In defense of "Bulldozer"

Page 7 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Chiropteran

Diamond Member
Nov 14, 2003
9,811
110
106
Scale starts at zero? Its never slower than AMD's previous high end CPU? (1100 thuban)

Slower wouldn't be faster, however you could have read my entire post before responding-

"It's single threaded performance is abysmal, in many situations worse than Phenom II."
 

Idontcare

Elite Member
Oct 10, 1999
21,118
58
91
Slower wouldn't be faster, however you could have read my entire post before responding-

"It's single threaded performance is abysmal, in many situations worse than Phenom II."

"0-15% faster" implies it is never slower, otherwise you say something along the lines of "anywhere from 10% slower to 15% faster".

The lower limit is relevant, as is the upper limit, otherwise you would not have bothered to have specified a lower limit in the first place.

And your comment in quotes reads to me that you think it is only slower when single threaded apps are involved...again it appear you ruling out the other side of the coin, that it is also slower when multi-threaded apps are involved.
 

Vesku

Diamond Member
Aug 25, 2005
3,743
28
86
Ah, was considering editing the word "everyone" out while I was typing the above think that there will be someone... someone who will ring that bell.

I'd just personally wait until revision 2 or 3 to start throwing 'flop' around. People gave Intel a decent amount of time to work on P4. By my recollection the Intel hating didn't get this bad until Athlon 64s showed up. Again, I partly blame AMD marketing for abusing the FX name and hyping the "eight" cores but really not that different than Intel's hyping of the GIGAHERTZ.
 

Chiropteran

Diamond Member
Nov 14, 2003
9,811
110
106
"0-15% faster" implies it is never slower, otherwise you say something along the lines of "anywhere from 10% slower to 15% faster".

The lower limit is relevant, as is the upper limit, otherwise you would not have bothered to have specified a lower limit in the first place.

And your comment in quotes reads to me that you think it is only slower when single threaded apps are involved...again it appear you ruling out the other side of the coin, that it is also slower when multi-threaded apps are involved.

I'm sorry my comment doesn't read the way you would prefer it to read, but it was completely accurate factually. While I may have omitted some information, I didn't make an outright lie, such as saying it's slower and hotter than phenom II, when reality is a different story.
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
Slower wouldn't be faster, however you could have read my entire post before responding-

"It's single threaded performance is abysmal, in many situations worse than Phenom II."

Slower wouldn't be faster? :rolleyes:

In many situations it is slower at multithreaded performance as well. Also, if you are so bent on getting the truth out there (which is usually a good thing), your 0% should actually be less than zero if you were to be completely honest.
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
I'm sorry my comment doesn't read the way you would prefer it to read, but it was completely accurate factually. While I may have omitted some information, I didn't make an outright lie, such as saying it's slower and hotter than phenom II, when reality is a different story.

No. It wasn't. Still isn't.
 

Chiropteran

Diamond Member
Nov 14, 2003
9,811
110
106
No. It wasn't. Still isn't.

So you are calling me a liar? What did I post that was a lie? Funny coming from you, who claimed that bulldozer was hotter & performed worse than Phenom 2. Do you have any proof to back up your statement, or are you going to admit that it was a lie as well? Or is it just your favored method of argument to focus on me instead of actually responding to the points I make?
 

Vesku

Diamond Member
Aug 25, 2005
3,743
28
86
The good news is that this thread shows a lack of PR to try to change the initial FX impression. Not enough screenshots of benchmarks and applications where FX shows it's transistor count off. Also, I haven't seen panning of how Intel does turbo versus FX.

At least AMD knows when to duck and cover.
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
So you are calling me a liar? What did I post that was a lie? Funny coming from you, who claimed that bulldozer was hotter & performed worse than Phenom 2. Do you have any proof to back up your statement, or are you going to admit that it was a lie as well? Or is it just your favored method of argument to focus on me instead of actually responding to the points I make?

It's just kind of hysterical watching one guy go against all the reviews. You act as if you hadn't seen a single one. But, you go with your bad self.

And @ Tri Teh Pony: That was funny.
 

Idontcare

Elite Member
Oct 10, 1999
21,118
58
91
I'm sorry my comment doesn't read the way you would prefer it to read, but it was completely accurate factually. While I may have omitted some information, I didn't make an outright lie, such as saying it's slower and hotter than phenom II, when reality is a different story.

I really have no idea what you are going on about.

If you want to lie by omission then feel free. Everyone is left to manage their credibility as they see best.

But if you wanted to omit the cases where the 8150 is slower than an 1100, why bother saying "0-15% faster"? Why not go the full spectrum of lying by omission and just said "14-15% faster" and ignore everything that was lower than 14% faster?

I took your post to be a genuine misunderstanding on your part, that perhaps you weren't looking at the same reviews I was. But now it appears you are just interested in misrepresenting the truth for the sake of making some point...the point is lost on me because your methods leave me suspicious of your motives.

But its not for me, I don't care to enter into these battles of semantics. You do what you do and never mind me, I'll be moving along now. Sorry to have bothered you. (seriously)
 

Chiropteran

Diamond Member
Nov 14, 2003
9,811
110
106
I really have no idea what you are going on about.

If you want to lie by omission then feel free. Everyone is left to manage their credibility as they see best.

But if you wanted to omit the cases where the 8150 is slower than an 1100, why bother saying "0-15% faster"?

Lying by omission? Why are you suddenly making such a big deal about this, while dozens of other posters are posting crap like "OMG phenom 2 beats bulldozer in everything" when the truth is far from that?
 

TemjinGold

Diamond Member
Dec 16, 2006
3,050
65
91
Can we end this already? Let's sum up.

BD sucks horribly.
There is no logical financial reason to buy one.
There can be plenty of fanboy and "let's keep AMD alive" reasons to buy one.
I am never buying one in a million years only to hope they fix it later.
A bunch of other people aren't going to either.
You probably will because you feel bad for AMD.
I will not change your mind and you will not change mine.

BD costs more than its predecessor and does not blow it away in performance, let alone what Intel has. That makes it an utter failure. I honestly don't care if it "sucks less than some people think" because it still sucks hard enough for me to never want one. I don't believe in supporting failure with my hard-earned money for the hope that they will be competitive in the future because anyone who thinks that works is kidding themselves. If AMD goes down and Intel charges $999 for a cpu, it will be because AMD failed to put out a competitive product, NOT because I didn't pay for their failed products.

If you are so bent on defending BD, just go buy some.
 

Obsoleet

Platinum Member
Oct 2, 2007
2,181
1
0
I wonder if the ignorant masses have seen these Bulldozer benches, just so slow! http://techreport.com/articles.x/21848/2
Not sure how blind one has to be to not admit that it's an imperfect, but more than capable CPU.

It will be very interesting what Piledriver brings for price/performance.. since the i5/i7/Bulldozer/PhenomII stuff were all overpowered for the vast majority of users.
A 256-300GB Intel SATA6 or Crucial M4 SSD, AMD 7-series and Piledriver on Windows8 sounds like one of the first worthy upgrades in a long time. Skimping on other elements to get a faster i7 is probably the worst move in 2011 and beyond.

Processors are simply overpowered and not the main attraction, Bulldozer has more than the performance needed.
 

Rifter

Lifer
Oct 9, 1999
11,522
751
126
I just ran across this useful bit of data, so I wanted to respond to this again.

http://www.tomshardware.com/reviews/fx-8150-zambezi-bulldozer-990fx,3043-22.html
and
http://www.tomshardware.com/reviews/fx-8150-zambezi-bulldozer-990fx,3043-12.html (score)



Look at that, uses the exact same power. Performs better, not by a lot, but it is faster. Pretty much the opposite of your "slower and hotter" statement.

Well i never said it ran hotter and used more power, it does but i never said it, i think you are mistaking me for bononos.

Also anands testing showed a 30w difference with the BD chip drawing more power than the X6, and i'll take anands testing over toms. But even if the power useage was the same a 2500k is still a better buy.
 

Rifter

Lifer
Oct 9, 1999
11,522
751
126
Does anyone see any "In defense of Sandy Bridge" threads? No? There's a reason for that - it didn't fail.

Oh and 100% this.

No one would be rushing to "Defend" BD if it wasnt a huge Fail.

Dont get me wrong i think it could be a ok chip for some people IF They cut the price by $60-100 per SKU. At current prices you would have to be a complete idiot or blind fanboi to purchase it over a i5.
 

Chiropteran

Diamond Member
Nov 14, 2003
9,811
110
106
Well i never said it ran hotter and used more power, it does but i never said it, i think you are mistaking me for bononos.

Also anands testing showed a 30w difference with the BD chip drawing more power than the X6, and i'll take anands testing over toms. But even if the power useage was the same a 2500k is still a better buy.

How many times does this need to be repeated? The Anand test showing BD using more power is the SAME TEST that shows the FX-8150 beating the phenom II by a similar percentage. No site anywhere has ever posted a benchmark showing the FX-8150 losing to a 1100T & using more power AT THE SAME TIME.

http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/9
8150 idle- 84.8
1100T idle- 109.4

Oh look, bulldozer uses less power.

8150 (load*)- 229
1100T- 200

But whats this?
41697.png


When FX-8150 uses 229 watts, it easily surpasses the i5-2500k, the 1100T, and is within margin of error of the i7-2600 in performance.

This is the ONLY test where Anandtech has shown bulldozer using more power than 1100T, and it is not even remotely "slower".

I'm curious what the power usage is in other situations, such as the probably common situation of 1-2 cores loaded and other cores idle... I suspect bulldozer would use less than 1100T in those cases but due to lack of data we don't know.

And I agree fully, the 2500k is a better buy in general. Weak CPU or not, there is no excuse to spread lies about bulldozer.
 
Last edited:

Rifter

Lifer
Oct 9, 1999
11,522
751
126
How many times does this need to be repeated? The Anand test showing BD using more power is the SAME TEST that shows the FX-8150 beating the phenom II by a similar percentage. No site anywhere has ever posted a benchmark showing the FX-8150 losing to a 1100T & using more power AT THE SAME TIME.

http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/9
8150 idle- 84.8
1100T idle- 109.4

Oh look, bulldozer uses less power.

8150 (load*)- 229
1100T- 200

But whats this?
41697.png


When FX-8150 uses 229 watts, it easily surpasses the i5-2500k, the 1100T, and is within margin of error of the i7-2600 in performance.

This is the ONLY test where Anandtech has shown bulldozer using more power than 1100T, and it is not even remotely "slower".

I'm curious what the power usage is in other situations, such as the probably common situation of 1-2 cores loaded and other cores idle... I suspect bulldozer would use less than 1100T in those cases but due to lack of data we don't know.

And I agree fully, the 2500k is a better buy in general. Weak CPU or not, there is no excuse to spread lies about bulldozer.

Look we can dance around this all day cherry picking benchmarks to make our point. Look at the cinebenck 11.5 multithread benches where a 1100T gets the same score as a 8150 while using 30w less power.

http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/7

The bottom line is BD was a flop, didnt beat there old best chip, and is overpriced at current prices for its performance.
 

Rifter

Lifer
Oct 9, 1999
11,522
751
126
AMD would have to be a complete idiot to cut prices when they sold out within days at the existing prices.

Just because AMD cant produce enough chips does not mean they are selling like hotcakes. It's not hard to sell out if you launch with very few CPU's to sell.
 

TemjinGold

Diamond Member
Dec 16, 2006
3,050
65
91
AMD would have to be a complete idiot to cut prices when they sold out within days at the existing prices.

Therein lies the problem with "buying it anyway to support AMD so they don't go down." What incentive does AMD have to actually throw out a good product when a crap product will be sold out anyway to people who are so fiercely loyal?
 

Chiropteran

Diamond Member
Nov 14, 2003
9,811
110
106
Look we can dance around this all day cherry picking benchmarks to make our point. Look at the cinebenck 11.5 multithread benches where a 1100T gets the same score as a 8150 while using 30w less power.



There are no power figures for that benchmark. Where do you get the idea that the 1100T uses less power? You can't just make this stuff up, it actually needs to be proven.

The only reason I picked the x264 2nd pass benchmark is because that is where the power figures are derived from.

"Multithreaded" doesn't mean it pegs all 8 cores at 100%, the cinecrap bench could well be using a fraction of the power used in the x264 2nd pass test.


Just because AMD cant produce enough chips does not mean they are selling like hotcakes. It's not hard to sell out if you launch with very few CPU's to sell.

What does that have to do with what I said? Irrelevant info is irrelevant. Fact is, if you are selling product faster than you can produce it, you should either increase the price or at least hold the price, it makes no sense at all to reduce price in this situation. The cause of the situation is irrelevant.
 

Rifter

Lifer
Oct 9, 1999
11,522
751
126
There are no power figures for that benchmark. Where do you get the idea that the 1100T uses less power? You can't just make this stuff up, it actually needs to be proven.

The only reason I picked the x264 2nd pass benchmark is because that is where the power figures are derived from.

"Multithreaded" doesn't mean it pegs all 8 cores at 100%, the cinecrap bench could well be using a fraction of the power used in the x264 2nd pass test.




What does that have to do with what I said? Irrelevant info is irrelevant. Fact is, if you are selling product faster than you can produce it, you should either increase the price or at least hold the price, it makes no sense at all to reduce price in this situation. The cause of the situation is irrelevant.

Have you ran cinebench 11.5? i have and it loaded all my cores to 100% both on my i7 and my PH II.