrong. You need to start spending more and everything will get cheaper so you dont have to worry about 50 watts. And you if you think the environment, even the most pisimistic treehugging scientific forcast give us about 300,000 years before we destroy the earth. That's like 50 times more years that the humans as we know them have been around.
Back on topic, AT, can you test/overclock with a good aftermarket cooler. First we can see the actual potential of the CPU on air and second we can see how effective is the new Intel cooler. Thanks!
This is why I keep checking Anandtech every 5 mins :) Screw RSS notifications, manual is better, and AT gets me coming back.
Perhaps someone has time for some detailed overclocking analysis. That is, not just overclocking the proc, but the combined synergy and sweet spot for overclocking the proc/ram/gpu in such a way that yields the best results.
By RAM this includes gpu ram and by overclocking, this includes memory timings.
If only I had the resources, I'd sit down for a week and test, test, test :)
Tweaking Ram timings, etc in modern CPU architectures doesn't yield anything outside of benchmarks. It was different back in the p4 era where the memory bus was a bottleneck, but since amd64/core architectures that hasn't been the case outside of pathological circumstances. At stock speeds the dual channel LGA1156 controller provides 75% of the worst case memory bandwidth consumption for a quadcore system; a level that only synthetic benchmarks can exceed. For quads LGA1366 provides 100%; for the 6core model it again drops down to about 75%. High end memory only mattes in that it becomes something you don't need to fiddle with while overclocking.
For some reason, I still have a hard time believing that modern systems are unaffected by such timings, not just CAS Latency, but the RAS-to-CAS delay, or even the tRAS. The memory bus may have increased, but the clock has basically stayed the same (266MHz and less) and the memory timings have gone up.
If timings have gotten worst and the core clock has remained the same, then adjusting the clock/timings on those sticks today should have even a greater impact on today's sticks - this includes real life and benchmarks (the difference b/t the two is how likely your real life use will reach the benchmark scenario).
---
What might even be nicer than this is to see if it's possible to fiddle with SSD RAM settings :) I'd like to see what kind of affect a 10% increase in clock speed, or 25% drop in timings would have! While I personally wouldn't mess with the stability of SSDs, this may be an option in the future, when I have spares laying around. As SSD controllers have become dependent on their RAM, I'd really like to see the impact this would have - perhaps the more common, cost-effective products on the market could surpass the 3Gbps threshold.
hello, I have also asked this question (I have a gulftown Q3QP 3.07 ghz) and I went to the intel site (http://ark.intel.com/Product.aspx?id=47932)">http://ark.intel.com/Product.aspx?id=47932) and the answer is that the x980 is not dual QPI,
I contacted EVGA to ask if my processor is compatible with the new motherboard classified sr-2, and I was told that there is no problem in any case it is better EVGA contact for more details, I hope to be helpful.
The reality is that in any modern CPU, with most programs, most memory requests are serviced by the cache. Once you get out of the cache you're totally screwed in any event.
That's why Core 2 managed to be competitive with (and often faster than) Athlon 64 despite having a dramatically slower memory subsystem.
Ironically, a design like Atom would benefit much more from faster memory than something like Nehalem. Atom has smaller caches and is in-order, so you get more capacity misses and the CPU can't order around them.
Similarly, having six cores will give little additional bang for the buck if (as on my 3.8GHz quad core system with fast hard drives) almost all of the bottlenecks are in disk I/O. Some video encoding and scientific apps will go a lot quicker with this chip, but gamers / enthusiasts will be much better off planning for the move to SSDs. For us, CPU speeds and especially number of cores are no longer the limiting factors on overall performance. Even graphics cards are ahead of the curve. Now it's disks - which is one reason why the excellent Anandtech columns on SSDs are becoming my favorite reads.
I laughed when I read that Intel told him he's not doing something right. They were like: we specificially hand tuned and binned these review chips to show us in the best light possible. So your definitely doing something wrong Anand!
WHY are you overclocking with the stock cooler?? Your results are still the worst on the web, sadly. 4.1Ghz is nothing to brag about. Surely you have some other HSF lying around you could really test with.
Because not everyone uses the same aftermarket cooler. However, there's a good chance that every cooler somoene might use has been reviewed and compared to it's stock counterpart. With that information in tow you have a good idea (regardless of CPU) what type of performance gain you can achieve with this CPU and your cooler of choice.
I heard before that having a number of cores, say 2, doesn't mean you can guesstimate its speed, say 2GHz is going to be similar to a 2x2=4GHz single core. I'm sure it can get complicated with "architecture differences" (I use that term with the tiniest level of understanding about what it is), but if you consider I'm a layman looking initially for a simple bitesize digestible tidbit, can you approximate what kind of relative proportion of speed you may expect?
For example, a dual core 2GHz may be expected to reach around what? 90% 80% 70% 60%? the speed of a single 4GHz core (other factors being similar)? Perhaps I could crudely call that efficiency.
And, having answered that question with dual cores, when you start increasing to 6, 8, or more cores, does the "efficiency" stay about the same, improve, or get worse generally speaking? And finally, is it possible to apply some kind of simplified crude equation or fraction to very roughly approximate the relationship between core number and "efficiency"?
If any of my 2ish long questions made any sense, thanks for any answers.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
57 Comments
Back to Article
drewintheav - Tuesday, March 16, 2010 - link
The INTEL i7 980X has dual QPI's and will run in a dual socket mainboard!!!Such as the EVGA W555 / Classified SR-2
magnes79 - Thursday, December 9, 2010 - link
That is not true. i7 980X has single QPI. check intel website. All i7 have single QPI. Thats why there is XEON series.Dainas - Friday, March 12, 2010 - link
So what motherboard was used? I took a close look at the intel DX58SO and that thing looks like a piece of crap. I cannot believe people pay that much for a motherboard with the build quality of a $95 Asrock. Don't bother saying 'reliability' or 'it just works' because on newegg and other sites this mobo is racking up the complaints due to its shoddiness like no other else.1stguess - Friday, March 19, 2010 - link
Hey, easy on Asrock. Personally, I find that Asrock boards are a great no fuss solution around(especially P55 chipset) when we speak of motherboards. Having built a few P55 and 1366 desktops, I find that they post virtually every time, fuss free. Overclock utilities just work, too. I have found them more stable than boards twice the price. I have yet to receive a DOA Asrock board. Not crazy about their color schemes though. My two cents.Skouperd - Friday, March 12, 2010 - link
Hi Anand, great site and awesome articles, thank you!I have a question, at the price range we are looking at for the new CPU, how will a dual socket motherboard, with 2 quad core Xeons (i.e. 8 cores, 16 threads) compare with a single 980X (6 cores, 12 threads). I am not refering to the top of the range XEONS but to have two systems, with more or less the same priced CPU's go up against each other.
I've checked quickly, there are a fair number of XEONS retailing (newegg) for less than $500, which will make it an interesting comparison.
What will be super special, is if the benchmarking on a dual socket board include some gaming benchmarks as well.
As I said, great site, I love it.
aigomorla - Friday, March 12, 2010 - link
http://i125.photobucket.com/albums/p73/aigomorla/P...">http://i125.photobucket.com/albums/p73/aigomorla/P...Ask and you shall receive.
However there's a lot more posted inside our forums.
Skouperd - Sunday, March 14, 2010 - link
Thank you, very nice indeed. So from my understanding of that screenshot, a dual socket, XEON setup with 8c, and 16t will outperform the $999 socket? Cool.Thanks for the response.
JumpingJack - Monday, March 15, 2010 - link
Pay attention, though, to clockspeeds, 8c/16t @ 3.2 Ghz, adjust for OC, price, etc accordingly... but yeah, it looks like it would be a winner.Skouperd - Monday, March 15, 2010 - link
Agree. Now please excuse me, I have never really dealt with server prcoessors, but those specific ones used in the specific test, how much would they retail for?Also, I am really curious as to how well that motherboard will hold up, with two un-overclocked XEONS, and a single graphics card (top of the range), in terms of gaming and FPS.
I am nearly due for an upgrade, closer to the end of the year. I still have the QX6700 with 2x8800GTX in SLI but am considering a dual socket board. I mainly play games on the PC, and granted, I can still play pretty much any game right now without issues, but I on average upgrade every 3 to 4 years and it is getting towards that time now...
Your opinion? Would you consider either the 980x, or rather a dual socket xeon setup at more or less the same price?
Thanks for the responses it is always interesting reading your views.
PyroHoltz - Friday, March 12, 2010 - link
Since the 980x is an EE chip, doesn't that mean its unlocked? Whats the deal with not adjusting the multiplier? Wouldn't that yield a better overclock?-Confused
aigomorla - Friday, March 12, 2010 - link
they are unlocked.But you get better scores from a higher QPI then a higher multi because of the reduced latency when having a higher QPI.
;)
vol7ron - Friday, March 12, 2010 - link
Just because you can max the multiplier, doesn't mean it's limitless on its overclock. Otherwise it would max to an infinite Hz.Sometimes it's better to lower the clock and raise the multiplier, sometimes it's not. Though, todays Intel CPUs have other things going on then simply raising/lowering clocks and adjusting multipliers, just because of the L3 Cache and other nuances.
PyroHoltz - Friday, March 12, 2010 - link
I agree its not infinite(limitless as you put it). But, why didn't Anand explore the option? That's my point...I do understand it's not his intention to reach overclocking limits though, thats left to other sites. I just feel the topic should have been discussed.Foggg - Friday, March 12, 2010 - link
When will we have the ability to just 'overclock' the TurboBoost maximums (for 1 core, 2 cores, 4 cores, etc.)?Why should any Corei5/i7 chip, with all its smarts, be forced to idle at +4Ghz?
strikeback03 - Friday, March 12, 2010 - link
With the power gating and all it might not be the cores drawing that power, could be either uncore stuff or other motherboard components (like the RAM).Spivonious - Friday, March 12, 2010 - link
What mobo did you switch to? I'd like to see an analysis as to why the Intel board was so much more limited.7Enigma - Friday, March 12, 2010 - link
This was the first thing that came to mind as well. Your original review placed a lot of the benefit of this chip in the fact that you could use an 18-month old mobo and get great performance. Now, yes at stock and slightly elevated OC-levels you can, I'm just wondering if the new mobo has something that the ones meant for the 920 don't or whether it's just a bios issue that can be fixed/improved with a firmware update.mapesdhs - Friday, March 12, 2010 - link
Why are all these tests being done with the stock cooler? Anyone
spending $1K on this CPU is not going to overclock with the supplied
cooler. Until we have numbers from using a better cooler like a TRUE
or something, it's hard to judge the chip's proper oc potential.
Re video encoding (the review article), more cores are all very well,
but not many apps support more than 4. Indeed, some codec paths
don't even support 2 (eg. MJPEG). Given the cost involved with an
X58/980X setup, it may be more sensible to simply get two separate
Lynfield i7 860 systems, process more than one video file at a time.
Just a shame the price is so high for the 980X, but until AMD has
some kind of performance option, Intel doesn't need to sell for
anything less (sensible business practice from their point of view
I suppose). I'd originally planned to get a 920 system last year
for video encoding, with an eye toward the 6-core upgrade option,
but I'm glad I didn't; with hindsight, Lynfield looks far more
sensible, the 860 inparticular. At the budget end of course, an oc'd
Athlon2 X4 630 is hard to ignore.
Btw, how high does the POV-ray test scale? ie. how many threads can
it spawn? I was wondering how the test would behave on an SGI
UltraViolet system which will take the new 6-core Gulftown, ie. a
single combined shared-memory system with up to 256 x 6-core XEONs.
Hmm... :D
Ian.
Shadowmaster625 - Friday, March 12, 2010 - link
No need to use two systems. You can process more than one video file on the same system. If you have 8 cores but the program only uses 4, then you can run two instances and it will use all 8. I've done it many times. As long as the program can run more than one instance. Most can be made to.Nfarce - Friday, March 12, 2010 - link
"Why are all these tests being done with the stock cooler? Anyonespending $1K on this CPU is not going to overclock with the supplied
cooler. Until we have numbers from using a better cooler like a TRUE
or something, it's hard to judge the chip's proper oc potential."
Oh good grief. Have you ever heard of "base line" overclocking? You establish a base line for future reference (IE: how high can you get without raising voltage or how high can you get on the stock cooler, etc.). From that point you then upgrade the cooler, raise the voltage, and on and on. Overclock much?
Visual - Friday, March 12, 2010 - link
Basic computer literacy advice - when typing text in a text area, do not insert line-breaks (means, do not press Enter) just because you think there is no more space left on the current line. Computers are not type-writers, they can wrap text to new lines automatically according to the available space, which will often be different for the different places that the text might be presented on, as well as for the different users with different user agents each with potentially different font settings...Shadowmaster625 - Friday, March 12, 2010 - link
Actually his post was easier to read than yours. lolformulav8 - Friday, March 12, 2010 - link
I definitely agree! :)Barneyk - Friday, March 12, 2010 - link
Yeah, I was a bit surprised when I saw the Swedish hardware site Nordichardware get it up to 4.4GHz without much difficulty as soon as I had finished reading Anandtechs review. :)But dont despair Anand, you guys are fantastic at other things. :)
aigomorla - Friday, March 12, 2010 - link
if you would take a few moments to browse our forums,You'll see i did it 3 months ago.
http://forums.anandtech.com/showthread.php?t=20450...">http://forums.anandtech.com/showthread.php?t=20450...
stw500 - Friday, March 12, 2010 - link
Why do you compare a high overclocked brand new cpu only with non overclocked cpus...that makes only little sens...Much more interesting is a comparing of for example a C2D @ 4 Ghz versus i7 980 @ 4 Ghz... how much faster are these cpus eye to eye...
sbrown23 - Friday, March 12, 2010 - link
So if you wanted an apples to apples comparison, the data is there for a stock i7 975 vs stock i7 980X. That's your comparison of like for like and gives you an idea how much faster the 980X is. You can always extrapolate the i7 975 to get a representation of it's performance overclocked against the 980X overclocked. I don't see why the authors would want to re-run all those tests on a 975 to cover that scenario.stw500 - Friday, March 12, 2010 - link
i meant which of these cpus and how much is faster... ;)glockjs - Friday, March 12, 2010 - link
just stable enough to run the benchmarks? or total system stability? always wondered about numbers sites show with oc's...theres a huge dif between being able to boot and show a screen and a system thats considered oc stable. with the numbers being thrown out can you run prime small fft stable for at least 3 hours? just curious :/killerclick - Friday, March 12, 2010 - link
Yeah, given that this processor will in most cases be used for more serious stuff than gaming, I'd like to know how many hours of stress testing can it take before producing an error.Holly - Friday, March 12, 2010 - link
ye, not to mention it really depends what you stress test with. Found out hard way few years ago that usual stress test can run over 24hrs no problem and then I ran optimized Seti@Home (using (S)SSEx) and got in a BSOD in few minutes resulting in need to push up voltage few steps higher to keep it rock stable.shin0bi272 - Friday, March 12, 2010 - link
"That award would have to go to the time I wrote a review and left out the 'l' in overclocking."Better than going off half-cocked. That could get you killed.
cjs150 - Friday, March 12, 2010 - link
Could be even worse if he was into under-cockingcactusdog - Friday, March 12, 2010 - link
Anand, can you quiz your Intel guy on the release of a mainstream version? AMD will have at least 3 Thuban 6 cores, Intel have a range of Xeon 32nm 6 cores, it doesnt make sense that this is the only desktop version. It would be a bad business decision and Intel are not stupid.!!yyrkoon - Thursday, March 11, 2010 - link
Yeah, as if the initial ~79W was not enough at idle lol.I suppose it is pretty cool that you *can* do that when you wish, but personally, I'd rather see Intel ( and AMD for that matter ) put out more efficient CPUs. Then while we're at it, I would like to see nvidia, and AMD do the same things with their GPUs . . .
Sivar - Friday, March 12, 2010 - link
That's total system power consumption, including loss at power supply, for a six core I7 Extreme with a Radeon 5870. 79W -- about one light bulb. Are you really criticizing power consumption here?medi01 - Friday, March 12, 2010 - link
1 ancient light bulb. Afaik production of those is even forbidden in EU, thay are THAT efficient. Modern version of the same power would eat 17-20 Watt.Eating 80 Watt when doing nothing? There is, obviously, huge room for improvements.
stephenbrooks - Sunday, March 14, 2010 - link
I have an electric heater that uses 1500 Watts for most of the winter. The day they come out with a 1500W CPU is the day I start caring what the power consumption is (as a bonus, I'd have something to replace that heater with).CptTripps - Friday, March 12, 2010 - link
Doing nothing? You think that when a computer is idle it does "nothing"? Does your car not use any gas while "idling"? Turn your computer off, then it will be doing "nothing" and will use 0 watts.Iketh - Saturday, March 13, 2010 - link
The day i have only SSDs in my system is the day i'll start turning my system off againmasouth - Saturday, March 13, 2010 - link
Actually that's not entirely true either. Just being plugged in a pc (turned off) and monitor (in standby) are probably drawing anywhere from 7-35 watts depending on the pc, pc power supply switch and the type of monitor.The only TRUE way to prevent power draw is to completely unplug everything.
jonup - Friday, March 12, 2010 - link
That's why I hate going back to EU. You worry about stuff like that. You have a government that tells what's right and what's wrong. You need to start spending more and everything will get cheaper so you dont have to worry about 50 watts. And you if you think the environment, even the most pisimistic treehugging scientific forcast give us about 300,000 years before we destroy the earth. That's like 50 times more years that the humans as we know them have been around.Back on topic, AT, can you test/overclock with a good aftermarket cooler. First we can see the actual potential of the CPU on air and second we can see how effective is the new Intel cooler. Thanks!
busta - Friday, March 12, 2010 - link
You really need to work on your understanding of science and historyOR
stop commenting on things you obviously don't understand.
medi01 - Friday, March 12, 2010 - link
1 ancientvol7ron - Thursday, March 11, 2010 - link
This is why I keep checking Anandtech every 5 mins :) Screw RSS notifications, manual is better, and AT gets me coming back.Perhaps someone has time for some detailed overclocking analysis. That is, not just overclocking the proc, but the combined synergy and sweet spot for overclocking the proc/ram/gpu in such a way that yields the best results.
By RAM this includes gpu ram and by overclocking, this includes memory timings.
If only I had the resources, I'd sit down for a week and test, test, test :)
vol7ron
DanNeely - Friday, March 12, 2010 - link
Tweaking Ram timings, etc in modern CPU architectures doesn't yield anything outside of benchmarks. It was different back in the p4 era where the memory bus was a bottleneck, but since amd64/core architectures that hasn't been the case outside of pathological circumstances. At stock speeds the dual channel LGA1156 controller provides 75% of the worst case memory bandwidth consumption for a quadcore system; a level that only synthetic benchmarks can exceed. For quads LGA1366 provides 100%; for the 6core model it again drops down to about 75%. High end memory only mattes in that it becomes something you don't need to fiddle with while overclocking.vol7ron - Friday, March 12, 2010 - link
That's good to know.For some reason, I still have a hard time believing that modern systems are unaffected by such timings, not just CAS Latency, but the RAS-to-CAS delay, or even the tRAS. The memory bus may have increased, but the clock has basically stayed the same (266MHz and less) and the memory timings have gone up.
If timings have gotten worst and the core clock has remained the same, then adjusting the clock/timings on those sticks today should have even a greater impact on today's sticks - this includes real life and benchmarks (the difference b/t the two is how likely your real life use will reach the benchmark scenario).
---
What might even be nicer than this is to see if it's possible to fiddle with SSD RAM settings :) I'd like to see what kind of affect a 10% increase in clock speed, or 25% drop in timings would have! While I personally wouldn't mess with the stability of SSDs, this may be an option in the future, when I have spares laying around. As SSD controllers have become dependent on their RAM, I'd really like to see the impact this would have - perhaps the more common, cost-effective products on the market could surpass the 3Gbps threshold.
I'd like to know what you think,
vol7ron
drewintheav - Sunday, March 14, 2010 - link
i heard that the i7 980x has dual qpi's and can run in a dual cpu configuration. like on an evga 270-gt-w555. can anybody confirm this?gigi127 - Saturday, March 20, 2010 - link
hello, I have also asked this question (I have a gulftown Q3QP 3.07 ghz) and I went to the intel site (http://ark.intel.com/Product.aspx?id=47932)">http://ark.intel.com/Product.aspx?id=47932) and the answer is that the x980 is not dual QPI,I contacted EVGA to ask if my processor is compatible with the new motherboard classified sr-2, and I was told that there is no problem in any case it is better EVGA contact for more details, I hope to be helpful.
bsoft16384 - Saturday, March 13, 2010 - link
The reality is that in any modern CPU, with most programs, most memory requests are serviced by the cache. Once you get out of the cache you're totally screwed in any event.That's why Core 2 managed to be competitive with (and often faster than) Athlon 64 despite having a dramatically slower memory subsystem.
Ironically, a design like Atom would benefit much more from faster memory than something like Nehalem. Atom has smaller caches and is in-order, so you get more capacity misses and the CPU can't order around them.
Arbie - Friday, March 12, 2010 - link
Similarly, having six cores will give little additional bang for the buck if (as on my 3.8GHz quad core system with fast hard drives) almost all of the bottlenecks are in disk I/O. Some video encoding and scientific apps will go a lot quicker with this chip, but gamers / enthusiasts will be much better off planning for the move to SSDs. For us, CPU speeds and especially number of cores are no longer the limiting factors on overall performance. Even graphics cards are ahead of the curve. Now it's disks - which is one reason why the excellent Anandtech columns on SSDs are becoming my favorite reads.Jalek99 - Thursday, March 11, 2010 - link
LMAOThanks for keeping up with the hardware reviews, I seem to be finding less of them these days.
formulav8 - Friday, March 12, 2010 - link
I laughed when I read that Intel told him he's not doing something right. They were like: we specificially hand tuned and binned these review chips to show us in the best light possible. So your definitely doing something wrong Anand!Hauk - Thursday, March 11, 2010 - link
Much better! Faith in AT's overclocking ability restored..RaistlinZ - Friday, March 12, 2010 - link
Anand,WHY are you overclocking with the stock cooler?? Your results are still the worst on the web, sadly. 4.1Ghz is nothing to brag about. Surely you have some other HSF lying around you could really test with.
Arkive - Friday, March 12, 2010 - link
Because not everyone uses the same aftermarket cooler. However, there's a good chance that every cooler somoene might use has been reviewed and compared to it's stock counterpart. With that information in tow you have a good idea (regardless of CPU) what type of performance gain you can achieve with this CPU and your cooler of choice.
Oliver55 - Thursday, August 26, 2010 - link
Hi, newbie in every O.C/non O.C sense.I heard before that having a number of cores, say 2, doesn't mean you can guesstimate its speed, say 2GHz is going to be similar to a 2x2=4GHz single core. I'm sure it can get complicated with "architecture differences" (I use that term with the tiniest level of understanding about what it is), but if you consider I'm a layman looking initially for a simple bitesize digestible tidbit, can you approximate what kind of relative proportion of speed you may expect?
For example, a dual core 2GHz may be expected to reach around what? 90% 80% 70% 60%? the speed of a single 4GHz core (other factors being similar)? Perhaps I could crudely call that efficiency.
And, having answered that question with dual cores, when you start increasing to 6, 8, or more cores, does the "efficiency" stay about the same, improve, or get worse generally speaking? And finally, is it possible to apply some kind of simplified crude equation or fraction to very roughly approximate the relationship between core number and "efficiency"?
If any of my 2ish long questions made any sense, thanks for any answers.