Who Says You Need Four Cores?

Status
Not open for further replies.
G

Guest

Guest
Well. It seem like virtualisation was left out as consider multi-core is critical for running virtualise application.
 
waste of time to read, its been known for years that you dont compare clock speeds (in this case, 2.53ghz) - you compare price points! Wheres an Intel Quad? or a lower end Intel like a E4600 etc? and after all that BS, why the cheap AMD board thats "$10 - $15 less" against that expensive ass intel board? pffftttt
 

The_Trutherizer

Distinguished
Jul 21, 2008
509
0
18,980
I completely don't get the point of the phenom 9850 in this review. Isn't this supposed to be a comparison of budget, workstation systems with dual core CPUs? Why put it in there? If you put a current Intel quad core in for consideration then it's power consumption would be high as well.

What exactly are you trying to prove here? In any case. Any idiot knows that currently Intel's Dual core is the ideal processor. Currently of course.

And what the hell were you thinking with the motherboard? A 740G? You even state in your conclusion that the 780G is a more fair comparison to the G45? Of course it is! Why did you even review the 740G then?

I mean what a conflicting hodge podge of an article!
 

genored

Distinguished
May 20, 2007
83
0
18,630
If you haven't bought a new computer in 6 years don't do a review about your epic fail of picking computer parts. I mean your just embarrassing....
 

rtfm

Distinguished
Feb 21, 2007
526
0
18,980
If you give a million monkeys a typwriter, one of them will write a T.H article... Seriously, most of the readers of this site are well informed, this king of waffle is no good
 

curryj02

Distinguished
Mar 28, 2008
88
0
18,640
so quad isn't worth it now... what about in six years. just as Hyper Threading has kept his P4 going so long, going quad will have the same effect. Quad doesn't scale now, but in six years? dual core will seem like single core is now - quad core = new dual core. Just my two cents
 

addiktion

Distinguished
Aug 19, 2008
11
0
18,510
I currently run a Q6600 (3GZ OC) and it has done wonders for me. Take it I do a lot of Adobe Photoshop, gaming, coding, and generally have about 20-30+ windows open at one time which I would consider my "business" & "entertainment" use.

If you add virtualization into the mix the quad core definitely has saved me. I don't experience any hiccups and now that I've migrated to 64 bit I've noticed a subtle gain in overall computing too.

I think the highest I've hit on all my cores with extensive testing is 60-70%. This was running a few browser windows + 4 scanning programs at the same time and I did get some slow down due to my hard drive read/write speeds maxing out but nothing from the CPU. Which to me leaves plenty of room for what I actually do.

Eventually when more software actually catches up to using 4 cores it'll be better utilized I suppose but for the most part I'm happy with it and I think you'll be happy with a dual or quad core.
 

fepple

Distinguished
Nov 5, 2007
9
0
18,510
Running HL2:EP2 as a benchmark is pretty silly when its only single threaded

"shocking news! new super car max speed only 30mph in residential areas"
 

KT_WASP

Distinguished
Apr 16, 2008
125
0
18,690
Wow...

So, six years ago..in 2002.. you purchased the second fastest CPU on the market (first being a P4 Northwood 3.02GHz). And this is he main reason it was able to last so long.. becuase you bought a CPU that was at the top of the performance "food chain" at that time.

Now your looking at upgrading to a PC that you hope to last another six years, but your looking at Hardware that is at the lower end of the same "food chain". How on earth do you think that will work ?

This article is useless dribble written by a person that has no buisness writing on a PC tech website....

 

iocedmyself

Distinguished
Jul 28, 2006
83
0
18,630
Seriously, i want an answer to this question Mr.Guilty editor..

HOW OUT OF TOUCH CAN ALL OF YOU AT THG POSSIBLY GET?!?

How can you possibly expect anyone to put any stock in your advice, your findings or even you basic hardware knowledge after starting the "article" out by admiting that you, someone that tests and reviews hardware for a living, is not only using a 6 year old CPU, but an Intel pentium 4...at 2.8ghz. Your cpu of choice is not only based on arguably the worst cpu arch produced in the last 15 years (possibly second only to the original itanium) but not even the best performing PoS out of the entire craptastic line. Performance is marginally better than a P3 at 1.8ghz, if at all, memory performance is laughable even by intel standards and the cpu can easily double as a hot-plate when you get a late night craving for scrambled eggs... and a nice thick steak. The fact that you have a workstation gpu paired with such a horrid cpu further baffles what kind of broken logic has been used over the years to justify continued use.

Given the massive shift in the quality of THG content, it's not a suprising confession, but still depressing in light of the fact that so many ill-informed readers trust the findings of THG articles at face value. There is really no exscuse. For ssssshhhhhhaaaaaammmmme.

Aside from the fact that you compare one of the older based AMD boards tech's against the newest considerably more expensive intel IGP board, the exscuse of "it being the cheapest of the low power/low price option is either a very thinly vieled means of manipulation, or is just "one of the cheap-er" options. The use of a 9850 phenom on a 740G board is just stupid. Aside from the painfully obvious performance issues created by sticking a $200 cpu on a $60 board (something that i can only picture being done by the most unknowledgable of people capable of building their own computer) a 780G is actually only $6 more expensive at...$66. Which gets you an IGP with DX10 capability from DX9.0b.

I can only assume that the idiotic trend of using 32bit vista as test system was continued here, though i didn't see my assumption confirmed or denied, it should be noted that 64bit Vista ultimate can actually be run on a 780G board using a single core 1.8ghz semperon and still be capable of smooth B-Ray/HD video playback. 32bit Vista only actually taking advantage of 2 gigs of ram, regardless of the 3.56 gigs that may be acknowledged by task manager makes for a much cheaper base componet cost.

Foxconn A7GM-S AM2+/AM2 AMD 780G ---------------------------------$ 66
AMD Sempron LE-1100 Sparta 1.9GHz Socket AM2 45W Single-Core------$ 25
PNY OPTIMA 2GB (2 x 1GB) DDR2 SDRAM DDR2 800 (PC2 6400)-----------$ 39
SAMSUNG Spinpoint F1 HD502lJ 500GB 7200 RPM SATA 3.0Gb------------$ 69
Total Base hardware cost------------------------------------------$199

That would be the cheapest Vista capable AMD system loosely following the system guidelines. However if you have any interest in an equally priced comparison yagainst the intel system.


G.SKILL 4GB (2 x 2GB) 240-Pin DDR2 SDRAM DDR2 1066 (PC2 8500)-----$ 89
GIGABYTE GA-MA790GP-DS4H AM2+/AM2 AMD 790GX ----------------------$135
AMD Phenom 9150E 1.8GHz Socket AM2+ 65W Quad-Core ----------------$170
SAMSUNG Spinpoint F1 HD502lJ 500GB 7200 RPM SATA 3.0Gb------------$ 69
Total base componet cost------------------------------------------$463

I list the 9150E as it's low power and aimed at bussiness systems, as this is what the build is supposed to reflect....an office desktop, i think it's fitting. Granted you could go with a regular old phenom for around the same price, could drop down to a 780g board and save $70 or so, chances are the average actual upgrade, and not new build will not require the cost of a hard-disk.

All the damaged evidence considered, is made even more pointless by your choice to go with the 780G board which you chose not to use in the review. Nor did you choose to do any virtualization benches....probably because that is yet another on the long list of things intel can't do well,,,,

I'm continually amazed, every other day i read an article on THG and actually say aloud to myself "they can't possibly do anything worse then what i just read" but you guys seem to prove me wrong at every turn.

You seem to always be able to become increasingly misinformed.
 

cangelini

Contributing Editor
Editor
Jul 4, 2008
1,878
9
19,795
[citation][nom]aneky[/nom]Well. It seem like virtualisation was left out as consider multi-core is critical for running virtualise application.[/citation]

While virtualization is indeed an important feature, it wasn't a factor in a business desktop build. Thank you for the feedback, though.
 

JPForums

Distinguished
Oct 9, 2007
104
0
18,680
Chris Angelini,
It has already been mentioned (even by you), but I feel it needs to be reiterated.
A 780G based board would have been much better for this comparison.
(It would be even better to have both the 740G and the 780G)

You can't really get a feel for the benefits of quad core if your hardware, admittedly, is bottlenecking it, but not the dual core.
Then there is the issue of the graphics performance.
If the graphics and/or video decode performance is important to you, then the ~$15 difference for the upgrade that will do it should be assumed. (Especially when comparing to a platform that you claim costs ~$110 more)

If you really wanted to answer the question of whether or not you should go with a quad core processor you should've done two things.
1) Use an AMD board that doesn't bottleneck one or the other processor.
2) Add a low end Intel quad core to the mix.

You'll likely still come to the same conclusion that the extra money for a quad core in a cheaper system like this is currently better spent on the graphics card, but you'll have data to back it up. You'll also have a better idea of what quad cores really bring to the table.

You constantly comment on how a 780G/790GX would show much better results. You also mention that a 780G only cost ~$15 more. Finally, after all of the testing, you conclude that you should disregard the article completely and you get a board based on a chipset that you mentioned, but didn't bother to test. I have to conclude that you know what you should be doing, but for some reason (Upper management, agenda, lack of time, laziness, money or lack there of, etc.) you don't do it. I encourage readers to do the same as you did. Disregard this article in its entirety and use previous knowledge to make your decisions.

If you don't find an article useful, why should we? Consider this the next time you decide to print an article that gives little or no relevant information to the reader. A few simple changes and/or additions would have made this article worth reading. I realize it takes more time, but I'm much more satisfied with an article that is late, but well thought out, than one that even the author doesn't feel is pertinent to his buying decisions.
 

cangelini

Contributing Editor
Editor
Jul 4, 2008
1,878
9
19,795
[citation][nom]apache_lives[/nom]waste of time to read, its been known for years that you dont compare clock speeds (in this case, 2.53ghz) - you compare price points! Wheres an Intel Quad? or a lower end Intel like a E4600 etc? and after all that BS, why the cheap AMD board thats "$10 - $15 less" against that expensive ass intel board? pffftttt[/citation]

As stated at the very beginning of the piece, the platforms weren't chosen to go head-to-head. Rather, I wanted to talk about an AMD platform that isn't getting much attention yet in contrast to a newer platform.

As for processor performance, the other interesting comparison would have indeed been a quad-core Intel chip--though the Core 2 Extreme QX9650 I'd normally use as a Q6600 wouldn't boot in this board =)
 

masterwhitman

Distinguished
Aug 19, 2008
14
0
18,510
[citation][nom]apache_lives[/nom]waste of time to read, its been known for years that you dont compare clock speeds (in this case, 2.53ghz) - you compare price points! Wheres an Intel Quad http://en.wikipedia.org/wiki/QUaD ? or a lower end Intel like a E4600 etc? and after all that BS, why the cheap AMD board thats "$10 - $15 less" against that expensive ass intel board? pffftttt[/citation]

I have to completely agree. A 780G board can be had for as little as $54 online. Why on earth would someone choose a lesser solution at the same price point? Unless, of course, it could have something to do w/cross-promotion of IDF or something...
 

cangelini

Contributing Editor
Editor
Jul 4, 2008
1,878
9
19,795
[citation][nom]The_Trutherizer[/nom]I completely don't get the point of the phenom 9850 in this review. Isn't this supposed to be a comparison of budget, workstation systems with dual core CPUs? Why put it in there? If you put a current Intel quad core in for consideration then it's power consumption would be high as well. What exactly are you trying to prove here? In any case. Any idiot knows that currently Intel's Dual core is the ideal processor. Currently of course.And what the hell were you thinking with the motherboard? A 740G? You even state in your conclusion that the 780G is a more fair comparison to the G45? Of course it is! Why did you even review the 740G then?I mean what a conflicting hodge podge of an article![/citation]

The 740G is in there because it represents AMD's most entry-level platform--and it, along with the Athlon X2 4850e, represents a price point well below the Intel configuration. The Phenom was included as comparison data. And as the conclusion reads, we'll have a follow-up shortly to add price and power to the frequency comparison--stay tuned!
 

cangelini

Contributing Editor
Editor
Jul 4, 2008
1,878
9
19,795
[citation][nom]curryj02[/nom]so quad isn't worth it now... what about in six years. just as Hyper Threading has kept his P4 going so long, going quad will have the same effect. Quad doesn't scale now, but in six years? dual core will seem like single core is now - quad core = new dual core. Just my two cents[/citation]

Thanks for the feedback Curry--and you bring up a good point. For those who are buying for the long-term, when the software industry is likely to have caught up, quad-core might be more worth the investment today. Cheers!
 
[citation][nom]fepple[/nom]Running HL2:EP2 as a benchmark is pretty silly when its only single threaded"shocking news! new super car max speed only 30mph in residential areas"[/citation]

Actually its a great test of CPU performance since Source is very CPU limited. If you take a older Pentium D vs a Core 2 in TF2 or HL2 you will see that the Core 2 will do better.
 

cangelini

Contributing Editor
Editor
Jul 4, 2008
1,878
9
19,795
[citation][nom]fepple[/nom]Running HL2:EP2 as a benchmark is pretty silly when its only single threaded"shocking news! new super car max speed only 30mph in residential areas"[/citation]

Ah ha, but given Valve's stated commitment to optimizing for multi-core processors, is it not interesting that EP2 still does not incorporate those enhancements?
 

cangelini

Contributing Editor
Editor
Jul 4, 2008
1,878
9
19,795
[citation][nom]DECENEU[/nom]ABSOLUTE RUBBISHWHAT A PIECE OF CRAP THIS ARTICLE WAScomparing bananas to apples, and elephants to serpentsthis is how low Toms has becomeWHERE ARE THE GOOD OLD DAYSthe days when this site had good reviews, the days when Wusy, Crashman , Jumping Jack and many other first class posters were spreading true knowledge to the masses???????????????EPIC FAILUREToms is nothing more than a failure these days[/citation]

Oy, at the very beginning of the article it's stated that the two platforms in question represent different price points entirely. The point is that you're able to build a respectable dual-core machine, hopefully saving some money today, and it's able to outperform a quad-core configuration. Think the 740G is the limiting factor in most of the benchmarks? Not true. That's why I swapped out for a 780G board to verify the results don't change unless you're running something that uses integrated graphics or Blu-ray decoding (thanks to Avivo HD)
 

cangelini

Contributing Editor
Editor
Jul 4, 2008
1,878
9
19,795
[citation][nom]jitpublisher[/nom]Nothing surprising, interesting, or useful about this. Am I missing something, was the article incomplete and posted early? Just don't get it.[/citation]

Nope! Just some of what I considered to be interesting data as I played around with some new hardware in the lab. Don't think many of us have seen or heard of AMD's 740G yet, nor has G45 received much attention.
 

cangelini

Contributing Editor
Editor
Jul 4, 2008
1,878
9
19,795
[citation][nom]KT_Wasp[/nom]Wow... So, six years ago..in 2002.. you purchased the second fastest CPU on the market (first being a P4 Northwood 3.02GHz). And this is he main reason it was able to last so long.. becuase you bought a CPU that was at the top of the performance "food chain" at that time.Now your looking at upgrading to a PC that you hope to last another six years, but your looking at Hardware that is at the lower end of the same "food chain". How on earth do you think that will work ?This article is useless dribble written by a person that has no buisness writing on a PC tech website....[/citation]

Thanks for the feedback Wasp. In fact, I certainly hope that this machine isn't going to have to last six years--when things slow down to the point of being unbearable, it's no fun at all.
 

cangelini

Contributing Editor
Editor
Jul 4, 2008
1,878
9
19,795
[citation][nom]iocedmyself[/nom]Seriously, i want an answer to this question Mr.Guilty editor..HOW OUT OF TOUCH CAN ALL OF YOU AT THG POSSIBLY GET?!?How can you possibly expect anyone to put any stock in your advice, your findings or even you basic hardware knowledge after starting the "article" out by admiting that you, someone that tests and reviews hardware for a living, is not only using a 6 year old CPU, but an Intel pentium 4...at 2.8ghz. Your cpu of choice is not only based on arguably the worst cpu arch produced in the last 15 years (possibly second only to the original itanium) but not even the best performing PoS out of the entire craptastic line. Performance is marginally better than a P3 at 1.8ghz, if at all, memory performance is laughable even by intel standards and the cpu can easily double as a hot-plate when you get a late night craving for scrambled eggs... and a nice thick steak. The fact that you have a workstation gpu paired with such a horrid cpu further baffles what kind of broken logic has been used over the years to justify continued use.Given the massive shift in the quality of THG content, it's not a suprising confession, but still depressing in light of the fact that so many ill-informed readers trust the findings of THG articles at face value. There is really no exscuse. For ssssshhhhhhaaaaaammmmme.Aside from the fact that you compare one of the older based AMD boards tech's against the newest considerably more expensive intel IGP board, the exscuse of "it being the cheapest of the low power/low price option is either a very thinly vieled means of manipulation, or is just "one of the cheap-er" options. The use of a 9850 phenom on a 740G board is just stupid. Aside from the painfully obvious performance issues created by sticking a $200 cpu on a $60 board (something that i can only picture being done by the most unknowledgable of people capable of building their own computer) a 780G is actually only $6 more expensive at...$66. Which gets you an IGP with DX10 capability from DX9.0b. I can only assume that the idiotic trend of using 32bit vista as test system was continued here, though i didn't see my assumption confirmed or denied, it should be noted that 64bit Vista ultimate can actually be run on a 780G board using a single core 1.8ghz semperon and still be capable of smooth B-Ray/HD video playback. 32bit Vista only actually taking advantage of 2 gigs of ram, regardless of the 3.56 gigs that may be acknowledged by task manager makes for a much cheaper base componet cost.Foxconn A7GM-S AM2+/AM2 AMD 780G ---------------------------------$ 66AMD Sempron LE-1100 Sparta 1.9GHz Socket AM2 45W Single-Core------$ 25PNY OPTIMA 2GB (2 x 1GB) DDR2 SDRAM DDR2 800 (PC2 6400)-----------$ 39SAMSUNG Spinpoint F1 HD502lJ 500GB 7200 RPM SATA 3.0Gb------------$ 69Total Base hardware cost------------------------------------------$199That would be the cheapest Vista capable AMD system loosely following the system guidelines. However if you have any interest in an equally priced comparison yagainst the intel system.G.SKILL 4GB (2 x 2GB) 240-Pin DDR2 SDRAM DDR2 1066 (PC2 8500)-----$ 89GIGABYTE GA-MA790GP-DS4H AM2+/AM2 AMD 790GX ----------------------$135AMD Phenom 9150E 1.8GHz Socket AM2+ 65W Quad-Core ----------------$170SAMSUNG Spinpoint F1 HD502lJ 500GB 7200 RPM SATA 3.0Gb------------$ 69Total base componet cost------------------------------------------$463I list the 9150E as it's low power and aimed at bussiness systems, as this is what the build is supposed to reflect....an office desktop, i think it's fitting. Granted you could go with a regular old phenom for around the same price, could drop down to a 780g board and save $70 or so, chances are the average actual upgrade, and not new build will not require the cost of a hard-disk.All the damaged evidence considered, is made even more pointless by your choice to go with the 780G board which you chose not to use in the review. Nor did you choose to do any virtualization benches....probably because that is yet another on the long list of things intel can't do well,,,, I'm continually amazed, every other day i read an article on THG and actually say aloud to myself "they can't possibly do anything worse then what i just read" but you guys seem to prove me wrong at every turn.You seem to always be able to become increasingly misinformed.[/citation]

Io--while I appreciate the feedback, I don't necessarily think that the decision to use a stable platform as the basis for a business workstation makes anyone out of touch. The reason it's so funny that I was running a Pentium 4 is precisely because I know that it's an old, hot architecture.

The workstation GPU was chosen because it supports four DVI outputs, and has nothing to do with the platform itself--nor does one affect the performance of the other.

Hopefully you'll check out Patrick's follow-up piece, which steps away from the academic frequency comparison (normalizing for the sake of seeing where the data lines up, not necessarily to say one is better than the other as many of the comments here assume) to address a more practical price and power consumption angle.
 
Status
Not open for further replies.