AMD PileDriver Discussion thread

[SickBeast]

Perhaps piledriver will pound bulldozer into the ground and give it an eternal resting place.

 

[NostaSeronx]

Bulldozer -> Piledriver
L1 DTLB has been increased to 64M for AMD Family 15h Models 10h-1fh processors. For
AMD Family 15h models 20h to 2fh processors, the L1 DTLB size has increased from 32 entries to
64 entries.

The following performance caveats apply when using streaming stores on AMD Family 15h cores.

• When writing out a single stream of data sequentially, performance of AMD Family 15h
processors is comparable to previous generations of AMD processors.
• When writing out two streams of data, AMD Family 15h version 1 processors can be up to three
times slower than previous-generation AMD processors. AMD Family 15h version 2 processor
performance is approximately 1.5 times slower than previous AMD processors.
• When writing out four non-temporal streams, AMD Family 15h version 1 can be up to three
times slower than previous AMD processors. AMD Family 15h version 2 processor performance
is comparable to previous AMD processors.
• Using non-temporal stores but not writing out an entire cacheline may cause performance to be up
to six times slower than previous AMD processors.

 

[IlllI]

what is there to discuss? It will still be worse than whatever intel has out at the time.

 

[daweyo]

Dear pubescent trolls,

Please leave and go post positive threads in your Intel forums, or go create an Anti-AMD forum somewhere else. Stop cluttering threads with your negative comments.
I know AMD is one of the most shorted stocks out there, and I am sure many of you have positions that stand to gain cash the farther AMD stock falls.
Please continue to go shop at Walmart and find the best prices for your Mountain Dew and Pringles.
And.. in between hate threads bashing a company that you have maybe one-thousandth of a percent of knowledge about operations and engineering, try to go find a girlfriend to take your frustrations out on instead of cluttering threads with your painfully repetitive rhetoric.

WE GET IT.

Thuban better. Intel better. Across ALL fields. Thanks for your input. Goodbye. Go to Intel threads. Let people who want to talk about constructive AMD stuff do their thang. I am SUPER happy you are so knowledgeable about benches, prices, instruction sets, IPC, Faps Per Second and overclocked gigglehurts. However, some of us like the mom and pop stores, so live with it.

GO AWAY. Otherwise you are a propagandist with a short position in a stock, or an angry pubescent boy who ... well... fill in the blanks...


We don't take this approach to our fellow forum colleagues here at the AnandTech Forums. You are new, and are given a "grace period" if you will. Immature posts like yours above are not acceptable. Clean up your attitude please.

For now, please take some time to familiarize yourself with the following portions of the AnandTech Forum Guidelines:

1) No trolling, flaming or personally attacking members. Deftly attacking ideas and backing up arguments with facts is acceptable and encouraged. Attacking other members personally and purposefully causing trouble with no motive other than to upset the crowd is not allowed.

We want to give all our members as much freedom as possible while maintaining an environment that encourages productive discussion. It is our desire to encourage our members to share their knowledge and experiences in order to benefit the rest of the community, while also providing a place for people to come and just hang out.

We also intend to encourage respect and responsibility among members in order to maintain order and civility. Our social forums will have a relaxed atmosphere, but other forums will be expected to remain on-topic and posts should be helpful, relevant and professional.

We ask for respect and common decency towards your fellow forum members.

Idontcare
Super Mod

 

[d4a2n0k]

Dear pubescent trolls,

SNIP

Shouldnt that be pre-pubescent?

 

[Ferzerp]

When are the mods going to stop this b.s.?

 

[3DVagabond]

When are the mods going to stop this b.s.?

You need to report the offending posts.

 

[soccerballtux]

at the rate dulldozer's going at the only thing it'll be driving is a pile of poo. That's why it's called piledriver.

 

[Keysplayr]

When are the mods going to stop this b.s.?

Yes, just use the report post function so that the mods may ban these trolls over and over again while they proxy it up.

 

[Keysplayr]

There's no harm in discussing Piledriver. It's interesting to see Keys start a witch hunt rather than contributing something useful.

I will say that I hope AMD can work out the issues in Bulldozer relatively quickly and bring us something decent. I have a feeling that there are a few bugs in the design that they can work out, particularly when it comes to power consumption.

That said, things do not look good for AMD. As others have said, 15% more performance is not enough.

If AMD can reign in the power consumption to be competitive with Intel, they will have an excellent workstation and server processor on their hands, which is their niche anyway.

BTW OP welcome to the AT forums.

The one who absolutely feasts on drama and unrest says he doesn't want to see a witch hunt. At least not one that I might start? Yeah ok, I buy that.
I think it might be time for you to be who you really are and not pretend you don't like this type of drama.. AMD is pretty much reaping what they have sown, as a friend of mine has just said to me.
I think it just that the pieces fall into place now and in the near future.

 

[Piano Man]

So many October specials coming out defending a horrid product. Lets just all move on.

 

[SickBeast]

The one who absolutely feasts on drama and unrest says he doesn't want to see a witch hunt. At least not one that I might start? Yeah ok, I buy that.
I think it might be time for you to be who you really are and not pretend you don't like this type of drama.. AMD is pretty much reaping what they have sown, as a friend of mine has just said to me.
I think it just that the pieces fall into place now and in the near future.

Ha! Your drivel made me laugh for once.

I will not engage you further lest this shall degenerate into a flame war. :thumbsup::thumbsup:

 

[LoneNinja]

If Pile Driver is 10% faster/better IPC than BD, atleast then AMD will finally beat an i7...

Maybe a 6 core Pile Driver will out perform a 6 core Phenom II lmao

 

[NostaSeronx]

Trailing Bit Manipulation (TBM):
TBM instructions provide instructions implementing individual bit 
manipulation operations such as isolating, masking, setting, resetting, 
complementing, and operations on trailing zeros and ones.

I can't find much more

But the Vishera CPUs will most likely have a lower clock that the Zambezi CPUs

As the Trinity APUs have a lower clock than the Llano APUs while scoring similarlyish(That is a feat)

Implying they increased instruction throughput somehow

 

[Keysplayr]

Ha! Your drivel made me laugh for once.

I will not engage you further lest this shall degenerate into a flame war. :thumbsup::thumbsup:

Good choice.

You and I != flamewar potential.

 

[Gaz550]

I was wondering who would be the first to post this reasoning. I really was!

And I'll also sit and truly wonder if you really believe what you just posted. Not publicly of course.

Extremely good reasoning that was. I registered here because I followed Bulldozer news and discussion here.

Or did I miss some subtle meaning?

 

[SuperH3art]

Is it a crime to start a thread on a upcoming product ??????

If bulldozer failed does it mean that Pildriver will also fail....

Is it a crime to be a "new reg"????? If that so ban me and I will never come back on this forum..

16000+ posts does give you license to abuse others..... Every one here was a new registered user once so everyone has right to discuss something which is helpful to the community...

If this thread is not for tech discussion then sorry I am in a wrong place...... If anyone don't like this thread... Ignore it....

Final decision with mods.... If you feel this thread is useless and there is something intresting to discuss delete this thread.......

 

[lifeblood]

I think I will do something radical, I will actually talk about what this thread is supposed to be about.

If we assume Piledriver (PD) comes out in 8 months, what improvements can we expect based on process maturity alone? Higher speeds, certainly, but how how? 300mhz? 400mhz? IMO, that alone is not enough to make a significant difference. But what would? How much faster does BD have to be to reasonably compete with a i5 2500k in lightly threaded apps? 1 Ghz?

Now they know the limits of the process, what architectual changes can they make to improve performance? nVidia did a pretty good job revising Fermi after the disaster of their initial version. I would think AMD could achieve similar results with PD.

Personally I think they need to fix their cache latency. Its not quiet twice a s slow as Intels, but its close. Does the PD talk include any reference to that? Or are we stuck with that until their next major CPU overhaul?

 

[Genx87]

Anybody else get the idea AMD watched idiocracy and the 1 night of rehabilitation event?

 

[Nemesis 1]

Is it a crime to start a thread on a upcoming product ??????

If bulldozer failed does it mean that Pildriver will also fail....

Is it a crime to be a "new reg"????? If that so ban me and I will never come back on this forum..

16000+ posts does give you license to abuse others..... Every one here was a new registered user once so everyone has right to discuss something which is helpful to the community...

If this thread is not for tech discussion then sorry I am in a wrong place...... If anyone don't like this thread... Ignore it....

Final decision with mods.... If you feel this thread is useless and there is something intresting to discuss delete this thread.......

Like I said , This topic is a bad idea. Hyping an AMD product right now on this forum or any other outsided of zoners isn't wise at this time .

You have a very interesting writing form . I have seen that befor.

 

[Ajay]

Trailing Bit Manipulation (TBM):
TBM instructions provide instructions implementing individual bit 
manipulation operations such as isolating, masking, setting, resetting, 
complementing, and operations on trailing zeros and ones.

I can't find much more

But the Vishera CPUs will most likely have a lower clock that the Zambezi CPUs

As the Trinity APUs have a lower clock than the Llano APUs while scoring similarlyish(That is a feat)

Implying they increased instruction throughput somehow

Thanks NostrSeronx, a couple great posts! Where did you get this info (links?)?

If Vishera will have lower clocks and a 10% increase in performance, then the Piledriver cores must be pretty good - that fits in well with BD being targeted mainly at servers and BDII bringing desktop performance (single thread) improvements. If they overclock well, they could a decent budget enthusiast chip (like PHII was).

 

[lifeblood]

I thought I would play around with some numbers. According to Anandtechs review:
Cinabench 11.5 single threaded

2500k --------- 1.47
Ph 2 1100T --- 1.11
BD ------------ 1.02
​

That means the 2500k is about 31% faster. To equal it BD would need a 1.8GHz speed boost to 5.2/6.0 GHz in core/turbo. To equal last generations Phenom II 1100T single threaded performance they would need a clock boost of “only” 9% to about 4.0/4.6 GHz.

In Single Core AIDA64 Queens benchmark

2500k -------- 8141
Ph 2 1100T --- 6091
BD ----------- 5297
​

That means the 2500k is 35% faster. Again, BD still needs about a 1.8 GHz speed boost. To equal the 1100T is needs about a 14% boost to 4.1/4.8GHz.

Did AMD really expect GloFlo's 32nm process to allow a 5.2/6.0 GHz speed? If they had gotten it to the point where it equaled the 1100T (4.1/4.8GHz) in per core performance, would that be considered a win? I am talking end performance here which is IPC plus clock speed. Will process maturity allow 4.1/4.8GHz in six months? Its "only" another 0.5 GHz.

If they had gotten its speed up to 4.1/4.8GHz AND maintained the IPC of the 1100T, that would definitely put them just below the 2500K on single performance, but faster than it in multithreaded. I think that would have been a definite win.

So if AMD can get piledriver's IPC up to 1100T performance, and GloFlo can get their 32nm process capable of producing 4.1/4.8GHz BD CPU, then AMD will be able to claim parity with this gens Intel CPU. Of course, Intels next gen is right around the corner so...

 

[Gaz550]

Like I said , This topic is a bad idea. Hyping an AMD product right now on this forum or any other outsided of zoners isn't wise at this time .

You have a very interesting writing form . I have seen that befor.

Why is it bad idea? Is it a opinion or a fact? I think it's great idea (opinion). I found this forum very good source for BD information and I am hoping that it will bevery good source for piledriver information.

You have very interesting writing style too. But what that actually has to do with piledriver?

 

[exar333]

Like I said , This topic is a bad idea. Hyping an AMD product right now on this forum or any other outsided of zoners isn't wise at this time .

You have a very interesting writing form . I have seen that befor.

Interestingly enough, I actually took a quick trip to AMDZone a few days back. They are as down on BD as we are here (justifiably so). I have to admit I went there looking for crazy justification, and a little comedy, but the comments were pretty spot-on. I was somewhat pleasantly suprised.

The reason I bring this up is that the BD launch should cause a more muted response to the next release (PD) and it will be interesting to see how both AMD fans and the AMD marketing team react.

 

[Idontcare]

Did AMD really expect GloFlo's 32nm process to allow a 5.2/6.0 GHz speed? If they had gotten it to the point where it equaled the 1100T (4.1/4.8GHz) in per core performance, would that be considered a win? I am talking end performance here which is IPC plus clock speed. Will process maturity allow 4.1/4.8GHz in six months? Its "only" another 0.5 GHz.

If I could get one to analyze, to determine the static vs. dynamic power contributions (which would then speak to its limitations) then we'd be one step closer to having some insight into this.