Thursday, October 13th 2011

Bulldozer Aims For 50% Improvement By 2014: Is This Really Enough To Counter Intel?

Oct 13th, 2011 19:14 Discuss (132 Comments)

The reviews are now out for AMD's brand new Bulldozer architecture, in the form of the Zambezi FX 8120 & FX 8150 processors and they don't paint a pretty picture of these flagship products. The chips use lots of power, run hot and significantly underperform compared to their Intel competition. On top of that, they are being marketed as 8 core processors, when they are actually 4 core with an advanced form of multi-threading, due to the siamesed nature of each dual processor module. Perhaps to counter this negative publicity and try to restore some faith in the AMD brand, they have released a roadmap for the planned improvements to the architecture, all the way to 2014 - an ambitious timeline, given how much and how unexpectedly things can change at the cutting edge of the technology world.

Looking at the chart, one can see that the various architectures Piledriver, Steamroller and Excavator all add up to between 30-50% projected improvement by 2014 (subject to change without notice, of course). These are all names designed to impart a tough-guy image to their products to give one the impression that they must perform very well, beating the competition into submission. Therefore, if they fail to perform competitively against Intel, those names will continue to be branding embarrassments like Bulldozer is, currently. As Intel is already 20-50% faster right now depending on the benchmark, how are these modest improvements possibly going to compete with Intel's future products? AMD has already had a change of management at the top recently, so we can only hope that the right CEO comes along and turns them around, otherwise they may end up not manufacturing x86 processors at all in future, possibly becoming a GPU company only.

The main problem with the current Bulldozer architecture is that it's very, very late to market. AMD started working on it four years ago in 2007, which is a very long time in the world of desktop processors, so AMD have effectively released a new "old" product. The two important things that it has going for it, are that it scales well with core count and clock speed - those 8GHz overclock marketing demos weren't completely without merit. What we need to see is AMD improving performance much more than the prediction slide they've released, more like 100% or more perhaps, which is not really such an unrealistic target to achieve in three years of design and process improvements. Perhaps discarding this whole architecture and starting afresh with fully discreet cores like on the Phenom might be the way forward? AMD has recently let go some of its top-level management, so perhaps their replacements can turn the company around?

So, even if AMD achieves this projected performance improvement and more, will it really be enough to counter Intel, or will Intel steamroller AMD's Bulldozer back into submission?Source:X-bit labs and Bulldozer block diagram courtesy of Hexus' FX 8150 review.

Add your own comment

132 Comments on Bulldozer Aims For 50% Improvement By 2014: Is This Really Enough To Counter Intel?

#51

lashton

2 things, why don't AMD add a massive amount like 20MB of level 3 cache and up the curre4nt Level 1 and 2 cache on their processors, sure decrease some overclock head room and increase a little TDP but it would add a SNOT load of performance, i think AMD is too stuck in the low power usage analogy, also Intel in 2014 will run out their core i line and will need to release another CPU architecture, and it may be like P4 Fail, so there are 3 years of important architectural changes coming up.

#52

TrackrI think we can all safely say..

If AMD didn't buy ATi, they wouldn't exist by now.

What are the known for, really?

Having better clock-per-clock performance 7-8 years ago? Is that really worthy of the admiration that I've seen over the past 6 years?

They've been making lousy CPUs for 6 god-damn years. I simply don't see any reason to be an AMD fanboy anymore. I never was one, I gave it the benefit of the doubt, now I longer do.

People who still like AMD get high out of betting on the under-dog.. and there's apparently a ton of them.

But don't get me wrong, in terms of competition, what is best for the world, the chances that computers will be merged with humans and achieve immortality is all dependent on Intel having some competition right about now.

I guess I'm just venting. Thanks a lot, AMD.

its been said a million times how if AMD goes under then we as consumers are screwed but it is true. I still buy AMD to support the underdog but also because they are by far the closest in competition to Intel. then there is the fact that Intel used illegal practices to stifle them instead of just trying to bury them through innovation so it is pretty hard for me to embrace a company like that

all of that aside though... AMD takes care of everything I throw at it so I am content. I am a tech junkie and a gamer so I put my rigs through lots of rigorous use through various ways. I dont need 10-30% better performance to play all of my games on the highest settings, I dont need it to stream 1080p movies, I dont need it to play lossless music, and I dont need it to transfer, sync, torrent, encode and compress files

I dont need Intel but I do need AMD to bring those prices down a tad to be more aligned with their relational performance to Intel

#53

FreedomEclipse

~Technological Technocrat~

lashtoni think AMD is too stuck in the low power usage analogy,

Have you actually read the some of the benchmarks???

they use a SNOT more power then their own current processors and also intels. and it gets worse when it comes to overlcocking...

jokes are already out that you need to buy a nuclear powerplant to run one of these

#54

nt300

TRWOVDon't be mistaken. That the PII X6 beats the FX-8150 in some benchmarks doesn't mean that it will beat it in all of them.

In your case it seems an issue with memory bandwidth, which PII lacks (Sandy Bridge has almost twice). The FX falls right in the middle in that regard.

With the fact that Piledriver is being released so soon I believe that AMD knew about the design's problems (branch prediction, pipeline flushing, cache trashing, decode unit not wide enough) but instead counted on higher frequencies to make up until Piledriver could be released. Anandtech's review also shows that cache latency is worse than Phenom II. Both problems can be blamed on Global Foundries' poor 32nm process. Cache latency can be increased and clockspeeds lowered to get higher yields.

I think that AMD saw the problems that needed reworking but decided that clockspeed would be enough to counter them for the time being but then a few months later they find that yields were too poor and had no choice but to launch as is.

Now you're thinking, I wouldn't be surprised if your statement is exactly how this came to be Good Job and hope you are right on :toast:
If you are Piledriver should become a CPU to recon with. :D

#55

qubit

Overclocked quantum bit

laszlopeople don't forget that amd barely managed to survive so R&D is not what should be; i don't even compare to Intel $7.3 billion R&D for 2011 when Amd total revenue is $6.49 billion in 2010 so we get what they afford nothing more or less .

Yes, I'll bet this is absolutely true and with bad management as well, they released a rubbish product and are misleading customers into thinking it's an octa core CPU, when it's only a quad core with multi-threading. They're outright lying.

I'll bet if someone rich enough sued them over this, they'd win. :shadedshu

#56

lashton

[H]@RD5TUFFI lol'd

That's what everyone said about bulldozer .. .. :shadedshu

That doesn't make sense, Bulldozer is NOT a bad CPU its anew tech which is very hard to release, I think people had too much high hopes for it, if you didn't (like me) then its a good CPU, the price will drop and then they will sell but AMD will make a killing on these because thier channel partners will sell these and AMD will be laughing all the way to the bank, you wait and see, the core I line is almost at an end, whats Intel gonna do now, create another P4?

#57

MikeMurphy

lashton2 things, why don't AMD add a massive amount like 20MB of level 3 cache and up the curre4nt Level 1 and 2 cache on their processors, sure decrease some overclock head room and increase a little TDP but it would add a SNOT load of performance, i think AMD is too stuck in the low power usage analogy, also Intel in 2014 will run out their core i line and will need to release another CPU architecture, and it may be like P4 Fail, so there are 3 years of important architectural changes coming up.

Are you prepared to pay $380 for this CPU with an additional 20mb of cache?

You know cache costs money, right?

#58

MikeMurphy

lashtonThat doesn't make sense, Bulldozer is NOT a bad CPU its anew tech which is very hard to release, I think people had too much high hopes for it, if you didn't (like me) then its a good CPU, the price will drop and then they will sell but AMD will make a killing on these because thier channel partners will sell these and AMD will be laughing all the way to the bank, you wait and see, the core I line is almost at an end, whats Intel gonna do now, create another P4?

I didn't have high hopes for it, I just expected it to be measurably better than the K10.5 line of CPUs it replaces.

You think the price will readily drop and AMD will make a "killing" on these? Do you have any idea how much these things cost to make? 2 billion transistors at 32nm node, is VERY expensive. MUCH more expensive than the superior Intel chips. I wonder if AMD will ever be competitive with these, which begs the question as to why they were ever released as desktop chips in the first place.

Intel is a few generations ahead of AMD, and given BD, just pulled even further ahead.

I love AMD at heart and write this from my fantastic Llano setup, but WTF. What happened to designing great chips?!? I'll keep my Llano, thanks.

#59

pantherx12

qubitYes, I'll bet this is absolutely true and with bad management as well, they released a rubbish product and are misleading customers into thinking it's an octa core CPU, when it's only a quad core with multi-threading. They're outright lying.

I'll bet if someone rich enough sued them over this, they'd win. :shadedshu

I said before man, from a technical stand point* it is a 8 core chip as it can run two threads on a module at the same time.

Hyper threading CANNOT do this ( it essentially makes a nice orderly que)

Stop being so mad dude because you didn't understand or didn't read into what bulldozer is.

They're was slides about 1 year ago showing off the design and TPU had a news story every other week about it.

It's an 8 core chip with some scheduling problems at the moment.

*Go on wiki pedia and read the central processing unit page and the core page.

#60

springs113

Hearing all this and talk reminds me of my amd CPUs....thorough rev A 2700, Barton 2500, Venice 3000, athlon x2 5200,x2 7700, p2 720be, p2 955be, n was hoping to put a bd system together but bd let me down not to mention my p2 955 system with dual 4850s is a mini space heater...really wanted to stay but couldn't wait anymore, maybe next build.

#61

ensabrenoir

m o n e y

Inspite of what we tech types think.... do we have any sales data ? Launch day #s anything. Even if A md,can't fool a couple thousand learned men and women....there are billions of lemmings left

#62

springs113

We also gotta remember that we are minority, as a previous poster amd probably gonna make a killing with OEM s

#63

HalfAHertz

As others have said, these cpus will be great for servers.

1) Servers use slower memory and cache is like steroids for them - BD has metric tons of L2
2) Servers need lots of memory and memory bandwidth - BD has much better IMC performance coming close to 75% of SB's memory performance and nearly doubling PH II's
3) cores are like money in the server world, so the more you have of them, the more important you are and BD server chips come with 16 of those - that's the equivalent diamond bling teeth
4)AMD's stuff sells for less and they're much more willing to compromise compared to $ntel - all the MBA managers will drop to da floo' once they see those savings
5)...
6)PROFIT!!?

#64

zx679

HalfAHertzAs others have said, these cpus will be great for servers.

1) Servers use slower memory and cache is like steroids for them - BD has metric tons of L2
2) Servers need lots of memory and memory bandwidth - BD has much better IMC performance coming close to 75% of SB's memory performance and nearly doubling PH II's
3) cores are like money in the server world, so the more you have of them, the more important you are and BD server chips come with 16 of those - that's the equivalent diamond bling teeth
4)AMD's stuff sells for less and they're much more willing to compromise compared to $ntel - all the MBA managers will drop to da floo' once they see those savings
5)...
6)PROFIT!!?

Any server application benchmarks out there that would let us see how bulldozer performs in that environment?

#65

Velvet Wafer

lashton2 things, why don't AMD add a massive amount like 20MB of level 3 cache and up the curre4nt Level 1 and 2 cache on their processors, sure decrease some overclock head room and increase a little TDP but it would add a SNOT load of performance, i think AMD is too stuck in the low power usage analogy, also Intel in 2014 will run out their core i line and will need to release another CPU architecture, and it may be like P4 Fail, so there are 3 years of important architectural changes coming up.

Like an AMD Engineer once stated: "Cache is the Fat of the Processor"
Its pretty expensive, and a good Architecture is performant, even with small caches (they just have to be fast enough)
gigantic caches take a lot of die space,and cost a lot of watt (and logically, heat),
if im not wrong;)
Bigger Caches give a little speed, for a lot of Tradeoffs... i bet if AMD would have been able to design BD with smaller caches, but at the same performance level, cost,temperature and wattage would be MUCH better;)

@all
Dont loose hope people, in a parallel Universe, BD earned a performance increase of several hundred percent, and was able to more than double the power of SB, at only 65w TDP, with 4.5ghz stock, 5ghz turboboost, 6ghz under air,7ghz under Water, 8 under DICE and Dual Stage, 12 under LN2, and 15ghz under He2. And that, for under 150$! :D
According to the multiversal theory, that really happened!
Just not here ;)

#66

HalfAHertz

Velvet WaferLike an AMD Engineer once stated: "Cache is the Fat of the Processor"
Its pretty expensive, and a good Architecture is performant, even with small caches (they just have to be fast enough)
gigantic caches take a lot of die space,and cost a lot of watt (and logically, heat),
if im not wrong;)
Bigger Caches give a little speed, for a lot of Tradeoffs... i bet if AMD would have been able to design BD with smaller caches, but at the same performance level, cost,temperature and wattage would be MUCH better;)

@all
Dont loose hope people, in a parallel Universe, BD earned a performance increase of several hundred percent, and was able to more than double the power of SB, at only 65w TDP, with 4.5ghz stock, 5ghz turboboost, 6ghz under air,7ghz under Water, 8 under DICE and Dual Stage, 12 under LN2, and 15ghz under He2. And that, for under 150$! :D
According to the multiversal theory, that really happened!
Just not here ;)

Me and my 10 sexy wives all use BD in the 2nd Universe!

#67

3volvedcombat

This slide is ludicrous,
Bulldozer- was not really what it was supposedly marketed in the first place, more like a fiat.
Piledriver- Only 10-15% boost, it will finally get up to the i72600-2500k reasonably well.. barly..
Steamroller- It will finally be the real marketed "bulldozer" we were all waiting for!!!!!!
Excavator- well I cant even predict how this will compete its only the future!!!!!

There going to make tons of money if they do it right, simply because of the number of releases in consecutive years!!!

Some people are shelling out 250$ fx8150's
It will probably be another 150-250$ for the Piledriver's
It will probably be another 100-250$ for the Steamroller's
It will probably be another 100-250$ for the Excavator's
Count in all the regular OEM sale's ect ect... And if they manufacture everything with reasonable stock management they will make bank

#68

techtard

qubitYes, I'll bet this is absolutely true and with bad management as well, they released a rubbish product and are misleading customers into thinking it's an octa core CPU, when it's only a quad core with multi-threading. They're outright lying.

I'll bet if someone rich enough sued them over this, they'd win. :shadedshu

It is a real octal core chip. They are just using 8 weak cores with some shared resources. They payed the price for gambling on heavilly multithreaded apps in the consumer userspace. We are still in firmly last gen when it comes to multithreading. They were looking too far ahead.
If they and the major OS distributors can rework the scheduling, they may improve performance slighty. Not enough to call this chip a success, but enough to know what to adjust for piledriver.

They should have released the server chips first, delayed a few more months and tailored the desktop chips for todays software.

It would have been better to suffer a delay and more internet moaning than the huge negative hype of a botched release.

Hopefully they can actually hit their 50% target with their future revisions.

#69

xenocide

3volvedcombatThis slide is ludicrous,
Bulldozer- was not really what it was supposedly marketed in the first place, more like a fiat.
Piledriver- Only 10-15% boost, it will finally get up to the i72600-2500k reasonably well.. barly..
Steamroller- It will finally be the real marketed "bulldozer" we were all waiting for!!!!!!
Excavator- well I cant even predict how this will compete its only the future!!!!!

There going to make tons of money if they do it right, simply because of the number of releases in consecutive years!!!

Some people are shelling out 250$ fx8150's
It will probably be another 150-250$ for the Piledriver's
It will probably be another 100-250$ for the Steamroller's
It will probably be another 100-250$ for the Excavator's
Count in all the regular OEM sale's ect ect... And if they manufacture everything with reasonable stock management they will make bank

That's why I didn't go AMD last time around. When I had my 3800+ X2, and wanted to upgrade, Phenom I and a Q6600 were my options. If I had gone with AMD, I would have needed to upgrade twice just to get the same performance I had for 2 solid years with my Q6600. It would have been a cheaper initial investment, but in the long run, I find Intel CPU's to have a better shelf life.

#70

Bjorn_Of_Iceland

3volvedcombatThis slide is ludicrous,
Bulldozer- was not really what it was supposedly marketed in the first place, more like a fiat.
Piledriver- Only 10-15% boost, it will finally get up to the i72600-2500k reasonably well.. barly..
Steamroller- It will finally be the real marketed "bulldozer" we were all waiting for!!!!!!
Excavator- well I cant even predict how this will compete its only the future!!!!!

There going to make tons of money if they do it right, simply because of the number of releases in consecutive years!!!

Some people are shelling out 250$ fx8150's
It will probably be another 150-250$ for the Piledriver's
It will probably be another 100-250$ for the Steamroller's
It will probably be another 100-250$ for the Excavator's
Count in all the regular OEM sale's ect ect... And if they manufacture everything with reasonable stock management they will make bank

Hope it turns out rock solid with them. Else theyre just digging their own coffin with those Excavators and Bulldozers and stuff

#71

archangel

maybe intel buyed AMD and we do not know about! now intel rise the price of all products... :laugh:

#72

erocker

I don't see AMD as a company by 2014. I don't see how they will be able to stay afloat. That 50% is needed now in 2011.

#73

laszlo

erockerI don't see AMD as a company by 2014. I don't see how they will be able to stay afloat. That 50% is needed now in 2011.

they will;low cpu prices and good GPU's will solve the financial problems so they'll float not like Intel in a luxury yacht just in a small rubber salvage one...

another variant will be to sell gpu dept or to remain only with it;the shareholders and big "oil" rich investors will step in when shares will fall 2 high.... and we'll see maybe amd falling apart to smaller companies

#74

yogurt_21

erockerI don't see AMD as a company by 2014. I don't see how they will be able to stay afloat. That 50% is needed now in 2011.

? so you think a company that just got all 3 major console companies to use their chips in the nex gen consoles isn't going to be around?

ati was too much for amd to take on and it nearly killed them. But right now amd pretty much reaps the benefit of that decision and it's about to get far more lucritive. By 2014 they'll have plenty in the bank. (nowhere near intel, but far closer to nvidia's bankroll)

sure the processor deivision failed to impress, what's new? Amd's still around now and they'll continue to be around and thriving by 2014.

#75

qubit

Overclocked quantum bit

yogurt_21? so you think a company that just got all 3 major console companies to use their chips in the nex gen consoles isn't going to be around?

ati was too much for amd to take on and it nearly killed them. But right now amd pretty much reaps the benefit of that decision and it's about to get far more lucritive. By 2014 they'll have plenty in the bank. (nowhere near intel, but far closer to nvidia's bankroll)

sure the processor deivision failed to impress, what's new? Amd's still around now and they'll continue to be around and thriving by 2014.

The way they're going, I reckon they'll end up closing their x86 business and focusing on things like graphics cards and branching out into other areas where they have no presence at the moment. An AMD Android smartphone, perhaps? However, I'd never count them out until it actually happens, so who knows?

Add your own comment

Bulldozer Aims For 50% Improvement By 2014: Is This Really Enough To Counter Intel?

132 Comments on Bulldozer Aims For 50% Improvement By 2014: Is This Really Enough To Counter Intel?

Latest GPU Drivers

New Forum Posts

Popular Reviews

Controversial News Posts

Bulldozer Aims For 50% Improvement By 2014: Is This Really Enough To Counter Intel?

Related News

132 Comments on Bulldozer Aims For 50% Improvement By 2014: Is This Really Enough To Counter Intel?

Latest GPU Drivers

New Forum Posts

Popular Reviews

Controversial News Posts