Overclocking Preview Bulldozer 8150

[Spikesoldier]

i wanted to see superpi 1M at those frequencies.

 

[garagisti]

http://semiaccurate.com/2011/09/13/amd-sets-clock-speed-world-record-with-bulldozer/

 

[RussianSensation]

Different chip processes, RussianSensation, something to look for upon BD release is maximum 24/7 voltage. Did Intel ever give guidance on max safe SB voltage or is it still OC community consensus?

Not officially.

The VID voltage max for 45nm i7 (Lynnfield) like i7 860-880 is 1.400V
The VID voltage max for 45nm i7 (Nehalem) like i7 920-960 is 1.375V
The VID voltage max for 32nm i7 (Gulftown) like i7 980/990 is 1.375V
The VID voltage max for 32nm Xeon 5000 series is 1.35V

A lot of experienced overclockers like aigo have commented that 32nm Intel processors can fail at 1.45V. So the general consensus is that after 1.38V-1.40V on air cooling is going beyond 'safe' specs. Of course if you change your processor every 6-12 months, or use exotic cooling, then sure 1.5-1.55V is not a problem.

 

[RussianSensation]

First of all you clearly are new to extreme overclocking and probably know very little about it. But in this task it has done what no other cpu before has achieved. Got it?

My bad, I didn't know that "extreme overclockers" get this defensive and personal about overclocking with one module and 1/2 the L3 cache disabled.

"Hot Hardware was on hand for the event and claims that the CPUs used for testing were early AMD FX-8150 processors, and that only two of the eight cores were enabled for the record-setting run."

I was not impressed by an 8.3ghz Celeron LN2 that gets destroyed in performance by a 4.0ghz air cooled X6 1100T. I guess the thread title was misleading then because it said nothing about LN2 overclocking for 1 minute.

Running from thread to thread with tear down AMD is getting kinda childish.

Simply discussing the merits of performance range for AMD's new processors. I didn't start 5-10 threads on BD. It would have been a lot easier to have 1 aggregate thread.

 

[psolord]

What about it? 1.5V on 32nm SB and 4.8ghz is horrible. We have seen people achieve that on our forum with lower volts.

You specifically said "We don't know the voltage they used to achieve 4.8-5.0ghz overclocks though" and I just pointed it out that we do. I am sorry if I misunderstood you mate.

I never said anything about 1.5V on the SB.

I did say earlier though, that from what I've seen in various reviews, it seems that SB hovers around 1.45V for 5Ghz. 1.5V for AMD's power scheme is not that much in comparison. And to be clear, that statement refers to current AMD power schemes. I don't know what power scheme they have chosen for BD, but given that they will later couple it with a Radeon GPU, which also follows the low current/high voltage scheme, I'd guess that BD is on the same trail. Heck they are destined to work on AM3+ mobos, which are already made with this scheme in mind, so...!

 

[RussianSensation]

You specifically said "We don't know the voltage they used to achieve 4.8-5.0ghz overclocks though" and I just pointed it out that we do. I am sorry if I misunderstood you mate.

Oh my bad, I missed that part on your part. I apologize. So they achieved a 4.8-5.0ghz on BD at 1.5V on air cooling then?

 

[psolord]

Oh my bad, I missed that part on your part. I apologize. So they achieved a 4.8-5.0ghz on BD at 1.5V on air cooling then?

No worries!

They achieved 4.8Ghz with 1.5V on an Antec Khuler something (water cooler), that the general consensus is that it's equal to high end Noctua air coolers.

 

[RussianSensation]

No worries!

They achieved 4.8Ghz with 1.5V on an Antec Khuler something (water cooler), that the general consensus is that it's equal to high end Noctua air coolers.

More info here:
http://www.techspot.com/news/45454-amd-sets-guinness-world-record-with-overclocked-fx-cpu.html

"A phase change cooler was installed at the next station and took another FX CPU to 5.894GHz using 1.632 volts."

I wonder what voltage was used to get 8.4ghz overclock? 1.9-2V?

 

[sawtx]

More info here:
http://www.techspot.com/news/45454-amd-sets-guinness-world-record-with-overclocked-fx-cpu.html

"A phase change cooler was installed at the next station and took another FX CPU to 5.894GHz using 1.632 volts."

I wonder what voltage was used to get 8.4ghz overclock? 1.9-2V?

2.02V for 8.4GHz, looks like there was a run at 8.14GHz and 1.944V. Seems BD can take more voltage than previous chips, AMD's previous record was 7.378GHz at 1.88V (per overclockers.com).

 

[Vesku]

It was 2 or 2.1V for the record run if I read correctly.

 

[busydude]

I wonder what voltage was used to get 8.4ghz overclock? 1.9-2V?

2.01x V.

 

[sm625]

So what happens when we get the cheapest 4 core BD and clock it up to 4.6GHz with a $20 cooler and it ends up being faster than anything intel sells for under $200?

 

[GaiaHunter]

As for IPC, I think this general rule will stick to be true: from 5% slower than Llano to 5% faster. When both cores in a module are being used, 180% performance out of 200%. Given this, you can definitely expect that an FX-4100 at 4GHz will be slower than a Phenom II X4 955 at 4GHz in multi-threaded.

A BD module do share resources, but with the exception of those shared resources (that are different from the equivalent on a stars core), a BD module also generally has more resources than 2 phenom II cores.

For example, even dividing the BD's FPU by 2 integer cores, that FPU can do more than 2 "stars core" FPU.

 

[AtenRa]

So what happens when we get the cheapest 4 core BD and clock it up to 4.6GHz with a $20 cooler and it ends up being faster than anything intel sells for under $200?

As i have said before, i expect AMD FX4100 close to $150-170. At that point it could be the best price/performance CPU for Desktop/gaming when OCed, add better featured motherboards than Socket 1155 at the same price and you have a winner

 

[psolord]

So what happens when we get the cheapest 4 core BD and clock it up to 4.6GHz with a $20 cooler and it ends up being faster than anything intel sells for under $200?

That would be great, but people would argue that a 2500k is marginally more expensive and at the end of the day, it would add a little more to the overall system cost. You also can OC the 2500K as well. So let's hope that the 4100 will stay as close to 150$ as possible, so the price difference will be around the 60-70$ mark.

I have two more questions though.

1) What happens if that 4100 has a good chance being unlocked to an 8100

2) What happens if AMD launches a FX X2 at some later time, unlocked and ready to hit 5Ghz, costing less than 100$? This could be the 2120k that never came.

 

[Edrick]

So what happens when we get the cheapest 4 core BD and clock it up to 4.6GHz with a $20 cooler and it ends up being faster than anything intel sells for under $200?

Considering the 2500K sells for $179 now, I highly doubt the cheapest 4 core BD will beat it.

 

[ransomlist]

As i have said before, i expect AMD FX4100 close to $150-170. At that point it could be the best price/performance CPU for Desktop/gaming when OCed, add better featured motherboards than Socket 1155 at the same price and you have a winner

The issue i suspect we might have is how much headroom we have in those chips- are they 4 'cores' because there are defects in the others (in two bulldozer modules)? Is the small turbo increase indicative of more than the max clock possible when staying within specified TDP?

If Bulldozer modules behave more like 4 cores than dual core w/ HT then the 2 Bulldozer module chip is going to be a potent price/performance contender. For average Joe, the FX4's with the nice features but reasonable £/$ boards is going to sell well- provided AMD get their heads screwed on right regards marketing and retailers realize that competition benefits them in the long term.

EDIT: Regardless, i want BD out before later October and competitive with an 2500K at 6 'cores' and above- i want! Oh christ, spelling mistakes everywhere.

 

[sequoia464]

1) What happens if that 4100 has a good chance being unlocked to an 8100

I'm using an Asus Sabertooth 990FX at the moment - it will not unlock my PhenomII 555X2. Not to say that they won't include a way to unlock in future a bios. I think i read somewhere thay MSI has a board that will unlock - at least the Phenoms. Not certain of that though - maybe someone can correct or confirm.

 

[Arkaign]

So what happens when we get the cheapest 4 core BD and clock it up to 4.6GHz with a $20 cooler and it ends up being faster than anything intel sells for under $200?

That's really what I'm hoping honestly. Real competition in the sub $200 area would be nice.

 

[Arkaign]

Considering the 2500K sells for $179 now, I highly doubt the cheapest 4 core BD will beat it.

What?

http://www.newegg.com/Product/Produc...82E16819115072

$219

http://www.amazon.com/Intel-i5-2500K...5941816&sr=8-1

$219

http://www.ebay.com/itm/Intel-Core-...NEW-/320702288071?pt=CPUs&hash=item4aab5890c7

$218

??????

Now, if the 4-core BD turns out to be a good value, then I could see Intel lowering the price down a bit. Is that what you're getting at?

 

[aigomorla]

So what happens when we get the cheapest 4 core BD and clock it up to 4.6GHz with a $20 cooler and it ends up being faster than anything intel sells for under $200?

well id like you to show me that..

because from what im seeing thats not going to be possible.

How about you pulling somewhere between 4.6ghz on your Bulldozer only to realize a stock 2600K is hugging your rear playing with you.

 

[ElFenix]

??????

microcenter

 

[Arkaign]

microcenter

Ah. I don't really consider that a fair price to compare, as not everyone is near a MC and they don't sell them online for that.

 

[psolord]

well id like you to show me that..

because from what im seeing thats not going to be possible.

How about you pulling somewhere between 4.6ghz on your Bulldozer only to realize a stock 2600K is hugging your rear playing with you.

We would be talking twice the price difference then.

It wouldn't be too bad for AMD's offering, considering that the 2600K already has more power than most people need. Even gamers. People on a budget would rejoice if they could extract 300$ worth of performance from a 150$ part.

 

[Neutralman]

We would be talking twice the price difference then.

It wouldn't be too bad for AMD's offering, considering that the 2600K already has more power than most people need. Even gamers. People on a budget would rejoice if they could extract 300$ worth of performance from a 150$ part.

Agreed. I would totally buy the 8core BD if its OC'ed self was similar to how a stock 2600k performs.