Fudzilla: Bulldozer performance figures are in

[sm625]

Buying a $350 Bulldozer over a $200 Intel 2500K with a small margin of performance gain just to give AMD your money is flat out stupid if you aren't even going to use it

AMD is pricing that chip to beat 2500k by at least 20% in single threaded performance. And to totally crush it in multithreaded performance. For $350 it should beat 2600k by 20% in multithreaded. If AMD cannot do this then they have an overpriced part. I am certain they wont bring it to market at $350 if it cant perform. The only question is... will they bring it to market lol.

 

[Davidh373]

AMD is pricing that chip to beat 2500k by at least 20% in single threaded performance. And to totally crush it in multithreaded performance. For $350 it should beat 2600k by 20% in multithreaded. If AMD cannot do this then they have an overpriced part. I am certain they wont bring it to market at $350 if it cant perform. The only question is... will they bring it to market lol.

My intent with that comment was for it to be an "if" statement, but there are people that would do it. There is no doubt in my mind the saying "A fool is easily parted with his/ her money" applies to many of the fanboys on both sides of the fence.

 

[Arkaign]

AMD is pricing that chip to beat 2500k by at least 20% in single threaded performance. And to totally crush it in multithreaded performance. For $350 it should beat 2600k by 20% in multithreaded. If AMD cannot do this then they have an overpriced part. I am certain they wont bring it to market at $350 if it cant perform. The only question is... will they bring it to market lol.

The K series is pretty limited run stuff though in the big picture. Maybe they're shooting for the i5-2500 non-K? Because most K-series customers will have a P67 or Z68 or whatever and have it running at 4.2-4.5ghz, which is a considerably higher target, not to mention on a performance plane that is far beyond the realm of even 580GTX SLI rigs @ appropriate settings (say 1920x1080 and beyond with max everything).

Now if the overclocking performance of a BD part can beat the 2500k @ ~4.4ghz in gaming? That'd be incredible.

 

[Idontcare]

No, but the title of this thread is "Fudzilla: Bulldozer performance figures are in." Let's stay on topic. ARM and GPUs have nothing to do with Bulldozer other than AMD. I don't see why people have to bring the company's other products into a debate when clearly bulldozer is a desktop processor. It isn't a cell phone, and it's not a graphics card. I'm fine talking about the company that makes it, and competing companies which have products that directly compete with it. Intel's SB and SB-E as well as Haswell will all compete and be compared with Bulldozer and it's future revisions. I don't see why people need to turn this into an orgy of fanboyism and snipe at one another.
I'm not a fanboy for either side, but for good reason I find myself agreeing with the Intel side. The fact AMD people are screaming over these numbers like it's the most amazing and unbelievable thing ever is clearly over-exaggeration. Those people make themselves look bad. Like you said, there are no official numbers yet, only rumored ones. Given they are rumored numbers, I would think they're more likely to be twisted in a positive direction for AMD.
The fact is both Intel and AMD have their ups and downs. So let's take the (rumored) numbers we have with a grain of salt, and (hopefully) come to the consensus that fanboyism is ignorant and foolish, and getting your hopes up for Bulldozer or SB-E to kick the other's ass is up there on the same level. They both make good processors, but normally for different markets. Let's be grateful for our budget Phenom IIs and our performance Intel products and stop trying to expect them to flip flop just so you can FINALLY get the numbers from the company you desire. Buying the best product is supposed to drive the market in the direction you want it to go. I have to wonder if all the fanboys in the world just bought what was best instead of biting their lip and going for either a worse product or a much more expensive one that both Intel and AMD would be closer in the race, and probably make even larger gains in that race. If you AMD people want a performance processor more often, buy an Intel every once in a while to show them you want them to put more effort into R&D. If you Intel fanboys are fed up with how much Intel charges for only slight gains, buy an AMD. I have both in my systems because they both excel in different areas. Buying a $350 Bulldozer over a $200 Intel 2500K with a small margin of performance gain just to give AMD your money is flat out stupid if you aren't even going to use it, just like buying a 990X from intel to get an extra 2 cores for an extra $750 is stupid. Just chill on the fanboyism already!

Why so angry bro? We are 500+ posts into a Fudzilla-based rumor thread...the topic is gonna wander around a bit.

Feel free to continue talking about the OP from 500 posts ago but there's not much value to come from chastising the relatively small portion of the posters in this thread for having moved on to more interesting (to them) side-discussions within the thread in the meantime.

Telling fanboys to chill on the fanboyism is like asking someone to stop breathing. Unrealistic expectations.

 

[Idontcare]

AMD is pricing that chip to beat 2500k by at least 20% in single threaded performance. And to totally crush it in multithreaded performance. For $350 it should beat 2600k by 20% in multithreaded. If AMD cannot do this then they have an overpriced part. I am certain they wont bring it to market at $350 if it cant perform. The only question is... will they bring it to market lol.

They may also be pricing the chip based on the fact that even if it doesn't beat a 2600K on price/performance it will still sell to a number of diehard fans provided it is simply "close enough".

AMD is just getting into 32nm...supply will be naturally tight. They need merely price them such that the limited initial supply is met with the demand that ensures most are sold. As supply increases, both from node maturity advances and capacity ramps, prices would then naturally be lowered so as to increase demand commensurately.

I would LOVE for supply/demand economics to NOT be in play here, but AMD is smarter than that. When supply was tight for Cypress parts they boosted prices to bring demand back inline with supply.

There is no reason for us to expect GloFo to be prepared to supply enough wafers at this stage for AMD to be able to support an Intel-level supply of 2600K competing chips and AMD is not unintelligent to the point of failing to manage this reality as they price out Zambezi.

It would be awesome for us if AMD wasn't ran so efficiently, it would be great if they did produce a limited supply of top-end Zambezi's and merely priced them accordingly to their price/performance competitiveness with SandyBridge regardless the fact the limited number of those top-end parts would sell out in the first 15 minutes. But somehow I just can't bring myself to assume AMD would do that.

A $1k 990X is not price/performance competitive with a 2600K, and I fully expect a top-end Zambezi to not be as well, at least not until GloFo fully ramps 32nm capacity and AMD begins to have more top-end Zambezi's to sell than the market will bear at the initial intro prices.

 

[ElFenix]

AMD is pricing that chip to beat 2500k by at least 20% in single threaded performance. And to totally crush it in multithreaded performance. For $350 it should beat 2600k by 20% in multithreaded. If AMD cannot do this then they have an overpriced part. I am certain they wont bring it to market at $350 if it cant perform. The only question is... will they bring it to market lol.

$350 for 20% better at multithreaded is probably a bargain. the 2600k is 50% more expensive than the 2500k and is not much more than 20% better at multithreaded (depending on the benchmark)

 

[AtenRa]

AMD is just getting into 32nm...supply will be naturally tight.

AMD main distributor in Greece has already dry out of AMD Llano APUs (shops still have APUs to sell), 32nm Yields must be low still and BD is bigger in size.

 

[Mopetar]

AMD main distributor in Greece has already dry out of AMD Llano APUs (shops still have APUs to sell), 32nm Yields must be low still and BD is bigger in size.

I'm surprised that they're selling so well given the economic problems. Either that or rioters broke into the shops and cleaned out all the parts

 

[ElFenix]

AMD main distributor in Greece has already dry out of AMD Llano APUs (shops still have APUs to sell), 32nm Yields must be low still and BD is bigger in size.

is it? llano is about 220 mm2

 

[soccerballtux]

still waiting for hard performance numbers from a legit site. When is this slated to ship? When do most here think it actually will ship? I thought this was supposed to be out in may or something...got pushed back to July? To September? Maybe it'll be out by January...

 

[Idontcare]

JF's comments were made before current top-bin MC's were released. He's said a few times that the 50% comment should only be compared to the top-bin MC at the time he said that.

From BSN:

The top-end 12-core MC part at the time was a 2.3 GHz Opteron 6176 SE, released on Mar 29, 2010.

Since then AMD has launched a 2.5 GHZ Opteron 6180 SE, on Feb 14, 2011.

It would require some seriously wierd clockspeed scaling going on for MagnyCours, for Interlagos to be 50% faster than a 2.3 GHz MC while only being 35% faster than a 2.5GHz MC.

If you crank out the math, back of the envelope style, it becomes rather obvious that it would appear AMD lost their intended top-end Interlagos clockspeed bin somewhere on the road between last Nov 9 and the earnings CC yesterday.

Given the power-consumption versus clockspeed we've seen for Llano, this would not surprise me one bit if it were true. As evidenced by Llano's clockspeed/W and performance/W, 32nm clearly is in the early stages of being production worthy.

Come Q4 we'll no doubt see new steppings and a more mature process tech, but coming out of the gate things are looking a little challenging (and not surprising, there is a reason why Zambezi launch was delayed after all).

 

[wlee15]

The Interlagos platform is our first server offering, optimized for today's cloud datacenters and the architecture excels at compute-intensive and HPC workloads, where it will deliver up to 35% performance improvements compared to our current offerings.

Seems to me that Thomas Seifert was referring to HPC performance increases.

 

[RampantAndroid]

$350 for 20% better at multithreaded is probably a bargain. the 2600k is 50% more expensive than the 2500k and is not much more than 20% better at multithreaded (depending on the benchmark)

I'm curious: Where do we have solid figures backing up these perf claims?

 

[Idontcare]

I'm curious: Where do we have solid figures backing up these perf claims?

http://www.anandtech.com/bench/Product/288?vs=287

 

[AtenRa]

I'm surprised that they're selling so well given the economic problems. Either that or rioters broke into the shops and cleaned out all the parts

It is easy to be sarcastic (A Greek word like many more that you use in your language ) but it is harder to put your brain in motion and understand that i have said the Main Distributor run out of CPUs but Shops still have Llano in Stock.

That means the distributor had a small quantity of Llano CPUs in the first place and they expecting a new shipment of Llanos in one month from now and the shops didnt sell that much because they still have them in stock.

 

[AtenRa]

Given the power-consumption versus clockspeed we've seen for Llano, this would not surprise me one bit if it were true. As evidenced by Llano's clockspeed/W and performance/W, 32nm clearly is in the early stages of being production worthy.

Since Llano is using different types of Xtor (smaller for higher density) than Bulldozer, i believe we cant compare Llano's clockspeed/w and performance/w vs Bulldozer.

And perhaps that's one of the reasons why Llano have that clockspeed/performance per Watt characteristics in the first place.

ps: im not saying the manufacturing process dont play a role.

 

[podspi]

The top-end 12-core MC part at the time was a 2.3 GHz Opteron 6176 SE, released on Mar 29, 2010.

Since then AMD has launched a 2.5 GHZ Opteron 6180 SE, on Feb 14, 2011.

It would require some seriously wierd clockspeed scaling going on for MagnyCours, for Interlagos to be 50% faster than a 2.3 GHz MC while only being 35% faster than a 2.5GHz MC.

If you crank out the math, back of the envelope style, it becomes rather obvious that it would appear AMD lost their intended top-end Interlagos clockspeed bin somewhere on the road between last Nov 9 and the earnings CC yesterday.

Given the power-consumption versus clockspeed we've seen for Llano, this would not surprise me one bit if it were true. As evidenced by Llano's clockspeed/W and performance/W, 32nm clearly is in the early stages of being production worthy.

Come Q4 we'll no doubt see new steppings and a more mature process tech, but coming out of the gate things are looking a little challenging (and not surprising, there is a reason why Zambezi launch was delayed after all).

Actually, my back-of-the-envelope math works out to ~ 38% faster. Maybe they just rounded down? (People are much more accepting of under-estimation than over-estimation).

Here are my (admittedly simplistic, it is on the back of an envelope after all) calculations:

Opteron 6176 SE:1.0000
Bulldozer (Est) :1.5000
Opteron 6180 SE:1.0869

% change from 6176-->6180:8.69%
% change from 6176-->BD :50%
% change from 6180-->BD :38%

No weird scaling issues or loss of top-bin required, just some rounding

Of course, there are some pretty aggressive assumptions embedded in this simple calculation, but iirc wasn't the original statement about SPECint, which scales relatively linearly wrt clockspeed AND cores?

 

[Mem]

still waiting for hard performance numbers from a legit site. When is this slated to ship? When do most here think it actually will ship? I thought this was supposed to be out in may or something...got pushed back to July? To September? Maybe it'll be out by January...

Saw this link posted over@OCUK about launch date ,http://wccftech.com/amd-bulldozer-launch-september-official-8-core-fx-processors-arrive-q1-2012/ .

Taiwanese Motherboard manufacturers have leaked details of AMD’s Upcoming Bulldozer CPU’s and Release Date. According to them, AMD will start Mass Production of B2 chips in August and the processors will officially launch in September (Q3 2011)"]Taiwanese Motherboard manufacturers have leaked details of AMD’s Upcoming Bulldozer CPU’s and Release Date. According to them, AMD will start Mass Production of B2 chips in August and the processors will officially launch in September (Q3 2011)

So looks like September.

 

[novasatori]

aaaaaaaaaahhhhhhhhhhhhhhhhhhhh

I am waiting for these CPUs... specifically the server ones...

I'm interested in building a server for encoding & virtual machines.. 2x8 core AMDs seem pretty good especially at their current prices($259 ea), but I figure I should hold out until the next ones come .... if only it would happen

 

[RussianSensation]

AMD is pricing that chip to beat 2500k by at least 20% in single threaded performance.

You must be dreaming. AMD is at least 50% behind in IPC compared to SB (Phenom II X4 970 @ 3.5ghz vs. 2600k @ 3.4ghz) with the current Phenom II core. For Bulldozer to beat a 2500k in single threaded apps by 20%, it would need to have an 80% faster IPC performance than Phenom II is (Phenom II = 1.0x, SB = 1.5x ==> BD = 1.8x / 1.5x = 20% faster), or it would have to launch with insanely high clock speeds. But 2500k overclocks to 4.5ghz with ease....So the chance of Bulldozer outperforming 2500k in single threaded apps is 0.

For $350 it should beat 2600k by 20% in multithreaded. If AMD cannot do this then they have an overpriced part. I am certain they wont bring it to market at $350 if it cant perform.

What? AMD motherboards cost less than Intel's on average. Therefore, Bulldozer just needs to be as fast as a 2600k to be competitive at $350. Why would it need to be 20% faster or it's "overpriced"? But since Bulldozer is an 8-core module CPU, it likely means AMD is aiming to have better performance in multi-threaded apps because they know they have no chance of competing in less threaded apps due to such a huge IPC deficit.

BD could be 30% faster in every benchmark and intel would still not drop their prices.

There has been so much hype behind BD, it's not even funny anymore. I bet 1 week before launch people will claim that it's 30% faster than SB and post some useless Cinebench rendering benchmarks that hardly more than 0.5% of people care about to "prove it". I was here when Phenom I was hyped up, I was here when Phenom II was hyped up. My personal expectation is that BD at best will match 1st generation Core i5 CPUs for IPC and AMD will net the rest with higher clock speeds just to be competitive. Their inability to get BD out in the last 6 months is most likely related to those low clock speeds, which seem imperative to be competitive.

The 8-core BD will likely walk out beating 2500k/2600k in synthetic heavily-multithreaded apps, but still remain slower for most real world apps once overclocked BD is pitted against an overclocked 2500k/2600k. Obviously anything is better at this point than Phenom II is -- so at least we won't be disappointed on that end .

 

[LOL_Wut_Axel]

From BSN:

The top-end 12-core MC part at the time was a 2.3 GHz Opteron 6176 SE, released on Mar 29, 2010.

Since then AMD has launched a 2.5 GHZ Opteron 6180 SE, on Feb 14, 2011.

It would require some seriously wierd clockspeed scaling going on for MagnyCours, for Interlagos to be 50% faster than a 2.3 GHz MC while only being 35% faster than a 2.5GHz MC.

If you crank out the math, back of the envelope style, it becomes rather obvious that it would appear AMD lost their intended top-end Interlagos clockspeed bin somewhere on the road between last Nov 9 and the earnings CC yesterday.

Given the power-consumption versus clockspeed we've seen for Llano, this would not surprise me one bit if it were true. As evidenced by Llano's clockspeed/W and performance/W, 32nm clearly is in the early stages of being production worthy.

Come Q4 we'll no doubt see new steppings and a more mature process tech, but coming out of the gate things are looking a little challenging (and not surprising, there is a reason why Zambezi launch was delayed after all).

That's not really an accurate assessment when it comes to comparing Llano CPU clock speed to Sandy Bridge CPU clock speed. AMD invested most of the die area and transistors into the GPU portion while Intel did the opposite. AMD was also able to get competitive power consumption. This is comparing apples to oranges.

 

[Accord99]

That's not really an accurate assessment when it comes to comparing Llano CPU clock speed to Sandy Bridge CPU clock speed. AMD invested most of the die area and transistors into the GPU portion while Intel did the opposite. AMD was also able to get competitive power consumption. This is comparing apples to oranges.

With power gating, the GPU's contribution to power usage when running purely CPU loads should be minimal. Even then, the power usage of a 2.9GHz Llano quad-core when running a purely CPU load is comparable, if not higher than a 2600K despite the 2600K having roughly 2x the throughput.

 

[videoclone]

 

[badb0y]

AMD better release these by September, my upgrade itch is getting unbearable.

 

[guskline]

Boy am I glad I got a 2500k!