[OFFICIAL] Bulldozer Reviews Thread - AnandTech Review Posted

[BlueBlazer]

Is there a bug or problem with Bulldozer? Initially when I saw that on Bittech and Hardware.fr reviews, I didn't think much of it. But then another two reviews (from Muropaketti and SweClockers.com) also reported the same problem, regardless of motherboard and BIOS settings. I know all processors have bugs (certain combination or sequences of instructions, sometimes in conjunction with certain processor states). However this is not a system error window popping up informing the user of illegal exceptions trapped. This is the dreaded BSOD. :hmm:

 

[frostedflakes]

Check out NewEgg reviews as well, could have sworn I saw one or two users on there mention stability problems with their setups.

 

[BlueBlazer]

Usually most compilers will avoid certain instruction sequences, or reorder them to avoid known processor bugs (such as padding with NOPs). There must be some sequence of instructions used in the Total War: Shogun 2 game that triggered it. Is there a errata list for Bulldozer? :hmm:

 

[NostaSeronx]

Usually most compilers will avoid certain instruction sequences, or reorder them to avoid known processor bugs (such as padding with NOPs). There must be some sequence of instructions used in the Total War: Shogun 2 game that triggered it. Is there a errata list for Bulldozer? :hmm:

Yes, and here it is unless you are being sarcastic(Making my post pointless if you are being sarcastic)

http://support.amd.com/us/Processor_TechDocs/48063.pdf

 

[BlueBlazer]

Surprisingly this game was part of AMD Gaming Evolved >>> AMD Gaming Evolved Total War Shogun 2, talk about screw-ups. Did the game developers get a Bulldozer for testing? :hmm:

 

[NostaSeronx]

Did the game developers get a Bulldozer for testing? :hmm:

The most obvious answer is usually the most correct answer

No

 

[Maximilian]

Is there a bug or problem with Bulldozer? Initially when I saw that on Bittech and Hardware.fr reviews, I didn't think much of it. But then another two reviews (from Muropaketti and SweClockers.com) also reported the same problem, regardless of motherboard and BIOS settings. I know all processors have bugs (certain combination or sequences of instructions, sometimes in conjunction with certain processor states). However this is not a system error window popping up informing the user of illegal exceptions trapped. This is the dreaded BSOD. :hmm:

IIRC that bluescreen is usually to do with not enough vcore when overclocking, could be that they havent got it 100% stable when overclocked.

 

[BlueBlazer]

IIRC that bluescreen is usually to do with not enough vcore when overclocking, could be that they havent got it 100% stable when overclocked.

That BSOD happened at stock speeds actually, not even at overclocked speeds yet. The game software have triggered a specific bug in Bulldozer. :hmm:

 

[drizek]

Or maybe they were using bad memory, or there was something wrong with the video card, or the sound card, or the PSU, or...

 

[BlueBlazer]

Or maybe they were using bad memory, or there was something wrong with the video card, or the sound card, or the PSU, or...

But, wait.... All 4 reviews having the same problem, with different motherboard models (one M5A99X EVO and the rest are Crosshair V), different graphic cards (one HD6970 and the rest are GTX580), different PSUs, etc. If there are problems with RAM then other software would have crashed as well. That's more than a coincidence! :hmm:

 

[Nemesis 1]

BB its nice that ya show the community this stuff . For guys like me and you . I think maybe its time to backoff. The public now has enough info to make their own minds up . I maybe an intel fanboy based on their position . But I don't want to be seen as a marketer for intel . the hand has been dealt the cards are exposed let the people read the cards for themselves.

Speaking of cards . A lot fun for ya . go to utude input illuminti card game invented in 1996 . Read the cards . You will have more to think about than microarch.

 

[theevilsharpie]

http://msdn.microsoft.com/en-us/library/ff557211(v=vs.85).aspx

The CLOCK_WATCHDOG_TIMEOUT bug check has a value of 0x00000101. This indicates that an expected clock interrupt on a secondary processor, in a multi-processor system, was not received within the allocated interval.

CAUSE: The specified processor is not processing interrupts. Typically, this occurs when the processor is nonresponsive or is deadlocked.

 

[Maximilian]

That BSOD happened at stock speeds actually, not even at overclocked speeds yet. The game software have triggered a specific bug in Bulldozer. :hmm:

Oh dear... this whole thing stinks of phenom I more and more each day.

 

[StrangerGuy]

Oh dear... this whole thing stinks of phenom I more and more each day.

Phenom I at least has a significant IPC increase and fabbed on a node too big that there was too little L3 cache and clocked too low.

This has no redeeming qualities to speak of.

 

[NostaSeronx]

This has no redeeming qualities to speak of.

It works well in a GPU bottlenecked area

HD 5550 and 440 GT

It is slightly faster than a sandy bridge(i7 2600K) setup

 

[3DVagabond]

Not to add insult to injury, because I, for on, was really hoping BD would perform.

http://www.pcper.com/reviews/Proces...ssor-Review-Can-Bulldozer-Unearth-AMD-Victory

Once you set them at the same speed, in this case 3.3GHz, 8 core BD is slower than 6 core Thuban. This isn't core vs. core. This is all 8 BD cores against 6 Thuban cores.

 

[Idontcare]

Not to add insult to injury, because I, for on, was really hoping BD would perform.

http://www.pcper.com/reviews/Proces...ssor-Review-Can-Bulldozer-Unearth-AMD-Victory


Once you set them at the same speed, in this case 3.3GHz, 8 core BD is slower than 6 core Thuban. This isn't core vs. core. This is all 8 BD cores against 6 Thuban cores.

It's so sad, this situation couldn't be any more Nov 2000 than it currently is.

http://techreport.com/articles.x/2784/7

Company A's current gen CPU can't beat its own last-gen CPU, let alone the fact that neither of them can beat Company B's current CPU...

Can't believe it took AMD 12 yrs to reverse-engineer Intel's "winning" strategy from the 90's.

 

[RussianSensation]

Phenom I at least has a significant IPC increase and fabbed on a node too big that there was too little L3 cache and clocked too low.

This has no redeeming qualities to speak of.

For me, the most shocking part of this entire launch has been the pricing. It's one thing to have a less power efficient CPU, a worse IPC per core CPU, a worse gaming CPU, a worse overclocking CPU, launching 10 months late, etc. But it takes a fail of epic proportions to combine all of these factors with a $60+ price increase over a 2500k and $100+ over an X6 1090T.

 

[drizek]

The FX 5800 works pretty good too. Slower than a 9700 Pro, totally destroyed by the 9800 Pro that came out a month afterwards. Used tons of power, was delayed due to manufacturing problems, was expensive, came with a fancy new cooler and was, in general, a giant disappointment.

http://www.anandtech.com/show/1062/1

 

[Vesku]

For me, the most shocking part of this entire launch has been the pricing. It's one thing to have a less power efficient CPU, a worse IPC per core CPU, a worse gaming CPU, a worse overclocking CPU, launching 10 months late, etc. But it takes a fail of epic proportions to combine all of these factors with a $60+ price increase over a 2500k and $100+ over an X6 1090T.

Yes, I'm not paying a price premium for the current FX CPUs. Although, apparently AMD is selling the FX 8 series in whatever volume they are shipping as it is sold out many places.

If B3 gets a handle on the power curve and they drop the price 10-20%, I'll probably get one. Mainly because I think my 990FX motherboard BIOS is going to prioritize FX performance over Phenom II. May just grab a x4 960T and see what Piledriver brings.

 

[frostedflakes]

MSRP for the 8150 is only $245. Sites charging $280 are just price gouging. NewEgg especially is notorious for this.

 

[NostaSeronx]

Just gonna link this but Passmark is going haywire!!!

http://img163.imageshack.us/img163/203/fx8120.png

 

[BlueBlazer]

Linux benchies for FX8150 are out >>> More AMD FX-4100, FX-8150 Bulldozer Linux Details. Bulldozer versus other CPUs >>> Bulldozer Benchmark Comparison.

 

[grimpr]

Yes, I'm not paying a price premium for the current FX CPUs. Although, apparently AMD is selling the FX 8 series in whatever volume they are shipping as it is sold out many places.

If B3 gets a handle on the power curve and they drop the price 10-20%, I'll probably get one. Mainly because I think my 990FX motherboard BIOS is going to prioritize FX performance over Phenom II. May just grab a x4 960T and see what Piledriver brings.

Looks there's indeed something brewing for Q1.

Heres a tip from Charlie@S|A.

http://semiaccurate.com/2011/09/07/bulldozer-finally-shipped-last-week/

"The first one is that there is a new stepping coming, SemiAccurate is hearing mid- to late Q1/2012 for the next rev. That rev is said to bump performance, specifically integer performance, up by quite a bit, and possibly improve clocks too. Either way, it looks like that stepping is the one to keep an eye out for. It isnʼt a Barcelona type fiasco, but it isnʼt an HD4870 launch either."

 

[Idontcare]

Just gonna link this but Passmark is going haywire!!!

http://img163.imageshack.us/img163/203/fx8120.png

Given the statistical nature of passmark's sampling methods, their graphs are useless IMO without some sort of error-bars included to give an estimate of the +/- range on the 90% confidence limits (or something comparable)...as well as an indication of the sample size from which each line is based upon.