AMD Piledriver rumours ... and expert conjecture

Page 166 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.
We have had several requests for a sticky on AMD's yet to be released Piledriver architecture ... so here it is.

I want to make a few things clear though.

Post a question relevant to the topic, or information about the topic, or it will be deleted.

Post any negative personal comments about another user ... and they will be deleted.

Post flame baiting comments about the blue, red and green team and they will be deleted.

Enjoy ...
 
Well, there are a lot of 'problems' that reduced BD's performance but wouldn't be that difficult to fix. It seems that AMD was able to reign in power consumption a bit, so we should see higher clocks and better efficiency. If that is all they did though, I would be pretty disappointed. Hopefully they had time to tweak the scheduler/cache mess so that it can actually feed all of those execution units. Sandy/Ivy will be very hard to beat at serial or lightly parallel tasks, however I hope that the new (8 "core") chip will finally beat the PII X6 while consuming less power, something BD had trouble doing.
 
Piledriver ~ Trinity A10 is looking good with 384 igp cores with 50% more accelerated opps
but Steamroller and it's 512 GCN cores sounds like a super coputer
by the time excavator shows up it's 7960D igp construction complete
the age of super computers has arrived.

I admire your optimism.

With Trinities welcome IPC increase and superior performance per watt in the mobile space, not to mention far superior graphics, at least they have something to grow their business in that area.

With SBE's superior IPC and headroom, and IVB's superior performance per watt and impending movement into the mobile space, there is plenty going on.


I'd like to finish by saying attacks against other users will not be tolerated ... please focus on the issue, situation, or behaviour that is presented ... not on the person.

Attacks against a user will be dealt with accordingly ... attacks where a number of users "join in" will result in aforsaid receiving same punishement the original offender receives.

If other users (in the last three pages) feel they want to edit their posts, then please do so ... it might be a good idea after my cursory glance today.

I'll check back later ...

:)
 
www.fudzilla.com/home/item/27235-amd-says-“enough”
AMD says “enough”
"AMD claims that users have all the processing power they need and if they want more then let them eat cake."

Reminds me of a long time ago when a certain CEO said most people wouldn't need more then 4GB of RAM (implying they also wouldn't need 64 bit CPU's) until 2007. That went over like a lead balloon on here. People do not like to be told what they should or shouldn't want.

If fudzilla's article is anywhere near correct, it means AMD won't be focusing on CPU enhancements in the near future - at least not for the laptop market.
 
www.fudzilla.com/home/item/27235-amd-says-“enough”
AMD says “enough”

Meh. Makes sense to me.

If you want/need a processing monster, buy Intel. If you fall anywhere even close to the realm of "average user," what more could you possibly need right now?

That said, I do agree that it is a silly move PR-wise.
 
I would ask the users in this thread to please refrain from any personal attacks. Any negative comment, direct or indirect targeting another user is in violation of the rules of conduct. Any breach of civility in this lively thread will be met with extreme sanction. I appreciate your cooperation.
 
www.fudzilla.com/home/item/27235-amd-says-“enough”
AMD says “enough”

Hahaha, this was taken out of context so bad...

This is just the confirmation of Mr. Read saying a while back "we will get out of the performance race". That's exactly what AMD did with Trinity. They only need improvements on their BD arch to get a "good enough" product and get OEMs. Notice Intel actually followed AMD in that regard. Why would Intel want to create QS and put a beefier GPU in ther CPUs? AMD was right in that call and Intel recognized it. This is in a business sense only, off course. Intel still has the upper hand in budget and process, like I've always said, they can turn the tables in no time (and they actually are with IB, somewhat).

Now, there's just a little contradiction in there. nVidia has developed that Cloud GPU thingy that kind of defies what AMD and Intel are doing today. So it would kind of triumph the GPU (or APU) race in a few years if it becomes successful; which I doubt BTW 😛

Anyway, current consensus is that PD won't be a miracle for AMD, just an improvement to get them closer to SB and IB and at least, no surprises there with the comment from Mr. Read. Remember that we won't need something that skyrockets performance against BD, just and improvement and we'll be happy campers; at least, I would if it's an improvement over PhII (same IPC, higher clocks are good for me -> upgrade).

Cheers!
 
www.fudzilla.com/home/item/27235-amd-says-“enough”
AMD says “enough”
Spudzilla has always been anti-AMD. I am sure this entire article was taken out of context and the meaning changed just so they can drum up some traffic. Looks like it was probably taken from a speech talking about thier future cloud based computing solutions.

http://sites.amd.com/us/business/it-solutions/cloud-computing/Pages/cloud-computing.aspx

If you take him saying that in a speech about cloud based computing, then absolutely there is enough cpu power in todays laptops (for cloud computing).

But outright saying there is enough cpu power in todays laptops.

2 entirely different meanings for 2 entirely different markets.

 
I'd like to observe a little prediction pole. If Sandybridge is a 10 and Bulldozer is a 1, where on the performance scale to you predict Piledriver may fall?
I think, overall, it would fall around 8. There is showing of an IPC increase over BD and K10.5. What is most promising about Trinity's cpu side is it's (IMO) absolute whooping of Llano's cpu. At the same time, it has less transistors, and uses less power. Performance/watt is greatly improved, and that is what matters.
 
I'd like to observe a little prediction pole. If Sandybridge is a 10 and Bulldozer is a 1, where on the performance scale to you predict Piledriver may fall?

The translation being: What percentage of the performance gap do you expect PD to close?

Because SW scaling will remain a problem, the hit you take with CMT cores, and the fact I don't expect much more then 20% IPC improvements [absolute best case average], I have to go with no more then a 4. PD will close the gap 40% on SB compared to BD.

And note, I am being VERY optimistic on the 4.
 
Assuming 1 is the worst and 10 is the best, and this is single core IPC, I'll put PD at a 2. That's about a 5% IPC improvement.

Efficiency will be better compared to BD, which could mean higher clocks, perhaps the 4.2ghz they tried to reach the first time. I don't expect massive overclocks though.
 
6. GPU compute context switch and GPU graphics pre-emption: GPU tasks can be context switched, making the GPU in the APU a multi-tasker. Context switching means faster application, graphics and compute interoperation.

That is incredibly important, most likely the single biggest step forward (for GPU) since HW assisted Rasterization.

Finally, AMD has stated from the beginning that our intention is to make HSA an open standard, and we have been working with several industry partners who share our vision for the industry and share our commitment to making this easy form of heterogeneous computing become prevalent in the industry.

Even more important, if it's Open and easy to implement then it will be implemented. HW does not drive SW, and SW isn't driven by "the best". SW is driven by how easy it is for developers to achieve their goal / create their products. Make it easy for the key pounders and they'll use it.


That is about far more then how many frames a CPU can render a motorbike at, or how many minutes it takes to render grandmothers birthday video. This is a fundamental shift in the way software is coded and hardware is utilized. Turning the GPU as an open ended vector co-processor that shares address space with the CPU and can interact on the fly is just .. wow. This means you can interact with both the CPU and GPU in the same machine level code stream, no need to context switch or segregate your code into "GPU land" and "CPU land". Create a construct in memory, do ~stuff~ to it, then reference that same construct in a GPU opcode without having to first re-create that construct in GPU memory.
 
www.fudzilla.com/home/item/27235-amd-says-“enough”
AMD says “enough”

To be fair, Read thinks that AMD needs to move to a more graphical focus

1. Over pay for ATI. Check
2. Kill ATI name, keep Radeon. Check
3. Realease CPUs that are about 1-2 gens behind competitions. Check
4. Kick CEO as company looses tons of money, gain new CEO from ATI purchse. Check
5. Can that CEO for not moving into UMD market fast enough while he tries to patch the sinking ship, hire new CEO from laptop company. Check
6. New CEO states no more trying to compete with Intel, then states users have enough. Check
7. Move away from CPUs and focus on GPUs, rename company AGD (Advanced Graphical Devices).

This is just another reason why I am not a fan of the new CEO. Instead of doing as Dirk Meyer and fixing the leaks and then focusing on the most profitable areas (server/HPC) this guy wants to not try and compete with Intel and instead act as if everything is fine and dandy and now Intel wont have the motivation to push harder.

I say screw the cloud. I want my power local, not controlled by some company. This is sad if its a legit quote....

New Hope = K8
Empire = Conroe
Jedi = ????

You forgot Phantom Menace = Bulldozer. 😛
 
The translation being: What percentage of the performance gap do you expect PD to close?

Because SW scaling will remain a problem, the hit you take with CMT cores, and the fact I don't expect much more then 20% IPC improvements [absolute best case average], I have to go with no more then a 4. PD will close the gap 40% on SB compared to BD.

And note, I am being VERY optimistic on the 4.


Looking at the benchmarks of several different sites. I come to this conclusion.

2600K 10
2500K 9.5
8150fx 6.5
1100T 6.3
980 6.0

The 1100T and the 8150fx can compete head on or slightly beat the 2500K on apps that use 6 or more cores but falls short when a App uses less and even loses on Resident evil 5 that can use all 6 cores.

Now per core per clock i'm still going to say a BD core is only half as strong as a Intel core(ivy).


In the next couple of years This design will most likely look better, No one not even huge Amd haters can deny that.

Side note

About 2 weeks before Bulldozer released i knew something wasn't right when all rumors on the processor were bad, The fanboy in me kept saying "its Amd they have great engineers just bad marketing" But i started to wonder and i said to myself "as long as its at least 10% faster per core and faster then the 2600K in multithreaded apps i'll buy it", Well let me just say i'm not letting this happen again with Piledriver! Not even going to save the money until it comes out and has full reviews(not amd only)

But something tells me i'll be happy with Piledriver! If not i think this is the end with me and Amd when it comes to their CPU.(Not APU's or GPU's)

I'm sure i'm not the only one who feels this way Amd will get one more chance with me unless i'm jumping to Intel and that makes me SICK to my stomach and also a little bit teary.
 
I'm sure i'm not the only one who feels this way Amd will get one more chance with me unless i'm jumping to Intel and that makes me SICK to my stomach and also a little bit teary.

Over the years I have swapped from ATI to Nvidia, to ATI, from Intel to AMD to Intel, in electrical goods I have swapped from Panasonic to Sony to Panasonic, in cameras from Canon to Panasonic to Nikon, and throughout it all, not a single tear shed.

Why do you feel teary?
 
Status
Not open for further replies.