Squaring is powers of two :-)

Each drop in process size gives you approx double the number of transistors (although processes from different manufacturers have different capabilities, and different designs may use the extra transistors in different ways).

---

Sorry.

I thought powers of 2 were 2^0 (= 1 core), 2^1 (= 2 cores), 2^2 (= 4 cores), 2^3 (=8 cores)

as opposed to squaring

2 (= 2 cores), 2^2 (= 4 cores), 4^2 (=16 cores), 16^2 (=256 cores)...

Hence, my comment about: (ok, very wishful thinking, lol !)

But hey, I never got past calculus !

Squaring is powers of two :-)

True... obviously getting a bit late for me ! :-)

---

Today, there are dual and quad-core chips in existance from AMD and Intel.

--
Mats

THERE IS NOT COMMERCIAL QUAD CORE FROM AMD!
your quad core is plan for Q2 said your CEO.

Come on Matt, you are better than that!

who?

The 80-core architecture is probably a fictive device. My mum has a book from the late 50's/early 60's that I used to read when I was little. It showed the future of our lives, including space-travel and personal aircrafts, in the next couple of decades. If you look out the windows (now more than 4 decades later), you'll notice the personal aircrafts are not flying around in the towns or countryside outside... ;-) It just goes to show how hard it is to foresee how fast the development is going to be in the decades ahead....

--
Mats

WEll, I am not going to comment on this.. I am not speaking for intel, but well, I saw the silicon, it does exist, and it does..... outch, I can 't say :)
And if you think it is science fiction, it is good for me :) please keep thinking this ... you'll make my work easier.

Stop FUDing!

Maybe it's just me, but "Who?" sounds like an Intel enthusiast.

Maybe it's just me. LOL

I would have pegged him as a Cyrix fanboy, lol !

Maybe it's just me, but "Who?" sounds like an Intel enthusiast.

Maybe it's just me. LOL

I would have pegged him as a Cyrix fanboy, lol !

Maybe it's just me, but "Who?" sounds like an Intel enthusiast.

Maybe it's just me. LOL

The AMD hype is collapsing ... so, they are back to: "don't believe intel ..." , then, "our next CPU will be better ..., if not K8L because "we" can not get frequency on it, it will be Doggy K9 or K10" (yep, VP of AMD just said they will not have Mhz, at least not as much as K8 ... but it is ok, because the Instruction per clock is higher ;-))

They said to the financial analyst that they will get 20% grow in 2007, to prove it, they started by losing 578M$ ... otherwise, it is too easy ;)
they will take the money from the new york state tax payer to cover that, so , no problem!

Bottom line, they are not going to beat core 2, except on SPEC_FP_RATE, so , they prepare the world for: "you don t need so much performance anymore"

I put 100$ on AMD using SPEC_FP_RATE in their slide at K8L launch (lunch for me)

The only problem is that SPEC_FP_RATE is like running N instances of your apps, on N cores, but without optimizing your application for data locality.
It is basically forgetting the 1st 3 months of your bachelor about computer architecture ... it is like running 4 iTune or 4 3DsMax ...
The applications are not program like this, if you run 4 copies of the same program, you ll get 4 different memory buffers allocated, giving a 100% fit to NUMA, while, in reality, 3DsMax allocate one (AND ONLY ONE) memory space, and all the cores and threads get the data from the 3D world database ... giving NO SPACE to NUMA (Non uniform Memory architecure). A good programmer will even add the tags for "read only" to the shared buffer, avoiding write combining between unshared caches. On the top of this, they are not going to speak about SPEC_INT_RATE, because they lose big time on it, and they are going to forget to tell you that server workloads are pure INTEGER based ... so, K8L compare to K8 will be only a little bit faster than K8, and it is not even sure, due to frequency lost. I don t even speak of the increased latency of the branches ... hehehehe

Bottom line, AMD is back to the "BS" (Banana Split) they are use to beliver, they are failling technically and Matt change of Attitude is in direct line with the new strategy they are unrolling.
very soon, they will say that Core 2 E serie are too cheap and it is uncompetitive practice! But sorry, it is all good for us! good CPU for cheap!

If K8L was that good, Matt should put it on Rosetta, and we shall see ... It is 8 instances of the same workload... ;) he should like it!


who? (I hope you guessed who I am working for who?)

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
This is my personal opinion, my employer is not responsable for my posting here, and I am sure they are not happy of having me speaking here free ;)

...
The applications are not program like this, if you run 4 copies of the same program, you ll get 4 different memory buffers allocated, giving a 100% fit to NUMA, while, in reality, 3DsMax allocate one (AND ONLY ONE) memory space, and all the cores and threads get the data from the 3D world database ... giving NO SPACE to NUMA (Non uniform Memory architecure).
...

From what you're saying, it sounds like Barcelona would be perfect for a heavily floating-point Boinc application then, since they run as different processes with different data areas, and hence NUMA will work well.

But this is all speculation, and 53xx isn't what Barcelona should be compared to anyway - it's the next generation of Intel quads which is the relevant comparison.

---

It has to be frustrating to work in an area like Intel/AMD or Sony/MSFT/Ninendo because of all the back-and-forth, so I will not get down on who? for letting off some steam.

Irregardless of the marketing angles that the companies take. It is clear that Intel currently has the lead with the chips that are on the market. Intel has been using 65nm process for over a year and AMD will not be releasing 65nm chips for several months still. Intel also got to market faster with their quad cores.

Though it is easy to get frustrated with the talk on the Internet, any IT server purchaser who is worth a darn will know that Intel currently has the lead (over AMD) on processing power, heat, and power usage.

---

Irregardless of the marketing angles that the companies take. It is clear that Intel currently has the lead with the chips that are on the market. Intel has been using 65nm process for over a year and AMD will not be releasing 65nm chips for several months still. Intel also got to market faster with their quad cores.

Though it is easy to get frustrated with the talk on the Internet, any IT server purchaser who is worth a darn will know that Intel currently has the lead (over AMD) on processing power, heat, and power usage.

While i agree that intel have the lead with C2D, and if I were buying I'd go for Intel over AMD for the first time in quite a few years, i'm not so sure it's clear cut that they're more energy efficient - platforms based on AMDs X2 EEs have been shown to be at least as electrically efficient as the C2Ds. Here's a good article.

Back to the thread topic though- two OC'd E6300's would match that 8 core cruncher with 1.6GHz CPUs, and the duos could be upgraded to quad cores once the prices come down, and you could probably build six or seven of them for the price an 8-core machine! If you're looking at overclocking the 1.6's then that's a different story, but even if you get them to 3.2GHz, it would still only match four OC'd E6300s. I don't think they're cost efficient yet (as crunchers).
I know which I'd go for...

---

...
The applications are not program like this, if you run 4 copies of the same program, you ll get 4 different memory buffers allocated, giving a 100% fit to NUMA, while, in reality, 3DsMax allocate one (AND ONLY ONE) memory space, and all the cores and threads get the data from the 3D world database ... giving NO SPACE to NUMA (Non uniform Memory architecure).
...

From what you're saying, it sounds like Barcelona would be perfect for a heavily floating-point Boinc application then, since they run as different processes with different data areas, and hence NUMA will work well.

But this is all speculation, and 53xx isn't what Barcelona should be compared to anyway - it's the next generation of Intel quads which is the relevant comparison.

You did not read?
threaded apps don't run separated memory banks ... I guess, you just don t want to get the point.

who?

Irregardless of the marketing angles that the companies take. It is clear that Intel currently has the lead with the chips that are on the market. Intel has been using 65nm process for over a year and AMD will not be releasing 65nm chips for several months still. Intel also got to market faster with their quad cores.

Though it is easy to get frustrated with the talk on the Internet, any IT server purchaser who is worth a darn will know that Intel currently has the lead (over AMD) on processing power, heat, and power usage.

While i agree that intel have the lead with C2D, and if I were buying I'd go for Intel over AMD for the first time in quite a few years, i'm not so sure it's clear cut that they're more energy efficient - platforms based on AMDs X2 EEs have been shown to be at least as electrically efficient as the C2Ds. Here's a good article.

Back to the thread topic though- two OC'd E6300's would match that 8 core cruncher with 1.6GHz CPUs, and the duos could be upgraded to quad cores once the prices come down, and you could probably build six or seven of them for the price an 8-core machine! If you're looking at overclocking the 1.6's then that's a different story, but even if you get them to 3.2GHz, it would still only match four OC'd E6300s. I don't think they're cost efficient yet.
I know which I'd go for...

The X2 EE uses less power that Core 2 ... AT IDLE!!!!!! (When you don t need your computer!) The delta on your electricity bill at the end of the year is around 5$ ...
White, if you run Rosetta full time, Core 2 is much more efficent, X2 EE power envellope goes to the sky if you try to get the same level of performance than the Core 2.
Very soon, we are going to hear: "AMD processor are better, because "AMD" is shorter than "Intel", and we use less ink to write it down, it is better for the environement. I am an engineer, and i am tired of hearing stupid marketing stuff and many many people repeating it without thinking ...
YES,IDLE POWER IS BETTER ON AMD WHEN YOU DONT NEED YOUR COMPUTER!!! AT IDLE! but the Pentium MMX beata it by FAR then!, so does the i386!!!!

From AMD point of view, i386 is the best CPU today!

who?
I am responsable for this posting

The X2 EE uses less power that Core 2 ... AT IDLE!!!!!! (When you don t need your computer!) The delta on your electricity bill at the end of the year is around 5$ ...
White, if you run Rosetta full time, Core 2 is much more efficent, X2 EE power envellope goes to the sky if you try to get the same level of performance than the Core 2.
Very soon, we are going to hear: "AMD processor are better, because "AMD" is shorter than "Intel", and we use less ink to write it down, it is better for the environement. I am an engineer, and i am tired of hearing stupid marketing stuff and many many people repeating it without thinking ...
YES,IDLE POWER IS BETTER ON AMD WHEN YOU DONT NEED YOUR COMPUTER!!! AT IDLE! but the Pentium MMX beata it by FAR then!, so does the i386!!!!

From AMD point of view, i386 is the best CPU today!

who?
I am responsable for this posting

According to that THG test, the X2 4600 system uses 14W less than the E6400 under load. I'm sure the E6400 is considerably quicker than the x2 for most applications, but then you could probably get a fair speed improvement from a faster x2 that uses the extra 14W. My point was that I don't think there's a huge gap in electrical efficiency when the whole platform is considered, whoever has the edge. Either way it's gotta be less environmentally unfriendly than netburst!

---

Using my P3 international "kill-a-watt" meter. (I have O'scope, amprobes, and multimeters too) My AMD64 X2 4800 toledo system consists of the following and draws the following:

AMD64 X2 4800 Toledo (939 Socket), 2x1Mb L2
WinXP home SP2, Asus A8V-VM mobo, 1G OCZ DDR400 Enhanced latency ram
Hitachi Deskstar HDT72251 6DLA380 Sata II 80G hard drive, Enermax
Noisetaker EG425P-VE SFMA 420Watt PSU, CD/DVD drive, Floppy drive, S3 Video
Reading no load load
Amps 1.1 1.73
Watts 81 134
VA 133 210
PF 0.6 0.64


My AMD64 X2 5200 Windsor data is:

AMD64 X2 5200 Windsor (AM2 Socket), 2x1Mb L2
WinXP Pro 64, Asus M2NPV-VM, 2 G (2x1G) Corsair XMS2 DDR2 800 ram,
Hitachi Deskstar HDT722525DLA380 250GB Sata II hard drive, Enermax
Noisetaker EG425P-VE SFMA 420Watt PSU, CD/DVD drive, S3 Video
Reading no load load
Amps 1.11 1.86
Watts 81 144
VA 134 223
PF 0.6 0.64


No load is defined as readings taken 5 min after startup without Boinc or other apps running 99-100 system idle process.
Load is defined as readings taken 2 minutes after starting boinc (both cores running boinc project app), and this ocurred directly after the "no load" test.

Comparisons between "seconds/decoy" of the same wu could be used to figure how they compare to actual work done.

what ya think???

dag nabbit, I just went and looked. The variations within similar wus is too varied for any reliability. Can these wu be run in standalone? We could pick a "reference" wu.??

Here's a list of the wu done by the x2 5200 and you can see the variation in time/decoy. The "decoys" column is the number of decoys done in 3 hour runtime pref.

decoys Seconds/ decoy WU

Decoys|seconds/decoy|wu name
16 667.88 1a32__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_1034_0
12 877.15 1a68__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_5141_0
7 1385.16 1acf__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_2092_0
19 565.28 1ail__BOINC_ABRELAX_frags83__1507_1022_0
18 571.44 1ail__BOINC_ABRELAX_frags83__1507_107_0
10 993.89 1aiu__BOINC_ABRELAX_frags83__1507_1553_0
29 370.70 1b3aA_BOINC_ABRELAX_frags83__1507_24_0
18 574.00 1cc8A_BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_4450_0
14 739.19 1cei__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_1978_0
19 551.93 1ctf__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_1985_0
25 419.51 1ctf__BOINC_POSE_ABRELAX_VARY_ALL_BOND_ANGLES_VARY_ALL_BOND_DISTANCES_NEWRELAXFLAGS_frags83__1450_825_0
21 495.81 1enh__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_4033_0
10 1013.70 1ew4A_BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_2355_0
9 1082.74 1ew4A_BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_4796_0
15 689.54 1fna__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_1986_0
15 709.00 1gvp__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_4716_0
15 690.06 1npsA_BOINC_ABRELAX_frags83__1507_921_0
25 427.14 1pgx__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_4238_0
31 342.79 1ptq__BOINC_ABRELAX_frags83__1507_1851_0
22 477.94 1r69__BOINC_ABRELAX_frags83__1507_1742_0
15 686.32 1ten__BOINC_ABRELAX_frags83__1507_1670_0
15 686.79 1ten__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_1994_0
15 702.58 1ten__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_5053_0
13 803.06 1tit__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_3141_0
17 616.72 1ubi__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_1976_0
16 668.67 1ughI_BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_410_0
12 872.47 1urnA_BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_2003_0
12 873.16 1urnA_BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_3384_0
14 768.60 1wit__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_1913_0
11 940.48 2acy__BOINC_ABRELAX_frags83__1507_1427_0
7 1542.09 2chf__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_1974_0
8 1348.29 2vik__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_5388_0
1 1465.03 2vik__BOINC_NOFILTERS_ABRELAX_SAVE_ALL_OUT_NEWRELAXFLAGS_frags83__1505_870_0
40 267.45 BENCH_01ABRELAX_SAVE_ALL_OUT_1hz6A_BARCODE_R51_R53_filters_1504_18_0
23 462.66 BENCH_01ABRELAX_SAVE_ALL_OUT_1ogw__BARCODE_R52_R59_filters_1504_437_0
76 140.65 DOC_1CHO_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_564_0
33 321.71 DOC_1FSS_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_593_0
33 322.96 DOC_1MAH_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_559_0
31 344.62 DOC_1MAH_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_565_0
39 277.16 DOC_1MDA_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_626_0
45 238.91 DOC_1MLC_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_559_0
42 253.31 DOC_1WQ1_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_564_0
69 155.40 DOC_2KAI_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_562_0
68 159.37 DOC_2KAI_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_584_0
44 241.97 DOC_2PCC_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_584_0
79 136.40 DOC_2PTC_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_584_0
65 166.85 DOC_2SIC_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_373_1
66 162.95 DOC_2SIC_R070115_pose_u_pert_bbmin_from_farlx_abs_tol_1495_534_0
95 113.30 DOC_2SNI_R070115_pose_u_pert_control_from_farlx_abs_tol_1496_3_0
24 440.36 s014__BOINC_LOOP_RELAX_IGNORE_THE_REST_hom001__IGNORE_THE_REST_mapback_hom009_S_00001_0008267_0_1447_41_0
20 547.86 s029__BOINC_ABRELAX_hom013__1497_3360_1
28 378.87 s029__BOINC_ABRELAX_truncate_hom001__1499_109530_0
30 354.19 s029__BOINC_ABRELAX_truncate_hom001__1499_114509_0
31 340.92 s029__BOINC_ABRELAX_truncate_hom001__1499_126770_0
34 312.27 s029__BOINC_ABRELAX_truncate_hom001__1499_138470_0
31 353.57 s029__BOINC_ABRELAX_truncate_hom001__1499_40023_0
31 343.75 s029__BOINC_ABRELAX_truncate_hom001__1499_46905_0
35 300.25 s029__BOINC_ABRELAX_truncate_hom001__1499_47470_0
37 285.38 s029__BOINC_ABRELAX_truncate_hom001__1499_73326_0
31 339.93 s029__BOINC_ABRELAX_truncate_hom001__1499_90265_0
32 332.48 s029__BOINC_ABRELAX_truncate_hom001__1499_94657_0
30 357.77 s029__BOINC_ABRELAX_truncate_hom002__1499_292_0
33 319.14 s029__BOINC_ABRELAX_truncate_hom002__1499_5610_0
30 354.05 s029__BOINC_ABRELAX_truncate_hom002__1499_6078_0
32 327.84 s029__BOINC_ABRELAX_truncate_hom002__1499_6697_0
36 294.78 s029__BOINC_ABRELAX_truncate_hom004__1499_4494_0
28 377.83 s029__BOINC_ABRELAX_truncate_hom006__1499_13452_0
31 338.05 s029__BOINC_ABRELAX_truncate_hom009__1499_273_0
32 336.94 s029__BOINC_ABRELAX_truncate_hom010__1499_7361_0
28 381.19 s029__BOINC_ABRELAX_truncate_hom011__1499_3982_0
30 361.23 s029__BOINC_ABRELAX_truncate_hom014__1499_13803_0
30 350.85 s029__BOINC_ABRELAX_truncate_hom014__1499_274_0
25 417.96 s029__BOINC_ABRELAX_truncate_hom014__1499_314_0
34 313.55 s029__BOINC_ABRELAX_truncate_hom017__1499_12534_0
29 360.05 s029__BOINC_ABRELAX_truncate_hom017__1499_1264_0
33 320.54 s029__BOINC_ABRELAX_truncate_hom017__1499_14605_0
29 363.89 s029__BOINC_ABRELAX_truncate_hom017__1499_292_0
35 303.88 s029__BOINC_ABRELAX_truncate_hom017__1499_5033_0
31 337.65 s029__BOINC_ABRELAX_truncate_hom018__1499_273_0
31 339.08 s029__BOINC_ABRELAX_truncate_hom019__1499_2098_0
33 320.60 s029__BOINC_ABRELAX_truncate_hom019__1499_7726_0
28 378.46 s029__BOINC_ABRELAX_truncate_hom021__1499_12881_0
31 348.37 s029__BOINC_ABRELAX_truncate_hom024__1499_14247_0
32 342.10 s029__BOINC_ABRELAX_truncate_hom024__1499_1627_0
36 296.32 s029__BOINC_ABRELAX_truncate_hom024__1499_762_0

Astro:
When this topic was brought up in the past, people that watched a WU on their system pointed out that each decoy was taking different amounts of time. So you can run a WU for 24 hours and give an average time for a WU with short running decoys - but they don't run in a set amount of time.

---

I think you're right. If standalone with a reference can't be done, the best would be to average over a LARGE sample. Since, we can't be guaranteed to run the same WU many times and then take that average. Maybe, running ONE or several at 24 hours might provide an idea of what it compares like. I resist changing my "run time" prefs, based solely on the fact that ALL the data I display is based upon the "not selected" or 3 hour run time prefs for comparability reasons. (I.E. I'd have to remember to exclude them from my charts).

And Yes, I know Who?s' rig will blow me out of the water for total production. I concede this point in advance, and bow to his greatness. I am curious if ONE $5000 system will outproduce 8 AMD 64 X2 5200's though(what I could buy for that much cash), even taking increased electrical usage into account.

The X2 EE uses less power that Core 2 ... AT IDLE!!!!!! (When you don t need your computer!) The delta on your electricity bill at the end of the year is around 5$ ...
White, if you run Rosetta full time, Core 2 is much more efficent, X2 EE power envellope goes to the sky if you try to get the same level of performance than the Core 2.
Very soon, we are going to hear: "AMD processor are better, because "AMD" is shorter than "Intel", and we use less ink to write it down, it is better for the environement. I am an engineer, and i am tired of hearing stupid marketing stuff and many many people repeating it without thinking ...
YES,IDLE POWER IS BETTER ON AMD WHEN YOU DONT NEED YOUR COMPUTER!!! AT IDLE! but the Pentium MMX beata it by FAR then!, so does the i386!!!!

From AMD point of view, i386 is the best CPU today!

who?
I am responsable for this posting

According to that THG test, the X2 4600 system uses 14W less than the E6400 under load. I'm sure the E6400 is considerably quicker than the x2 for most applications, but then you could probably get a fair speed improvement from a faster x2 that uses the extra 14W. My point was that I don't think there's a huge gap in electrical efficiency when the whole platform is considered, whoever has the edge. Either way it's gotta be less environmentally unfriendly than netburst!

you are perfecty right, the difference of few watts is not a big deal ... a i386 even use less power ... the processing is very much better with Core 2 and I don't think anybody will contest this today,and at least up to August.
In august, I expect to have a bigger smile ;)

And Yes, I know Who?s' rig will blow me out of the water for total production. I concede this point in advance, and bow to his greatness. I am curious if ONE $5000 system will outproduce 8 AMD 64 X2 5200's though(what I could buy for that much cash), even taking increased electrical usage into account.

well, your 8 Honda civics may transport more people than my Lambo ... but i still have the bragging righ ;)


who?

so that's a "no" to the data comparison with actual facts?

I really am curious. I see benchmark result of the E6600, E6700, E6800, and QX6700. I'm curious about how they really compare. I suspect my AMD64 5200 falls between the E6700 and E6800, and would like to know. I've even been looking at the Asus 1LN64 mobo with two FX74 chips, although I think they're a bit (ok alot)Pricey.

OK, I'll tell you what, you keep your computers visible and provide me with your watt VA, and operating voltage readings for a "no load" and "loaded (all 8 cores)", and I'll gather the data myself. I'll do the heavy lifting.

tony