Overclock.net › Forums › Industry News › Hardware News › [Official] AMD Bulldozer Reviews Thread
New Posts  All Forums:Forum Nav:

[Official] AMD Bulldozer Reviews Thread - Page 172

post #1711 of 2308
Quote:
Originally Posted by matty0610;15291261 
In what case would you need 3 6970s at 1920x1200 for gaming?

The case where you're using a Bulldozer chip apparently.
post #1712 of 2308
The clock speed difference would have been somewhat legit if they tested common frequencies these chips can achieve, but 5.2ghz is a good bit more than an average sb 24/7 OC.
Dunno
(17 items)
 
  
CPUMotherboardGraphicsRAM
Intel i5 3550 Asrock P67 Extreme4 Gen3 MSI HD7850 1GB Kingston Hyperx Blu 2x2GB 
RAMHard DriveHard DriveCooling
Kingston Hyperx Blu 2x4GB Samsung Spinpoint F3 1tb Intel 330 60Gb SSD Scythe Mugen 2 
OSMonitorMonitorPower
Windows 8 Pro Acer AL2416W Samsung 940n Chieftec 750w 
Case
Thermaltake V4 
  hide details  
Reply
Dunno
(17 items)
 
  
CPUMotherboardGraphicsRAM
Intel i5 3550 Asrock P67 Extreme4 Gen3 MSI HD7850 1GB Kingston Hyperx Blu 2x2GB 
RAMHard DriveHard DriveCooling
Kingston Hyperx Blu 2x4GB Samsung Spinpoint F3 1tb Intel 330 60Gb SSD Scythe Mugen 2 
OSMonitorMonitorPower
Windows 8 Pro Acer AL2416W Samsung 940n Chieftec 750w 
Case
Thermaltake V4 
  hide details  
Reply
post #1713 of 2308
Quote:
Originally Posted by lordikon;15292639 
Surely you're not comparing clock speed between two entirely different architectures as if that matters somehow.

So if I compare a Pentium 4 to a Sandy Bridge I need to compare them at the same clock speeds? It's probably better to instead test any CPUs you're comparing at both stock, and the fastest you can get them to run on air/water/etc.

If you want to go down that road then maybe we also rig the test to run both CPUs at the same TDP as well, just to be fair...

ill point this out, i was trying to make a point. YOU KNOW FOR A FACT intel people would be complaining about the fact that BD is clocked higher if it beat the 2600k in tests and say well they should overclock the 2600k then to make it fair. i can bring up hundreds of posts showing that(people complaining of higher amd clocks when amd wins in some tests) if you want. its all over overclock.net. to me it dosnt matter. i have a 2500k and all this BD release did was make me save some more $ and look at my gpu to upgrade now.
Quote:
Originally Posted by staryoshi;15292656 
AMD is facing supply-side deficiencies due to poor 32nm yields as well as ramp and manufacturing issues with GF Dresden. Demand may exceed their limited supply, but that reflects supply-side issues more so than strong demand.

ill have to go back and re-read it somewhere but i thought the number 250K was put out of the number of BD chips that have shipped to suppliers.
post #1714 of 2308
Quote:
Originally Posted by Papas;15292409 
true, never crossed my mind that a company like newegg and tigerdirect would carry limited quantities of a item that is such a huge release and has been touted about for over a year.

On another note. people talking about how BD performance cant be increased should really look at the 2500k/2600k. during initial testing for the release date, 2500k/2600k were scoring within 14 points of each other in vantage and 100 points in 3dmark 11(while getting beat by the i7 950 and i7 875k) now they are scoring hundreds points more(almost 1000 for the 2600k) and beating the i7 950 and i7 875k, so again, how can BD performance not increase due to some magical driver when intel released the same magical driver that made there cpu's perform better.

Yields are terrible. The launch was probably comprised of mostly paper.
I don't expect some magic 50% performance increase.

Linux developers seem to have had (still having?) a difficult time redesigning the scheduler to suit the new architecture. I assume that Windows developers are having similar issues. The issue talked about earlier in this thread (where disabling the second integer core in each module improves performance/clock). While this could show a problem with the decode unit but the decode unit is probably not the biggest problem. Data and instructions reach the integer cores only after being decoded. A decoder bottleneck will exist regardless of which integer core or cores are working downstream(the max no. of instructions decoded per unit of time is the max no. of instructions which can be executed per unit of time regardless of how many execution units are present); however, rearranging the chip in software (via disabling some cores) to simulate a more traditional architecture shows that the performance gains are more likely to be due to better scheduling optimization and less cache thrashing (not "getting more instructions to fewer core").

AMD's problem is that the CPU (apparently) can't effectively use all the available integer cores due to poor scheduling by the OS, large cache latencies, not enough decoders, and poor branch prediction. That is to say that the integer units (and possibly the FPU's) are being bottlenecked.

The poor scheduling can be fixed and (based off Anandtech's Windows 8 not-fully-optimized alpha test) will decrease normal power consumption (unused cores can downclock) while increase performance 10% or maybe more (some 4c/4cu benches showed >20% increases). The cache latencies were probably increased over expected numbers to increase yields due to poor 32nm performance at Globalfoundries. They will probably increase with the next stepping or two (side note: one of AMD's goals was near linear performance increase with clockspeed (something SB doesn't achieve) and getting the 30% clockspeed advantage over Deneb that was initially expected will also be a side affect of fab improvements).

Improving decoders (if necessary) and improving branch prediction require a complete reworking of the front-end of the processor. With normal development times for simple chip redesigns being a couple of years, I suspect that AMD knew months ago about the poor decode and branch prediction. This is the only explanation for how soon piledriver is being released (just a few months rather than a couple of years for a major redesign). AMD likely counted on the 30% greater stock clockspeed to carry them until the redesign was finished (notice that the 4.6Ghz overclock benches (roughly 30% faster than Deneb designs) were fairly competitive), but AMD was screwed by the bad fab (though I believe AMD to be at fault as well for shipping a faulty design).


My prediction (please don't quote me later, I am being optimistic, but in reality, I have little faith)

Between a 10-15% average OS performance increase (this seems fairly definite), better fabs giving (I guess) 20% increase in clockspeed rather than 30% (scaling almost linearly), better fabs giving nearly 80% improvement in cache latencies (to match Deneb latencies should be completely possible, cache is cache), and 10-15% IPC improvement (also seems fairly definite) due to more decode and better branch prediction, I believe that the next iteration will show more of the theoretical potential.

edit: At best, 15% from OS and 15% from redesign gives 30% IPC performance boost (making it 20% faster than Deneb and 20% slower than Sandybridge). Better cache latencies are a mixed bag; they may give less than 2% for some applications or they may give >20% for others. If clockspeed can be increased the total increases give between 35-90% increases in performance (that's a huge delta). Even with a 70-90% increase in overall performance, performance per transister would still be worse than Sandybridge.

This seems to be the only explanation for why a chip half the transistor count of bulldozer can have better performance. As the chip is currently, I couldn't recommend anyone buy one (I don't think that I could recommend one even if the OS problem went away).
Edited by hajile - 10/13/11 at 9:37am
post #1715 of 2308
All of u make me want to buy it after all biggrin.gif I will test it myself thumb.gif (fx-6100) i see there is sick demand in my country for all of them. Today was 30+ in stock fx-6100 and fx-8120 (very big not detalic reseller; not sure if its right word) after one hour 0 eek.gif . So instead of crying i will try it myself . I should order it today and will be getting it tommorow. Too bad now i must wait to monday tongue.gif. Maybe if i get one ill post some of myfindings (most important for me how much power hungry it is).
post #1716 of 2308
Quote:
Originally Posted by elito;15292481 
may i ask why reviewers ALWAYS, LIKE 99% ALWAYS bench these things on DIFFERENT CLOCKS? i mean..honestly, we all knew SB beats BD. but really..whats the point if youre gonna do 5.2ghz vs. 4.75ghz?? whats wrong with leaving both at 4.7??!?!? i dont get these retarded reviewers man. having better/stronger/faster architecture is one thing, but putting them on clocks that gives them PURE WIN is another.. ferrari with 600hp vs. porsche with 500hp. hmm...ferrari def wins..but thanks to the great differnce in the engine, its now PURE WIN.
p.s.. - they dare to say "apples to apples" well obv. u failed if you cant even have both systems configured accordingly to eachother.

they put them at their highest overclock, we are testing the speed of these processors.
post #1717 of 2308
so whats the verdict on this puppy? are we looking at a faulty schedular, poor cache latencies, and maybe lacking proper windows support for it? can we expect a fermi on this where performance would increase after an update rather then its netburst rehash?
post #1718 of 2308
Quote:
Originally Posted by Domino;15292816 
so whats the verdict on this puppy? are we looking at a faulty schedular, poor cache latencies, and maybe lacking proper windows support for it? can we expect a fermi on this where performance would increase after an update rather then its netburst rehash?

im guessing its a little bit of everything with lower performance than everyone expected. im not expecting BD to wipe the floor once everything is fixed, im expecting it to (hopefully) run along side SB.
post #1719 of 2308
Quote:
Originally Posted by Jagged_Steel;15292522 
And selling a product faster than you can make them is bad? AMD is also selling more Llanos than they can produce. AMD is also selling more 6950/70/90s than they can produce. So they have several new products that people are standing in line for, and what I hear (in this forum anyways) is that somehow this = failure for AMD which of course is completely ridiculous. I operated a manufacturing business for a decade or so, and I can tell you from experience that having higher demand than you can keep up with is a heck of a lot better position to be in than being able to crank out lots of products that are not wanted.

I agree with you here, AMD's liano and graphics cards are selling great. bulldozer is just bad news for enthusiasts.
post #1720 of 2308
Here's to hoping that the next iteration will be like Phenom --> Phenom II...
Blackie
(13 items)
 
  
CPUMotherboardGraphicsRAM
AMD Phenom II X4 970 ASUS M5A88-V EVO Radeon HD5850 1GB 4 GB DDR3 OCZ Platinum 
Hard DriveOptical DriveOSMonitor
Barracuda+Caviar Blue (1 TB) Samsung DVD burner Windows 7 Home Premium 64-bit Samsung T240HD 24" LCD (1920x1200) 
KeyboardPowerCaseMouse
Razer Lycosa Antec EA650 CM 690 Logitech MX518 and Razer Salmosa 
Mouse Pad
Steelpad 5L 
  hide details  
Reply
Blackie
(13 items)
 
  
CPUMotherboardGraphicsRAM
AMD Phenom II X4 970 ASUS M5A88-V EVO Radeon HD5850 1GB 4 GB DDR3 OCZ Platinum 
Hard DriveOptical DriveOSMonitor
Barracuda+Caviar Blue (1 TB) Samsung DVD burner Windows 7 Home Premium 64-bit Samsung T240HD 24" LCD (1920x1200) 
KeyboardPowerCaseMouse
Razer Lycosa Antec EA650 CM 690 Logitech MX518 and Razer Salmosa 
Mouse Pad
Steelpad 5L 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [Official] AMD Bulldozer Reviews Thread