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

[Official] AMD Bulldozer Reviews Thread - Page 201

post #2001 of 2308
Quote:

Not really.

This is nothing more than a terrible attempt at damage limitation, and trying to buy themselves more time.

New boards / immature bios / new platform - so must perform crap right?

I don't recall there being such issues with LGA1155 P67 boards with UEFI BIOS's.
post #2002 of 2308
I suspect there may be some truth in this:

Quote:
On paper bulldozer is a lovely chip. Bulldozer was on the drawing board (people were even working on it) even back when I was there. All I can say is that by the time you see silicon for sale, it will be a lot less impressive, both in its own terms and when compared to what Intel will be offering. (Because I have no faith AMD knows how to actually design chips anymore). I don’t really want to reveal what I know about Bulldozer from my time at AMD.

What did happen is that management decided there SHOULD BE such cross-engineering ,which meant we had to stop hand-crafting our CPU designs and switch to an SoC design style. This results in giving up a lot of performance, chip area, and efficiency. The reason DEC Alphas were always much faster than anything else is they designed each transistor by hand. Intel and AMD had always done so at least for the critical parts of the chip. That changed before I left – they started to rely on synthesis tools, automatic place and route tools, etc. I had been in charge of our design flow in the years before I left, and I had tested these tools by asking the companies who sold them to design blocks (adders, multipliers, etc.) using their tools. I let them take as long as they wanted. They always came back to me with designs that were 20% bigger, and 20% slower than our hand-crafted designs, and which suffered from electromigration and other problems.

That is now how AMD designs chips. [sarcasm]I’m sure it will turn out well for them [/sarcasm]
http://www.insideris.com/amd-spreads...ee-speaks-out/

AMD used to lead the way. From 2000 to 2007 all my systems were AMD.
Nyarlathotep
(17 items)
 
Cxaxukluth
(19 items)
 
Yog-Sothoth
(10 items)
 
CPUMotherboardGraphicsRAM
Intel Core i7 3970X ASUS X79 ROG Rampage IV Extreme Bios 2105 EVGA KingPin Classified 780 Ti 3GB Tri-SLI  G.Skill Trident X F3-2400C10Q-16GTX 16GB (... 
Hard DriveHard DriveOptical DriveCooling
OCZ Vertex 4 SSD Seagate Barracuda 1TB ST31000524AS Lite-On IHAS124 24X DVDRW 3 x XSPC Rasa 750 RX360 Universal CPU Water Coo... 
CoolingOSMonitorPower
3 x EK-FC780 GTX Classy - Nickel Windows 7 Professional x64 3 x Dell U2412M SilverStone 1500W V2.2 x 2 
Case
Lian Li Pitstop T60 Testbench 
CPUMotherboardGraphicsRAM
i7 - 5960X Asus Rampage Extreme V Gigabyte Titan X 12GB G.SKILL Ripjaws 4 series 16GB (4 x 4GB) 288-Pin... 
Hard DriveHard DriveHard DriveOptical Drive
Samsung 850 Pro Series 1TB SSD 3 x Seagate 2TB  Seagate 3TB LG 
CoolingCoolingOSMonitor
Phobya Xtreme 1260 Radiator XSPC D5 Photon 170 Reservoir/Pump Combo Microsoft Windows 7 Professional 64bit with SP1  Samsung S34E790C 34in Ultrawide Curved LED Mo... 
KeyboardPowerCaseMouse
Microsoft Wireless 3000 Silverstone ST1500 Antec Eleven Hundred Microsoft Wireless 3000 
AudioAudioAudio
Audioengine D1 Premium 24-Bit DAC Audioengine 5+ Premium Powered Speakers Black Audioengine S8 Black 
CPUMotherboardGraphicsRAM
i7 3960X Asus Rampage IV Extreme Gainward "GOOD" GTX 580 x 2 AData x 2 
Hard DriveCoolingCoolingOS
Seagate XSPC Rasa EX 360 crossflow kit x 2, XSPC Rasa E... EK GW-580 Waterblocks Win 7 Pro 
PowerCase
SilverStone Strider ST1500 InWin D Frame 
  hide details  
Reply
Nyarlathotep
(17 items)
 
Cxaxukluth
(19 items)
 
Yog-Sothoth
(10 items)
 
CPUMotherboardGraphicsRAM
Intel Core i7 3970X ASUS X79 ROG Rampage IV Extreme Bios 2105 EVGA KingPin Classified 780 Ti 3GB Tri-SLI  G.Skill Trident X F3-2400C10Q-16GTX 16GB (... 
Hard DriveHard DriveOptical DriveCooling
OCZ Vertex 4 SSD Seagate Barracuda 1TB ST31000524AS Lite-On IHAS124 24X DVDRW 3 x XSPC Rasa 750 RX360 Universal CPU Water Coo... 
CoolingOSMonitorPower
3 x EK-FC780 GTX Classy - Nickel Windows 7 Professional x64 3 x Dell U2412M SilverStone 1500W V2.2 x 2 
Case
Lian Li Pitstop T60 Testbench 
CPUMotherboardGraphicsRAM
i7 - 5960X Asus Rampage Extreme V Gigabyte Titan X 12GB G.SKILL Ripjaws 4 series 16GB (4 x 4GB) 288-Pin... 
Hard DriveHard DriveHard DriveOptical Drive
Samsung 850 Pro Series 1TB SSD 3 x Seagate 2TB  Seagate 3TB LG 
CoolingCoolingOSMonitor
Phobya Xtreme 1260 Radiator XSPC D5 Photon 170 Reservoir/Pump Combo Microsoft Windows 7 Professional 64bit with SP1  Samsung S34E790C 34in Ultrawide Curved LED Mo... 
KeyboardPowerCaseMouse
Microsoft Wireless 3000 Silverstone ST1500 Antec Eleven Hundred Microsoft Wireless 3000 
AudioAudioAudio
Audioengine D1 Premium 24-Bit DAC Audioengine 5+ Premium Powered Speakers Black Audioengine S8 Black 
CPUMotherboardGraphicsRAM
i7 3960X Asus Rampage IV Extreme Gainward "GOOD" GTX 580 x 2 AData x 2 
Hard DriveCoolingCoolingOS
Seagate XSPC Rasa EX 360 crossflow kit x 2, XSPC Rasa E... EK GW-580 Waterblocks Win 7 Pro 
PowerCase
SilverStone Strider ST1500 InWin D Frame 
  hide details  
Reply
post #2003 of 2308
Quote:
Originally Posted by microfister View Post
since when is blender a benchmarking tool? what did they render with it?
If you render the same scene with any 3D application, they're benchmarking tools.

Probably better ones in fact than the synthetic ones because they'll use the memory controllers and the RAM in general a lot more, you get a better idea of the system's power and how well the CPU accesses data in addition to simply how well it processes it.
2P Workstation
(13 items)
 
  
CPUCPUMotherboardGraphics
Xeon E5 2699 v3 Xeon E5 2699 v3 ASUS Z10PE-D16 EVGA TITAN X 
RAMHard DriveHard DriveCooling
128GB Crucial DDR4 @ 2133MHz Intel 750 Series 1.2TB 4 x 4TB Constellation RAID5 2 x XSPC Raystorms 
CoolingCoolingMonitorMonitor
RS480 MCP655 Dell U3011 Dell U3011 
Case
Silverstone TJ07 
  hide details  
Reply
2P Workstation
(13 items)
 
  
CPUCPUMotherboardGraphics
Xeon E5 2699 v3 Xeon E5 2699 v3 ASUS Z10PE-D16 EVGA TITAN X 
RAMHard DriveHard DriveCooling
128GB Crucial DDR4 @ 2133MHz Intel 750 Series 1.2TB 4 x 4TB Constellation RAID5 2 x XSPC Raystorms 
CoolingCoolingMonitorMonitor
RS480 MCP655 Dell U3011 Dell U3011 
Case
Silverstone TJ07 
  hide details  
Reply
post #2004 of 2308
Quote:
Originally Posted by mad0314 View Post
Its not so simple. They COULD have shrunk Phenom II and gotten better performance out of it. But the thing is, they are focused more on server chips, and this is a server based chip for desktops. Its not that their R&D can't do it, its that it would cost them a lot more. I was going to say that their budget can't do it, but I don't know that. In the end, they are focused more on server chips and this is a chip that reflects that.
mrcool63 is right, if they could have made a high-performing 4 core chip to compete with intel, they would have done it, but they couldnt.. it IS that simple.
Has nothing to do with focusing on server chips.. AMD is not dumb, I am sure even they know that achieving the same performance with lesser cores is better than having more but weak cores.

You make it sound like a die-shrunk Phenom II would be able to keep up with SB core for core, but thats laughable. I am pretty sure it wouldnt even match Nehalem IPC, so thats still MILES away from having a competitive 4 core.
Fact is that AMD has to use more cores in order to keep up because they simply cant build a faster core, its hopeless.

A die-shrunk Phenom II X8 would have had bit more performance thats right, but it would still be 8 vs 4 cores which means it would still fail in most benchmarks that dont really utilize all of them.
Edited by toX0rz - 10/14/11 at 4:48am
Rebel
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5-2500k 4,6 GHz @ 1,32 V AsRock Z68 Extreme3 Gen3 PNY GeForce GTX 660 Ti SLI @ 1228 / 3312 MHz 8GB G.Skill Ripjaws DDR3-1600 
Hard DriveOSMonitorPower
Samsung 1TB + 500GB Windows 7 Ultimate 64 Asus VH22H Xilence XQ 850W with a Scythe Kama Flex 
CaseMouse
Aerocool VS-9 Xmas Edition :P Logitech G500 
  hide details  
Reply
Rebel
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5-2500k 4,6 GHz @ 1,32 V AsRock Z68 Extreme3 Gen3 PNY GeForce GTX 660 Ti SLI @ 1228 / 3312 MHz 8GB G.Skill Ripjaws DDR3-1600 
Hard DriveOSMonitorPower
Samsung 1TB + 500GB Windows 7 Ultimate 64 Asus VH22H Xilence XQ 850W with a Scythe Kama Flex 
CaseMouse
Aerocool VS-9 Xmas Edition :P Logitech G500 
  hide details  
Reply
post #2005 of 2308
These CPUs are officially an epic fail. Prices of the 2500K/2600K and Intel SB motherboards, went up with the release of these processors. Way to go, AMD...
    
CPUMotherboardGraphicsRAM
Intel Celeron G530 Haven't decided yet. GTX 460 768MB 4GB Ripjaws 
Hard DriveOSMonitorPower
Samsung F3 1TB (SSD coming later) XP/Windows 7 Same old... Yet to decide. 
Case
Fractal Design Define R3 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel Celeron G530 Haven't decided yet. GTX 460 768MB 4GB Ripjaws 
Hard DriveOSMonitorPower
Samsung F3 1TB (SSD coming later) XP/Windows 7 Same old... Yet to decide. 
Case
Fractal Design Define R3 
  hide details  
Reply
post #2006 of 2308
Quote:
Originally Posted by mad0314 View Post
You cannot say that a Honda Civic is as fast as a Bugatti Veyron when they are only allowed to go the speed limit.
Well, you unknowingly proved his point here...bravo!

I'll FTFY:

Quote:
Originally Posted by mad0314 View Post
You can say that a Honda Civic is as good as a Bugatti Veyron when they are only allowed to go the speed limit.
post #2007 of 2308
Quote:
Originally Posted by Dublin_Gunner View Post
Not really.

This is nothing more than a terrible attempt at damage limitation, and trying to buy themselves more time.

New boards / immature bios / new platform - so must perform crap right?

I don't recall there being such issues with LGA1155 P67 boards with UEFI BIOS's.
No, not with the BIOS. But as I recall, Intel's chipset for Sandy Bridge was not without its problems.

...From Intel's own technology blog...

No one is perfect.

Like the performance evals, I will now wait for multiple sources to do comparisons. In fact, I am gonna message a reviewer I have talked to and see if he's going to look into this. It would make a great article for his website.
2009-2010 Build
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II x4 555BE @ 3.96GHz MSI 790FX-GD70 2x Sapphire 5850s 2x4GB Corsair XMS3 DDR3-1333 
Hard DriveOptical DriveOSMonitor
2x OCZ VertexII RAID0 + 2xSamsung F1s RAID0 Lite-On BD-ROM/DVD-RW WinXP Pro x64/Win 7 Pro x64/Kubuntu 9.10 x64 Acer G24 Ferrari 
KeyboardPowerCaseMouse
merc SteelSeries Enermax Evolution 1050W Azza Solano 1000 Microsoft Sidewinder X5 
  hide details  
Reply
2009-2010 Build
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II x4 555BE @ 3.96GHz MSI 790FX-GD70 2x Sapphire 5850s 2x4GB Corsair XMS3 DDR3-1333 
Hard DriveOptical DriveOSMonitor
2x OCZ VertexII RAID0 + 2xSamsung F1s RAID0 Lite-On BD-ROM/DVD-RW WinXP Pro x64/Win 7 Pro x64/Kubuntu 9.10 x64 Acer G24 Ferrari 
KeyboardPowerCaseMouse
merc SteelSeries Enermax Evolution 1050W Azza Solano 1000 Microsoft Sidewinder X5 
  hide details  
Reply
post #2008 of 2308
Quote:
Originally Posted by alancsalt View Post
I suspect there may be some truth in this:

http://www.insideris.com/amd-spreads...ee-speaks-out/

AMD used to lead the way. From 2000 to 2007 all my systems were AMD.
Computer fab processes are just a few years away from not being able to become smaller. A 2D chip (note: "3D transistors" are not 3D, just turned on edge. They are not stacked.) scales to a power of 2; in order to increase performance once minimum fab size is achieved, a 2D chip must get wider and longer. A 3d chip will allow the same base size (and keep devices small).

Computer designed chips is the future. Some research shows that there is little or no heat output with graphene chips. This makes 3D designs feasible. A cubic chip would have so many transistors and be so complicated (due to hundreds of 3D interconnects) that a computer is the only feasible method of design.

AMD designing chips with a computer (if true) simply shows that the company is forward thinking. Getting experience and patents (I don't like patents, but it is "legit" business strategy) would put AMD ahead of the curve. As experience with computer design increases, efficiency will as well (note: AMD must still create and input basic structures for a computer program to work with). Also important to note is that AMD would still examine the computer designs and tweak the less efficient places by hand (as is done in computerise engineering in every other engineering profession).

That said, GPU's have been computer designed for years and it works fairly well.

edit: I would be remiss if I did not say that Intel undoubtedly uses computers for at least the more boring jobs like designing wire trace layers. If computers are not used for current designs, Intel still likely has a team doing research into this area.

To answer some other persons, I think I'll just quote myself to save typing

Quote:
Originally Posted by hajile View Post
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 - 10/14/11 at 6:34am
post #2009 of 2308
Heh, Intel's chipset problem had nothing to do with performance and was just a messed up storage controller. My theory on all of this unsupported supposition about ASUS being the culprit is that the whole thing is an overblown wishful-thinking conspiracy theory. My reasoning:

First and foremost AMD isn't about to send every reviewer a board that doesn't use their chip to its potential. Agreement or no, that would just be ridiculously stupid to do.

Second, ASUS' BIOS was as up-to-date as it got. The date on the UEFI AMD gave us was little more than a week before the board & cpu were shipped to us. There was one less than a week old (which I flashed) on their FTP available for our use in reviews.

Third I find it hard to believe (with all due respect to the other companies) that ASRock and MSI have some sort of secret sauce to possess a significantly stronger BIOS than ASUS or Gigabyte. Not only AMD's reputation but ASUS' as well could be hurt by screwing the pooch with a UEFI that messes up (nearly) everyone's reviews.

Subset of #3: ASRock doesn't even have a production BIOS on their web site that can function with BD right now. 1.30 for the 990FX Extreme 4 does not POST with FX chips, you have to get a non-public beta from them.

Last, but certainly not least, if that were even a remote possibility why would AMD not have said anything at all to us? All of the reviewers are on the same mailing list and we have received no word what-so-ever from AMD about this. Any reasonable person would see that as a tell-tale sign that it performed right where they thought it would. At the very least AMD would reach out and say 'hey guys, anybody have an ASRock board to test this out on and make sure we didn't screw something up?'

Sorry, I just don't think that theory (and the quite-a-few others like it over the last two days) holds water. Use common sense people.
post #2010 of 2308
Quote:
Originally Posted by hokiealumnus View Post
Heh, Intel's chipset problem had nothing to do with performance and was just a messed up storage controller....
I wasn't saying it was a performance issue. Just that Intel's products aren't always perfect either.

Besides, Intel put out a solution for the issue. I'm sure if there is one with AMD's product that is resolvable, they will too.
2009-2010 Build
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II x4 555BE @ 3.96GHz MSI 790FX-GD70 2x Sapphire 5850s 2x4GB Corsair XMS3 DDR3-1333 
Hard DriveOptical DriveOSMonitor
2x OCZ VertexII RAID0 + 2xSamsung F1s RAID0 Lite-On BD-ROM/DVD-RW WinXP Pro x64/Win 7 Pro x64/Kubuntu 9.10 x64 Acer G24 Ferrari 
KeyboardPowerCaseMouse
merc SteelSeries Enermax Evolution 1050W Azza Solano 1000 Microsoft Sidewinder X5 
  hide details  
Reply
2009-2010 Build
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II x4 555BE @ 3.96GHz MSI 790FX-GD70 2x Sapphire 5850s 2x4GB Corsair XMS3 DDR3-1333 
Hard DriveOptical DriveOSMonitor
2x OCZ VertexII RAID0 + 2xSamsung F1s RAID0 Lite-On BD-ROM/DVD-RW WinXP Pro x64/Win 7 Pro x64/Kubuntu 9.10 x64 Acer G24 Ferrari 
KeyboardPowerCaseMouse
merc SteelSeries Enermax Evolution 1050W Azza Solano 1000 Microsoft Sidewinder X5 
  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