Do you own a Bulldozer CPU?

[Broheim]

hahahah no.. once MS is done with their patch an fx-6100 generally speaking sits next to the core i7 920, one bad bench doesnt cost you a fight, ESPECIALLY when the 2500k lost to a 3.5ghz 3870k in most linux benchies :whiste:
the fx-4100 then sits next t0 the x4 980 sometimes better.

realistically speaking bulldozer wouldn't be too bad once the MS patch arrives to boost everyone by 15%+ and once the prices go down. when the 4100 hits south of 100 bucks, hater you cant stop that ----
a 980 for -100 is practically a must by haha

http://www.youtube.com/watch?v=64wZO36OuPU

link to a reputable source...

 

[Broheim]

anyways, OP imma just leave this here

 

[Chiropteran]

Good point Broheim,

If your most common task in using your computer involves compiling chromium with VS 2008, FX-8150 is not the best CPU.

On the other hand, I can also cherry pick a benchmark:

Or, you could just look at a full comparison like I linked early, which shows that the 8150 is faster than the 1100T in most applications- without the extra 10-20% performance gained by the infamous patch. Once you account for the patch, 8150 isn't just clearly better, it's clearly a lot better.

In short, if you are telling someone to buy a 1100T over an 8150 you are doing them a disfavor, unless you know for a fact that person is only going to be running one of the few edge case applications where the 1100T wins.

 

[Rifter]

In short, if you are telling someone to buy a 1100T over an 8150 you are doing them a disfavor, unless you know for a fact that person is only going to be running one of the few edge case applications where the 1100T wins.

Wrong, a 1100T is a much better buy. Its cheaper by about 30-40%, and offers 80%+ of the performance of the 8150. All while working on older/cheaper AM3 motherbords and while using less power. How is that not a better buy than the 8150. The only way a 8150 would be better is if you really need that extra 20% of performance in which case the i5-2500 is a much better option than the 8150, and its cheaper too, and the power consumtion of the 2500 is so much lower than the 8150 it makes it look like a bad joke.

Edit: grammer

 

[Broheim]

Good point Broheim,

If your most common task in using your computer involves compiling chromium with VS 2008, FX-8150 is not the best CPU.

On the other hand, I can also cherry pick a benchmark:
-snip-

yeah it's not like the OP said

This machine is being used (mainly) for software development

or anything....

 

[Chiropteran]

yeah it's not like the OP said

or anything....

He also has something about linux in his sig. I doubt he is using VS.

Here are some linux compilation benchmarks.

http://www.phoronix.com/scan.php?page=article&item=amd_a8_3870k&num=3

 

[Chiropteran]

Wrong, a 1100T is a much better buy. Its cheaper by about 30-40%, and offers 80%+ of the performance of the 8150.

I don't think so.

8150 has been OOS for awhile (obviously enough people think it's a good CPU for the money), but amazon's price for it if it was in stock is $239.99 1100T at amazon is $223.99. That is a 7% difference in price. Additionally, the 1100T is out of stock all over and NEVER coming back, while the FX-8150 is going to continue to be manufactured and the price will only go down with time.

You didn't dispute my claims ( and it'd be silly to, the benchmarks I linked make it clear that 8150 > 1100T in most applications), so lets say 8150 after the patch will be 20% faster than 1100T.

7% more money for a 20% performance boost is a wonderful value.

As far as power consumption, 8150 uses less than 1100 in idle. Unless you are crunching data 24/7, overall it will use less than a 1100T system, even if it's peak power usage is higher.

I'm not really interested in comparing with 2500k or any other intel CPU, because the whole point is that I have a 990FX based motherboard with an old athlon dual core sitting around unused so this would be a drop in CPU upgrade. I don't want to deal with buying a new motherboard and possibly other parts to support a CPU that doesn't work in an AM3+ socket.

 

[Broheim]

He also has something about linux in his sig. I doubt he is using VS.

Here are some linux compilation benchmarks.

http://www.phoronix.com/scan.php?page=article&item=amd_a8_3870k&num=3

what are we supposed to do with that benchmark, there's no x6 in that comparison.

 

[Chiropteran]

what are we supposed to do with that benchmark, there's no x6 in that comparison.

Nevertheless, it's infinitely more informative than a benchmark on a program the OP doesn't use. Also, while it lacks the 1100T, it does show that FX-8150 can hold it's own and performs better than any intel CPU in it's price range, at that task.

 

[BD231]

You don't need all these colorful graphs to prove a piece of crap is a piece of crap guys. Bulldozer issues are simple, the design is to complicated for the process it's on. I was *STOKED* to get 32nm on AM3 boards, only to find out AMD added so much into their designs that they completely negated any power consumption benefits we might of had. Brand new node, brand new architecture, different thread scheduling within operating systems that AMD didn't address with OS makers, bad yields, horrible single threaded performance, ect.

It needs time to mature as there are just to many negatives with BullDozer right now to be considered a $200 dollar piece of silicone.

Anyone who buys an 8150 is literally screwing themselves out of performance they can obtain for cheaper on a much more mature process. All bulldozer is good for is running at its rated speeds, and as all these colorful graphs show you're not doing yourself any favors buying one if you already have something even remotely comparable.

Save your money for the second coming at least, I'm sure you're not hurting for performance to terribly.

 

[Broheim]

Nevertheless, it's infinitely more informative than a benchmark on a program the OP doesn't use. Also, while it lacks the 1100T, it does show that FX-8150 can hold it's own and performs better than any intel CPU in it's price range, at that task.

but the OP doesn't care about the FX-8150, he cares about the FX-4100 or the FX-6100.

and no the chart doesn't show AMD does better than intel in that price range because the chart is also lacking the 2600K which can currently be had for $279.99 at microcenter, putting it in the FX-8150's price range.

 

[Rifter]

I don't think so.

8150 has been OOS for awhile (obviously enough people think it's a good CPU for the money), but amazon's price for it if it was in stock is $239.99 1100T at amazon is $223.99. That is a 7% difference in price. Additionally, the 1100T is out of stock all over and NEVER coming back, while the FX-8150 is going to continue to be manufactured and the price will only go down with time.

You didn't dispute my claims ( and it'd be silly to, the benchmarks I linked make it clear that 8150 > 1100T in most applications), so lets say 8150 after the patch will be 20% faster than 1100T.

7% more money for a 20% performance boost is a wonderful value.

As far as power consumption, 8150 uses less than 1100 in idle. Unless you are crunching data 24/7, overall it will use less than a 1100T system, even if it's peak power usage is higher.

I'm not really interested in comparing with 2500k or any other intel CPU, because the whole point is that I have a 990FX based motherboard with an old athlon dual core sitting around unused so this would be a drop in CPU upgrade. I don't want to deal with buying a new motherboard and possibly other parts to support a CPU that doesn't work in an AM3+ socket.

Perhaps on amazon, local(to me, in vancouver) retailers have been blowing out the thubans for under $200 lately while the 8150 is $250-300 around here if you can even find one.

Either way, you are correct that you will only be able to get your hands on a thuban for a few more weeks. I dislike(alot) that AMD is discontinuing a better chip(price/performance) to force feed you the crap that is BD.

In any event unless you are upgrading from a AM3+ platform it makes zero senes to build AMD right now, and if you already are AM3+ and have a thuban its not worth the money to sidegrade to BD.

 

[Chiropteran]

Perhaps on amazon, local(to me, in vancouver) retailers have been blowing out the thubans for under $200 lately while the 8150 is $250-300 around here if you can even find one.

Sure, I'll give you that- if you gotta go out and buy a CPU right now today, the FX-8150 isn't a good deal simply because supply issues drove up the price. But if you are researching for a purchase a month or two down the line, I'd say it's pretty likely amazon will be back in stock by then at $239.99 per piece, and if anything the 1100Ts will be more expensive rather than cheaper.

but the OP doesn't care about the FX-8150, he cares about the FX-4100 or the FX-6100.

and no the chart doesn't show AMD does better than intel in that price range because the chart is also lacking the 2600K which can currently be had for $279.99 at microcenter, putting it in the FX-8150's price range.

So let me get this right, your line of reasoning is that the 1100T is > the FX-8150 for the OP because it's better at VS 2008. But when I point out that the FX8150 is actually faster in the majority of other applications, suddenly we aren't comparing the 1100T and 8150 anymore, instead we are comparing the 4100 and 6100.

Well, okay then. We can both play this game. The OP doesn't care about the 2600K, so it's price and performance is irrelevant. See how it works when you use your own sort of logic?

 

[CaptainAx]

I have one of these purely as a test machine. I run distributed.net clients on all of my machines and it performs as well as my 2500k at a similar clock. The problem is it uses about 150 more watts than the 2500k at 5.0GHz. The Corsair H100 was maxed out on RPMs which is why I was asking about temperature sensing in Linux on the Crosshair V. I cut it back to 4.8 and dropped the voltage to 1.40625 from 1.425 to get the H100 to actually not run at max.

That being said, any optimized code for that chip should make it rock. I was trying to compile Open64 and didn't get it to go. I'm interested in seeing if compiling the distributed.net client with FMA4 optimizations will improve performance.

Pat.
------------
I7-3960X @ still testing | Corsair H100 | Asus P9X79 Deluxe | 8Gig Turbulence DDR3-2200(@2133) 7-10-10-24 | 2x EVGA SC GTX 580 | Hyperx 240G Sata3 SSD | Antec TPQ-1200 | Lanboy Air Black
I7-980X @ 4.5GHz | Corsair H100 | Gigabyte GA-X58-UD7 | 3x4GB Ripjaws DDR3-1333 7-7-7-21 | EVGA SC GTX580 | OCZ Revodrive 240GB | 300GB Velocraptor | Antec TPQ-1200 | Lanboy Air Yellow
I7-2600K @ 4.6GHz| Corsair H100 | Gigabyte P67A-UD7-B3 | 2x4GB Ripjaws DDR3-1333 7-7-7-21 | EVGA SC GTX580 | OCZ Vertex 60GB | 300GB Velociraptor | Corsair AX850 | Lanboy Air Yellow
I7-2600 @ 3.8GHz | Stock | Gigabyte GA-Z68-D2H-B3 | 4x4GB Ripjaws DDR3-1333 7-7-7-21 | XFX 5970 BE | 120GB WDC | Corsair AX1200 | Coolermax Elite 341
I7-2500K @ 5.06GHz | Corsair H100 | ASRock Professional Gen3 | 2x4GB Ripjaws DDR3-1333 7-7-7-24 | 300GB Velociraptor | Corsair AX750 | Lanboy Air Blue
I7-2500T @ 2.3GHz | Stock | ASRock Professional Gen3 | 2x4GB Ripjaws DDR3-1333 7-7-7-21 | OCZ Solid3 64GB | Kingwin STR-500 Fanless | HAF912
FX-8150 @ 4.8GHz | Corsair H100| ASUS Crosshair V | 4x4GB Ripjaws DDR3-1866 | Seagate 500GB | AMD HD6450 | Corsair AX750 | Lanboy Air Black
I7-950 @ 3.06 | Stock | Gigabyte GA-X58-USB3 | 3x2GB OCZ DDR3-1333 | 300GB Velocirapter | 8TB Raid5 Sans Digital external | PNY GTX275 | Dead Kingwin STR-500

 

[jacktesterson]

I'd go 1090T right now for AMD.

 

[Maximilian]

hahahah no.. once MS is done with their patch an fx-6100 generally speaking sits next to the core i7 920, one bad bench doesnt cost you a fight, ESPECIALLY when the 2500k lost to a 3.5ghz 3870k in most linux benchies :whiste:
the fx-4100 then sits next t0 the x4 980 sometimes better.

realistically speaking bulldozer wouldn't be too bad once the MS patch arrives to boost everyone by 15%+ and once the prices go down. when the 4100 hits south of 100 bucks, hater you cant stop that ----
a 980 for -100 is practically a must by haha

http://www.youtube.com/watch?v=64wZO36OuPU

I want some of what you are smoking...

i7 920 crushes all bulldozer offerings, until bulldozer can soundly beat stars then it cant come close to nehalem, never mind sandy bridge or the soon to be released ivy bridge.

Bulldozer = POS.

 

[Chiropteran]

I want some of what you are smoking...

i7 920 crushes all bulldozer offerings, until bulldozer can soundly beat stars then it cant come close to nehalem, never mind sandy bridge or the soon to be released ivy bridge.

Bulldozer = POS.

I guess you didn't see it the first time.

http://www.phoronix.com/scan.php?page=article&item=amd_a8_3870k&num=3

i7 920 not crushing anything, more like i7 920 gets crushed.

 

[jhu]

I want some of what you are smoking...

i7 920 crushes all bulldozer offerings, until bulldozer can soundly beat stars then it cant come close to nehalem, never mind sandy bridge or the soon to be released ivy bridge.

Bulldozer = POS.

You mean most. It doesn't crush the FX-8150.

 

[996GT2]

hahahah no.. once MS is done with their patch an fx-6100 generally speaking sits next to the core i7 920, one bad bench doesnt cost you a fight, ESPECIALLY when the 2500k lost to a 3.5ghz 3870k in most linux benchies :whiste:
the fx-4100 then sits next t0 the x4 980 sometimes better.

realistically speaking bulldozer wouldn't be too bad once the MS patch arrives to boost everyone by 15%+ and once the prices go down. when the 4100 hits south of 100 bucks, hater you cant stop that ----
a 980 for -100 is practically a must by haha

http://www.youtube.com/watch?v=64wZO36OuPU

 

[Makaveli]

I guess you didn't see it the first time.

http://www.phoronix.com/scan.php?page=article&item=amd_a8_3870k&num=3

i7 920 not crushing anything, more like i7 920 gets crushed.

Those are stock numbers.

I would be interested in see these numbers with the 920 @ 3.6-4ghz where most people are running them. My money says even an overclocked 8150 won't be looking so hot then. pun intended ;-)

Also the 920 is a 2008 cpu and bulldozer is current its kinda embarrassing anyway you look at it.

And for the OP good luck this is a tough choice with the Current landscape amd has created for itself with the previous gen giving performance that is equal or slightly less than Next gen but for a ton more power consumption.

 

[Chiropteran]

Also the 920 is a 2008 cpu and bulldozer is current its kinda embarrassing anyway you look at it.

Also despite being 3 years old, the i7 920 still costs more than the FX-8150.

You said it was faster than the 8150, and we proved you wrong. Now you are trying to hide your error by bringing up irrelevant information. If you really think an OCed 920 is faster than an OCed 8150, go ahead and do those tests and let us know.

After all that work, you might be able to show that a $300 overclocked CPU can be faster than an overclocked $240 CPU.

 

[Broheim]

So let me get this right, your line of reasoning is that the 1100T is > the FX-8150 for the OP because it's better at VS 2008. But when I point out that the FX8150 is actually faster in the majority of other applications, suddenly we aren't comparing the 1100T and 8150 anymore, instead we are comparing the 4100 and 6100.

Well, okay then. We can both play this game. The OP doesn't care about the 2600K, so it's price and performance is irrelevant. See how it works when you use your own sort of logic?

sigh... I thought it would be obvious that if the "fastest" Bulldozer CPU is not as "fast" as the x6, then it's safe to assume that the "slower" Bulldozer CPUs wont be either... jesus

 

[Chiropteran]

sigh... I thought it would be obvious that if the "fastest" Bulldozer CPU is not as "fast" as the x6, then it's safe to assume that the "slower" Bulldozer CPUs wont be either... jesus

Uh, the slower bulldozer CPUs cost about half as much as the 1100T X6. Not sure why you would expect them to be faster.

Also, it should be "obvious" to you that the fastests bulldozer is indeed a lot faster than the fastest Phenom X6, as the benchmarks show. So your entire premise is flawed. Your idea of showing that it wasn't as fast was simply you cherry picking one of the edge cases where it does poorly. Nice try though.

 

[Maximilian]

Those are stock numbers.

I would be interested in see these numbers with the 920 @ 3.6-4ghz where most people are running them. My money says even an overclocked 8150 won't be looking so hot then. pun intended ;-)

Also the 920 is a 2008 cpu and bulldozer is current its kinda embarrassing anyway you look at it.

And for the OP good luck this is a tough choice with the Current landscape amd has created for itself with the previous gen giving performance that is equal or slightly less than Next gen but for a ton more power consumption.

Indeed, my old i7 920 was a pretty average overclocker and even it managed 3.6ghz at stock volts. The fx 8150 needs significantly more volts and power to reach a 1ghz overclock and it is slower clock for clock as well:

http://www.tomshardware.com/reviews/bulldozer-efficiency-overclock-undervolt,3083-19.html

Bulldozer is a huge failure.

 

[Broheim]

Uh, the slower bulldozer CPUs cost about half as much as the 1100T X6. Not sure why you would expect them to be faster.

Also, it should be "obvious" to you that the fastests bulldozer is indeed a lot faster than the fastest Phenom X6, as the benchmarks show. So your entire premise is flawed. Your idea of showing that it wasn't as fast was simply you cherry picking one of the edge cases where it does poorly. Nice try though.

yeah I'm cherry picking the only software development bench anand does... wait how can you cherry pick something when there's only one choice

erm, since when is 150 half of 180? but you can just as easily save $20 and go with the 1090T since there really isn't any difference other than you might have better luck in the silicon lottery with the 1100T.