In the article you seem to imply a tremendous difference in that 2meg to 3meg jump but your single data point doesn't show that conclusion.
In your example there is a ~9% increase in performance when going from 2 to 3 megs, but there is also a 4.5% increase in clockspeed between the 2 chips. So at best (in this example) you have a 5% difference, which while significant, I would't say it was tremendous.
Can you comment on whether you did any further tests with identical frequencies?
1. AMD Athlon II X2 250 results missing a dot, becoming slower than Intel Atom 230.
2. Better using "Intel Pentium Dual Core E5300" for E5300 as the name Intel printed on the CPU.
3. AMD Athlon LE-1620 notes (platform side notes) should be same as AMD Athlon LE-1640.
Thanks for adding more CPUs for comparison in AnandTech Bench, doing the full test in latest software is really time consuming especially for old CPUs. Take care.
I never had much luck running Cool n Quiet when I used to run AMD processors. Even in games the CPU was put into a low clock mode running at half speed. Sounds like a buggy implementation to this day.
And I cant say in my experience it was the Mother Boards fault either. I enabled Cool and Quiet in the bios and nothing happened. It wasnt until I installed the Cool n Quiet driver all hell broke loose.
However one has to keep in mind the bug comes from Redmond.
The Win 5.x scheduler is a piece of garbage and it allways was. It is just that it shows with different symptoms:
1 CPU core => crappy multi-tasking on heavy loads
2+ CPU cores => unable to get a grasp that power management not a thing of the future ...
A no, being from 2000 is no excuse, proper OS scheduler were there 30yrs ago.
Had MS decided to do so, their Win 5.x schedules would be a proper one from the start.
As for AMD, it is a price one pays for being ahead of the market.
Remember, it took Intel 5yrs (2003-2008) to get comparable power management capabilities as AMD with Nehalem.
YAY 3D max, also that E7500 is beating the crap out of everything but AMD's top of the line. That really solitifies in my mind that picking the E8400 6 months ago for my friends gaming computer was the right choice. Stupid AMD, I REALLY want them to be competitive again so Intel is FORCED to stop overcharging for EVERYTHING.
Just like that article that you linked to with the name CnQ bug, I don't believe that a hardware manufacturer creates a bug if the OS scheduler can't seem to do the sensible thing and not bounce threads around. Although various factors in the AMD chips make the problem worse, bouncing threads across different cores on an Intel chip also has a bit of impact, since L1 and the L2 shouldn't be shared across cores on the i7 platform.
So please, stop calling it a bug. The bug lies in Windows, not the chips.
I agree with almost everything you say, I only have a small caveat.
Intel chips will suffer much less from this than AMD, since they have an inclusive cache architecture, and can readily read the information from the L3 cache. I still think AMD has an exclusive cache arrangement, which I really think they should change with regards to the L3 cache for reasons just like the one you mention.
For what's it's worth, Microsoft screwed Intel 14 years back when the Pentium Pro was released. Naturally, Intel got the blame for having miserable 16-bit performance (it was related to segmentation, which was part of all the 16-bit modes, and technical, even 32-bit mode even though it was transparent), because Microsoft told them the world will be 32-bit by then. Of course Windows 95 had a lot of legacy code, and Windows NT, which we called "Not There" at the time, was about as common as a 20 year old virgin in western Europe. So, Intel took the blame, just like AMD is now, despite, once again, Microsoft's incompetence.
Really, if you think about it, the ability to clock the processors differently could be a very useful features, except for the fact Winblows can't use it properly.
First, I like seeing the Pentium 4s in the benchmarks, it was kind of interesting. They did better than I thought they would, and it makes me even more curious what they would be like on 45nm, since their clock speeds would probably be astronomical (since 45nm has much better power characteristics, and the clock speed limiter on the Pentium 4 was power use/heat).
But, anyway, why not use the Pentium 4 670 (3.8 GHz), or Pentium EE 965 Extreme Edition (3.73 GHz) processors? Why use the next to fastest ones?
Don't get me wrong, it was still informative, but I would have liked to see the fastest measured against today's processors, not one step removed. Even so, it was nice to see them, so it's just a minor complaint. I'm looking forward to seeing the Nano.
Might not have had any around. Figure the "best of breed" were the most likely to be either sold or go in a system for some family member when they were no longer needed for comparison duties.
I am curious as to why the current cheapest Intel quad core were not included in the benches yet you bothered to go grab the previously $999 Pentium 4 and then also included the Phenom X4 940.
It's more than likely due to the fact that Intel has much higher cache densities than AMD does. It probably had very little to nothing to do with the actual process (well, beyond the geometry size, that is).
Typically, cache is very dense, so you will notice transistor count increasing disproportionately to size as you add cache.
With respect to the Athlon II X2 being larger than the Penryn, that's not really a bad thing, since it does more too; the Penryn needs a memory controller on the chipset that the Athlon II does not.
Based on above die size chart. Core2Duo 107mm2 and 410million transistors while Athlon II has only 234million transistors on 117mm2. It's almost half of the number of Intel used on theirs. IMC take that much space?
Well, take a look at the Pentium version, and you'll see the what I was saying about the cache. We both can agree it's the same core, but one has a larger cache.
The Pentium is 82 mm2, with 228 million transistors, with a 2.064 megabytes of cache (L1 + L2). But, since 1 megabyte is disabled, it's really 3.064, like the other Wolfdale's have. The 6 MB version of the Core 2 is 107 mm2 with 410 million transistors.
So, you can see that adding 3 MB of cache increased the transistor count by 182 million, but the size by only 25 mm2. Or, in other words, it increased transistors by about 80%, but size by about 31%. So, cache does increase transistor count disproportionately to die size.
Oh, and yes, the IMC is quite large. You can view some of the pictures of the CPU die to see it, but it's far from insignificant in size.
its really hard to believe ad a lil concern for upcoming c2d processors --- why that 6300 sample didnt overclock? wat was wrong 3.7 i xpct normally but i m amazed as its a 45nm.... not 90 or 65
I like how you included the overclockability at stock voltage. Since all of my computers use stock heatsinks, overclocking at stock voltage is pretty much all I can do.
I used the standard heatsink off the Phenom II 940 on the X2 250, 1.4V, and hit 3.825GHz in Win7 64, idle temps around 33C, full load at 54C. I have retail CPUs coming with the new cost reduced heatsink to see what happens but any stock AM2+ heatsink from a Phenom/PhenomII works wonderfully for overclocking these two CPUs.
It's more of a timing issue than anything else. I spent a lot of last week working on Lynnfield and there's one more CPU review before this week is up. I am curious about it and will look into it shortly though :)
Hi Anand. Great review, it is much appreciated. The first thing I looked for was a comparison to the Brisbane CPU. To me, a comparison to Brisbane and a comparison to Penryn are the two most interesting.
I'm guessing if Phenom II X2s use harvested cores, they'll eventually be a limited commodity. Still, the Athlon II X2 seems to have a lot of scaling room left so higher clocked Athlon II X2s will probably do just as well.
I just flipped through the benchmarks this morning, but since I was one of those commenting on wanting to see the best of dual core Netburst processors like the 965EE for comparison, I wanted to thank you for deciding to include the 955EE.
As a follow-up is AMD still making 90nm dual cores or have they converted the fab to 45nm now? I believe the previous fastest dual core was still the 3.2GHz 90nm Athlon 64 X2 6400+. It's too bad it wasn't included here.
The X2 6400+ is in the charts now and you can always use our Bench tool to compare a whole litany of processors against each other. AMD is currently phasing out of 90nm production and even several 65nm products will be phased out this year as they ramp the 45nm production.
WOW, it really amazes me how little performance has improved. Athlon II X2 750 (3ghz) is barely faster in most benchmarks than a Athlon X2 6400+ (3,2ghz), and loses in 1 or 2.
So the phenom core redesign buys around 300mhz around 3ghz, or only 10%. Everything else that improved in phenom is uncore.
And this while the original is at 90nm and the new one is 45nm, what a waste of potential. It seems to me AMD could've tried a little harder with the Athlon II.
There's a small util/service out there that brings phenom II cnq behaviour from vista over to windows xp (phenom II's cnq behaves like phenom I's cnq under windows xp). It does this by disabling standard cnq (set power management not on "minimal") and implementing pstate changing itself
I'm beginning to wonder whether AMD/Intel are making the same mistake
we saw last year with gfx cards, ie. too many different options. What
is the target market for the new AMD CPUs? Many retailers seem to
offer just a small selection.
Any chance you could add an i7 920 and a 6000+ to the tables please?
The former for completeness, the latter to show how the newer AMD
parts stack up against a typical older product. I'd been hoping for
a suitable replacement for the 6000+ in my ASUS board, but still nothing (no BIOS support).
Atm it looks like my next system will be an i7 920 setup (core task
is video encoding). In the past there's been lots of talk about the
higher cost of an i7 system, but looking around yesterday, I was
surprised at how small the difference has now become. The i7 920 is
only 18% more than the Ph2 955 BE. Expecting a larger difference for
the mbd cost, I found an X58 board from Gigabyte (the GA-EX58-UD3R,
135 UKP from LambdaTek) right in the middle of the price range of
typical AM3 boards. They both use DDR3, so that isn't a factor. I
was going to build a Ph2 955 BE system for my brother as his next
gaming rig, but with such small price differences now in play, the
i7 looks more sensible.
Oh, a typo on the first page perhaps? Surely it should be 2 cores for
the Athlon64 X2?
Judging by other reviews your choice of using the x264 HD Bench Pass 1 for the Power Consumption comparison doesn’t give a true representation of the situation.
As expected other reviews shows the Athlon II X2 having a noticeably lower power consumption under typical loads.
The Phenom II X2 has too much extra circuitry to have lower power consumption and I’m surprised that you didn’t deduce that something was amiss.
As I mentioned on the power consumption page, I'm guessing it has more to do with the current level of BIOS support for the Athlon II's power management. AMD is expecting a much better situation in the coming weeks.
can you include linux kernel compilation tests, or something similar or larger (gcc, libqt, X) ??? would help me much more than gaming and 3d rendering benches :-)
Anand, I BEG you to include some kind of compilation tests in the "bench" application; some of us are actually programmers that spend more time building than watching or transcoding movies ;)
A Linux Kernel bench + some kind of MS Visual C++ benchmark would be extremely welcome.
Btw, when could we expect the old CPUs to be added to Bench? I am specifically waiting for Athlon XP and P3/P4's.
10x
I really do want to include a software build test, the question is what is the simplest to setup and run, most representative and most repeatable test I can run?
I'd prefer something under Windows because it means one less OS/image change (which matters if you're trying to run something on ~70 different configurations) but I'm open to all suggestions.
Thoughts? Feel free to take this conversation offline over email if you'd like to help.
You could try building a CGAL demo program (http://www.cgal.org/FAQ.html)">http://www.cgal.org/FAQ.html). It is cross platform and big enough (but not too big).
I am really a Linux programmer but I could try to help if you are not a programmer. I haven't booted Windows for years but, hey, we have virtual machines nowadays :)
I have no experience with VS 2008. Can it be manualy set to certain amount of compile threads ? make has a command line parameter for this, so you can even test a single threaded compile and scale the number of threads used to exploit the drop off limit (where more threads do not yield better performance).
qt is quite huge, but that's ok, since a compilation of a few minutes (linux kernel) won't tell much in the future, when processing power increases.
Anand, the x264 first pass encoding graph is "higher is better" but the processors are arranged with the shortest bar (slowest processor) on the top. Please fix that.
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.
55 Comments
Back to Article
7Enigma - Thursday, June 4, 2009 - link
Hello Anand,In the article you seem to imply a tremendous difference in that 2meg to 3meg jump but your single data point doesn't show that conclusion.
In your example there is a ~9% increase in performance when going from 2 to 3 megs, but there is also a 4.5% increase in clockspeed between the 2 chips. So at best (in this example) you have a 5% difference, which while significant, I would't say it was tremendous.
Can you comment on whether you did any further tests with identical frequencies?
FlameDeer - Thursday, June 4, 2009 - link
Here are some corrections of AnandTech Bench.Refer to here:
http://www.anandtech.com/bench/default.aspx?b=25&a...">Adobe Photoshop CS4 results of few CPUs from AnandTech Bench
1. AMD Athlon II X2 250 results missing a dot, becoming slower than Intel Atom 230.
2. Better using "Intel Pentium Dual Core E5300" for E5300 as the name Intel printed on the CPU.
3. AMD Athlon LE-1620 notes (platform side notes) should be same as AMD Athlon LE-1640.
Thanks for adding more CPUs for comparison in AnandTech Bench, doing the full test in latest software is really time consuming especially for old CPUs. Take care.
MadAd - Wednesday, June 3, 2009 - link
would have liked to see phenom (non II) compared, especially the low end x3s, users may be thinking of upgrading from those about nowFlameDeer - Thursday, June 4, 2009 - link
You can always conveniently & interactively comparing them here.http://www.anandtech.com/bench/">AnandTech Bench
Elementalism - Wednesday, June 3, 2009 - link
I never had much luck running Cool n Quiet when I used to run AMD processors. Even in games the CPU was put into a low clock mode running at half speed. Sounds like a buggy implementation to this day.And I cant say in my experience it was the Mother Boards fault either. I enabled Cool and Quiet in the bios and nothing happened. It wasnt until I installed the Cool n Quiet driver all hell broke loose.
mino - Wednesday, June 3, 2009 - link
However one has to keep in mind the bug comes from Redmond.The Win 5.x scheduler is a piece of garbage and it allways was. It is just that it shows with different symptoms:
1 CPU core => crappy multi-tasking on heavy loads
2+ CPU cores => unable to get a grasp that power management not a thing of the future ...
A no, being from 2000 is no excuse, proper OS scheduler were there 30yrs ago.
Had MS decided to do so, their Win 5.x schedules would be a proper one from the start.
As for AMD, it is a price one pays for being ahead of the market.
Remember, it took Intel 5yrs (2003-2008) to get comparable power management capabilities as AMD with Nehalem.
mino - Wednesday, June 3, 2009 - link
sorry for the spelling mess..Hrel - Wednesday, June 3, 2009 - link
YAY 3D max, also that E7500 is beating the crap out of everything but AMD's top of the line. That really solitifies in my mind that picking the E8400 6 months ago for my friends gaming computer was the right choice. Stupid AMD, I REALLY want them to be competitive again so Intel is FORCED to stop overcharging for EVERYTHING.mmpalmeira - Wednesday, June 3, 2009 - link
Did you try to overclock the uncore of the AMD's CPUs?calyth - Tuesday, June 2, 2009 - link
Just like that article that you linked to with the name CnQ bug, I don't believe that a hardware manufacturer creates a bug if the OS scheduler can't seem to do the sensible thing and not bounce threads around. Although various factors in the AMD chips make the problem worse, bouncing threads across different cores on an Intel chip also has a bit of impact, since L1 and the L2 shouldn't be shared across cores on the i7 platform.So please, stop calling it a bug. The bug lies in Windows, not the chips.
TA152H - Tuesday, June 2, 2009 - link
I agree with almost everything you say, I only have a small caveat.Intel chips will suffer much less from this than AMD, since they have an inclusive cache architecture, and can readily read the information from the L3 cache. I still think AMD has an exclusive cache arrangement, which I really think they should change with regards to the L3 cache for reasons just like the one you mention.
For what's it's worth, Microsoft screwed Intel 14 years back when the Pentium Pro was released. Naturally, Intel got the blame for having miserable 16-bit performance (it was related to segmentation, which was part of all the 16-bit modes, and technical, even 32-bit mode even though it was transparent), because Microsoft told them the world will be 32-bit by then. Of course Windows 95 had a lot of legacy code, and Windows NT, which we called "Not There" at the time, was about as common as a 20 year old virgin in western Europe. So, Intel took the blame, just like AMD is now, despite, once again, Microsoft's incompetence.
Really, if you think about it, the ability to clock the processors differently could be a very useful features, except for the fact Winblows can't use it properly.
Good
TA152H - Tuesday, June 2, 2009 - link
First, I like seeing the Pentium 4s in the benchmarks, it was kind of interesting. They did better than I thought they would, and it makes me even more curious what they would be like on 45nm, since their clock speeds would probably be astronomical (since 45nm has much better power characteristics, and the clock speed limiter on the Pentium 4 was power use/heat).But, anyway, why not use the Pentium 4 670 (3.8 GHz), or Pentium EE 965 Extreme Edition (3.73 GHz) processors? Why use the next to fastest ones?
Don't get me wrong, it was still informative, but I would have liked to see the fastest measured against today's processors, not one step removed. Even so, it was nice to see them, so it's just a minor complaint. I'm looking forward to seeing the Nano.
strikeback03 - Wednesday, June 3, 2009 - link
Might not have had any around. Figure the "best of breed" were the most likely to be either sold or go in a system for some family member when they were no longer needed for comparison duties.ShangoY - Tuesday, June 2, 2009 - link
I am curious as to why the current cheapest Intel quad core were not included in the benches yet you bothered to go grab the previously $999 Pentium 4 and then also included the Phenom X4 940.Gary Key - Tuesday, June 2, 2009 - link
http://www.anandtech.com/bench/default.aspx?b=2">http://www.anandtech.com/bench/default.aspx?b=2 - You can compare them here.Kenzid - Tuesday, June 2, 2009 - link
Does any body know why AMD transistor density is very low compare to Intel? Is this because of Intel High K metal process or the architechture?Goty - Tuesday, June 2, 2009 - link
It's more than likely due to the fact that Intel has much higher cache densities than AMD does. It probably had very little to nothing to do with the actual process (well, beyond the geometry size, that is).TA152H - Tuesday, June 2, 2009 - link
What are you basing that on?Typically, cache is very dense, so you will notice transistor count increasing disproportionately to size as you add cache.
With respect to the Athlon II X2 being larger than the Penryn, that's not really a bad thing, since it does more too; the Penryn needs a memory controller on the chipset that the Athlon II does not.
Kenzid - Tuesday, June 2, 2009 - link
Based on above die size chart. Core2Duo 107mm2 and 410million transistors while Athlon II has only 234million transistors on 117mm2. It's almost half of the number of Intel used on theirs. IMC take that much space?TA152H - Tuesday, June 2, 2009 - link
Well, take a look at the Pentium version, and you'll see the what I was saying about the cache. We both can agree it's the same core, but one has a larger cache.The Pentium is 82 mm2, with 228 million transistors, with a 2.064 megabytes of cache (L1 + L2). But, since 1 megabyte is disabled, it's really 3.064, like the other Wolfdale's have. The 6 MB version of the Core 2 is 107 mm2 with 410 million transistors.
So, you can see that adding 3 MB of cache increased the transistor count by 182 million, but the size by only 25 mm2. Or, in other words, it increased transistors by about 80%, but size by about 31%. So, cache does increase transistor count disproportionately to die size.
Oh, and yes, the IMC is quite large. You can view some of the pictures of the CPU die to see it, but it's far from insignificant in size.
vajm1234 - Tuesday, June 2, 2009 - link
its really hard to believe ad a lil concern for upcoming c2d processors --- why that 6300 sample didnt overclock? wat was wrong 3.7 i xpct normally but i m amazed as its a 45nm.... not 90 or 65crimson117 - Tuesday, June 2, 2009 - link
I'm getting image not found for this first image on the last page of the review:http://www.anandtech.com/reviews/cpu/amd/athlon2/l...">http://www.anandtech.com/reviews/cpu/amd/athlon2/l...
ShawnD1 - Tuesday, June 2, 2009 - link
I like how you included the overclockability at stock voltage. Since all of my computers use stock heatsinks, overclocking at stock voltage is pretty much all I can do.Gary Key - Tuesday, June 2, 2009 - link
I used the standard heatsink off the Phenom II 940 on the X2 250, 1.4V, and hit 3.825GHz in Win7 64, idle temps around 33C, full load at 54C. I have retail CPUs coming with the new cost reduced heatsink to see what happens but any stock AM2+ heatsink from a Phenom/PhenomII works wonderfully for overclocking these two CPUs.RamarC - Tuesday, June 2, 2009 - link
I know AMD has 'nudged' mobo makers to prevent this, but some still allow it. I'm curious why Anand didn't experiment with it.AnnonymousCoward - Wednesday, June 3, 2009 - link
I figured those extra cores have defects.Anand Lal Shimpi - Tuesday, June 2, 2009 - link
It's more of a timing issue than anything else. I spent a lot of last week working on Lynnfield and there's one more CPU review before this week is up. I am curious about it and will look into it shortly though :)Take care,
Anand
flipmode - Tuesday, June 2, 2009 - link
Hi Anand. Great review, it is much appreciated. The first thing I looked for was a comparison to the Brisbane CPU. To me, a comparison to Brisbane and a comparison to Penryn are the two most interesting.Just thoughts.
ltcommanderdata - Tuesday, June 2, 2009 - link
I'm guessing if Phenom II X2s use harvested cores, they'll eventually be a limited commodity. Still, the Athlon II X2 seems to have a lot of scaling room left so higher clocked Athlon II X2s will probably do just as well.I just flipped through the benchmarks this morning, but since I was one of those commenting on wanting to see the best of dual core Netburst processors like the 965EE for comparison, I wanted to thank you for deciding to include the 955EE.
ltcommanderdata - Tuesday, June 2, 2009 - link
As a follow-up is AMD still making 90nm dual cores or have they converted the fab to 45nm now? I believe the previous fastest dual core was still the 3.2GHz 90nm Athlon 64 X2 6400+. It's too bad it wasn't included here.Gary Key - Tuesday, June 2, 2009 - link
The X2 6400+ is in the charts now and you can always use our Bench tool to compare a whole litany of processors against each other. AMD is currently phasing out of 90nm production and even several 65nm products will be phased out this year as they ramp the 45nm production.Spoelie - Thursday, June 4, 2009 - link
WOW, it really amazes me how little performance has improved. Athlon II X2 750 (3ghz) is barely faster in most benchmarks than a Athlon X2 6400+ (3,2ghz), and loses in 1 or 2.So the phenom core redesign buys around 300mhz around 3ghz, or only 10%. Everything else that improved in phenom is uncore.
And this while the original is at 90nm and the new one is 45nm, what a waste of potential. It seems to me AMD could've tried a little harder with the Athlon II.
Spoelie - Tuesday, June 2, 2009 - link
There's a small util/service out there that brings phenom II cnq behaviour from vista over to windows xp (phenom II's cnq behaves like phenom I's cnq under windows xp). It does this by disabling standard cnq (set power management not on "minimal") and implementing pstate changing itselfhttp://home.comcast.net/~pmc650/site/?/page/CNQ_Ph...">http://home.comcast.net/~pmc650/site/?/page/CNQ_Ph...
Maybe it can do the same for the Athlon II X2 on vista...
mohindar - Tuesday, June 2, 2009 - link
Hello Anand,It will be very nice to provide some benches regarding desktop virtualization, like how windows-xp usage on this chip and so on...
plonk420 - Tuesday, June 2, 2009 - link
i'm curious to see this on the bench!mapesdhs - Tuesday, June 2, 2009 - link
I'm beginning to wonder whether AMD/Intel are making the same mistake
we saw last year with gfx cards, ie. too many different options. What
is the target market for the new AMD CPUs? Many retailers seem to
offer just a small selection.
Any chance you could add an i7 920 and a 6000+ to the tables please?
The former for completeness, the latter to show how the newer AMD
parts stack up against a typical older product. I'd been hoping for
a suitable replacement for the 6000+ in my ASUS board, but still nothing (no BIOS support).
Atm it looks like my next system will be an i7 920 setup (core task
is video encoding). In the past there's been lots of talk about the
higher cost of an i7 system, but looking around yesterday, I was
surprised at how small the difference has now become. The i7 920 is
only 18% more than the Ph2 955 BE. Expecting a larger difference for
the mbd cost, I found an X58 board from Gigabyte (the GA-EX58-UD3R,
135 UKP from LambdaTek) right in the middle of the price range of
typical AM3 boards. They both use DDR3, so that isn't a factor. I
was going to build a Ph2 955 BE system for my brother as his next
gaming rig, but with such small price differences now in play, the
i7 looks more sensible.
Oh, a typo on the first page perhaps? Surely it should be 2 cores for
the Athlon64 X2?
Ian.
smilingcrow - Tuesday, June 2, 2009 - link
Judging by other reviews your choice of using the x264 HD Bench Pass 1 for the Power Consumption comparison doesn’t give a true representation of the situation.As expected other reviews shows the Athlon II X2 having a noticeably lower power consumption under typical loads.
The Phenom II X2 has too much extra circuitry to have lower power consumption and I’m surprised that you didn’t deduce that something was amiss.
Anand Lal Shimpi - Tuesday, June 2, 2009 - link
As I mentioned on the power consumption page, I'm guessing it has more to do with the current level of BIOS support for the Athlon II's power management. AMD is expecting a much better situation in the coming weeks.Take care,
Anand
Eeqmcsq - Tuesday, June 2, 2009 - link
The one comparing various Athlon X2 specs. The table says the Athlon 64 X2 has 4 cores.Anand Lal Shimpi - Tuesday, June 2, 2009 - link
Woops, thank you :)-A
haplo602 - Tuesday, June 2, 2009 - link
can you include linux kernel compilation tests, or something similar or larger (gcc, libqt, X) ??? would help me much more than gaming and 3d rendering benches :-)virvan - Tuesday, June 2, 2009 - link
Anand, I BEG you to include some kind of compilation tests in the "bench" application; some of us are actually programmers that spend more time building than watching or transcoding movies ;)A Linux Kernel bench + some kind of MS Visual C++ benchmark would be extremely welcome.
Btw, when could we expect the old CPUs to be added to Bench? I am specifically waiting for Athlon XP and P3/P4's.
10x
Anand Lal Shimpi - Tuesday, June 2, 2009 - link
I really do want to include a software build test, the question is what is the simplest to setup and run, most representative and most repeatable test I can run?I'd prefer something under Windows because it means one less OS/image change (which matters if you're trying to run something on ~70 different configurations) but I'm open to all suggestions.
Thoughts? Feel free to take this conversation offline over email if you'd like to help.
Take care,
Anand
virvan - Wednesday, June 3, 2009 - link
You could try building a CGAL demo program (http://www.cgal.org/FAQ.html)">http://www.cgal.org/FAQ.html). It is cross platform and big enough (but not too big).I am really a Linux programmer but I could try to help if you are not a programmer. I haven't booted Windows for years but, hey, we have virtual machines nowadays :)
adiposity - Tuesday, June 2, 2009 - link
A fairly decent size build that I do is Qt under VS 2008.Instructions are here:
http://wiki.qtcentre.org/index.php?title=Qt4_with_...">http://wiki.qtcentre.org/index.php?title=Qt4_with_...
Download source here:
http://www.qtsoftware.com/downloads/windows-cpp">http://www.qtsoftware.com/downloads/windows-cpp
You can use VS2008 Express.
-Dan
haplo602 - Wednesday, June 3, 2009 - link
I have no experience with VS 2008. Can it be manualy set to certain amount of compile threads ? make has a command line parameter for this, so you can even test a single threaded compile and scale the number of threads used to exploit the drop off limit (where more threads do not yield better performance).qt is quite huge, but that's ok, since a compilation of a few minutes (linux kernel) won't tell much in the future, when processing power increases.
smitty3268 - Wednesday, June 3, 2009 - link
Yes, you can add the /MP parameter in Visual Studio.adiposity - Wednesday, June 3, 2009 - link
From the page I linked before:
Add these line to the .pro file for release version:
QMAKE_CXXFLAGS_RELEASE += -MP[processMax]
-Dan
smitty3268 - Wednesday, June 3, 2009 - link
All of Qt might be a bit large for a simple benchmark.Something like Paint.NET or NDepend might make a good C# test.
adiposity - Wednesday, June 3, 2009 - link
Use:nmake sub-src
It only compiles qt libraries, not the tools or examples.
It really does not take very long (less than 10 minutes on a Core2Duo 2.4).
-Dan
adiposity - Friday, June 5, 2009 - link
10 minutes is wrong, it's more like an hour for sub-src. My mistake.adiposity - Wednesday, June 3, 2009 - link
Oh, andconfigure -release
Will build only release target, speeds things up as well.
adiposity - Wednesday, June 3, 2009 - link
make thatconfigure -release -no-webkit
ssj4Gogeta - Tuesday, June 2, 2009 - link
Anand, the x264 first pass encoding graph is "higher is better" but the processors are arranged with the shortest bar (slowest processor) on the top. Please fix that.Ryan Smith - Tuesday, June 2, 2009 - link
Noted and fixed.