AMD CPU speculation... and expert conjecture

Page 398 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
AMD owns several world-wide OC records thanks to SOI, but that is under extreme cooling. On air, the average OC of an i5-3570k is 4663MHz, which is slightly higher than average OC of an FX-6350: 4645MHz. On watter the tendecy inverts and FX achieves slightly better frequencies than the i5: 5063MHz (FX) vs 4819MHz (i5). In any case these are differences of 3-5%. Thus for average overclockers, bulk is not massively limiting anything.

Kaveri will not break world-records, but it will have good OC.
 


Oh not denying that, actually most people can get a 8350fx at 4.6-4.8Ghz on air and a I5 haswell at 4.4(Usually)
 


Were are you getting that information?
 
http://www.xbitlabs.com/news/cpu/display/20131209231050_AMD_We_Are_Not_EOL_ing_FX_Line_of_Microprocessors.html

"AMD will continue to supply AM3+ and AMD FX processors for the foreseeable future, as per AMD's official roadmap update at APU'13 [above]. Recently, AMD launched the FX-9000 series, AMD's fastest desktop processor to date. As AMD's business continues to evolve, AMD will focus on the areas of growth including support for the desktop PC enthusiast leveraging AMD's world-class processor design IP, including heterogeneous compute. AMD's FX branded products will continue to evolve and we look forward to sharing those updates in the future," said James Prior, an AMD manager of APU/CPU product reviews

But, But the APU. Again listen to noob222 and Me when we say we know people in Amd who say just that FX lady's and gentlemen ain't going nowhere 😉
 


Ah, but remember the BASE clocks:

FX8350: 4Ghz
i7-4770k: 3.5GHz

So Intel actually gets a little bit more OC ABOVE its base clock (~900MHz average improvement versus ~600-800MHz), so at max OC, a typical i7-4770k versus a typical 8350 will do better then a 4770k versus a 8350 at stock.

Point being, the whole "overclockability" myth of AMD is just marketing; Intel chips OC better. And if Intel cared about making its chips run at below 0C temps, I'm sure OC records using LN2 would show that as well.
 


There is a difference between continuing to sell FX chips, and making significant upgrades to the design.

Seriously people, read between the lines and read through the marketing. Its not that hard to do.
 


Intel bulk != GloFo bulk.

I remain very pessimistic over GloFo 28nm bulk.

I'm expecting Kaveri to top out around 4.3ghz to 4.6ghz on air and 4.8ghz on water.

Which might seem fine if you compare it to 4m/8c SOI parts, but it won't look so peachy if you're comparing it to Richland overclocking.

Hawaii didn't show up on any roadmaps or anything either. In fact, we were looking at "stable throughout 2013" for 7000 series.

It's hard to tell if Hawaii was a fluke or not, but perhaps we're seeing AMD with a new tactic of not letting people know about their products until they're about to be released? It works pretty darn well for Apple after all.

There are some good rumors if you dig deep enough that haven't reached the rumormill websites yet that should make one feel optimistic. But if you think about why HSA won't work on AM3+ or a generic platform update with PCIe 3.0 it'll make a little more sense. But if AMD were to do something radical like that for a gaming platform, they wouldn't want to be talking about it just yet. Catch my drift?

APU solves the problem of HSA having to have high bandwidth communication between CPU and GPU, but it's only short term because it's not viable on higher end platforms. Whether you think AMD wants to solve the problem of fast interconnects between GPUs and CPUs depends on if you think AMD is going to be content with mobile, ARM, and low-end gaming PCs or if they want to be viable in server, workstation, and high end gaming PCs again.
 
+1 on gf vs intel silicon. Amd is usually very aggressive when it comes to base clocks and kaveri tops out = intel.

Id be suprised if kaveri overclocks over 4.4 ghz, putting gf's 28nm roughly the same as intel's 22nm.

Smaller nodes are not good for thermal density.
 
you guys, blackkstar and others, should stop worrying so much about kaveri's overclockability. the cpu would have around 50% of the die area built on a smaller node and possibly with more transistors. there will also be the integrated u.n.b. on the die. all these already limit o.c. potential on the apus. the lower binned cpus (sans igpu) would be better for o.c. but amd always releases those near the last part of the cycle.
what i'm trying to say is kaveri as it is right now is not a total deal breaker as long as amd properly implements turbo and thermal monitoring. also if the apu doesn't throttle when stressed at stock. turbo is the most important factor here, as amd'd previous turbo implementations were garbage (in reality. on paper it is superior (the irony!!)), imo. auto-overclocking is factory guaranteed unlike manual o.c.
 
Probably a dumb question that has already been touched on but did anyone else see that while we were all busy arguing about ARM and roadmaps VIA had joined the HSA foundation? Does this mean they have intentions of producing their own APUs for industrial applications?
 
Amd's turbo is the same if not better than Intel's since the FX series. Yes phenom turbo was trash.

http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/4

single core, the 8150 at least tried to get 4.2 ghz where the 2500k never even hit 3.7 ghz.

The advantage intel has is the 2500k never ran at its base clock, aka the fake base clock intel claims gives intel an" ipc" boost.

AMD could achieve the same results by introducing more turbo states, but that requires more transistors and the advantages Would be minimal (0.5% according to AT).
 

Piledriver v Haswell and even Ivy, AMD DESTROYS Intel in OC headroom.. A H60 can take a 4670K to 4.3, a 8350 4.6... Hell, an H100 or D14 can get em up to 5 pretty easily..
 
jdwii, the words of the PR James Prior have been cited now three times in this thread. My response is the same than in the two previous occasions.

blackstar, I was answering a very specific question made by etayorius about OC capabilities of Intel chips.

I cannot comment on AMD tactic, because I would ruin it ;-)

Regarding APUs, sorry but it is just the contrary of what you think: the ultra-high-end platform is APU only. There is no possibility to reproduce performance at that scale using the old CPU + discrete GPU combo. The laws of physics are very clear. That is why Nvidia, Intel, and AMD are developing their fastest designs around APUs. The future APU that AMD has in mind has about 10TFLOP of performance and 200-250W. Nvidia own design gives more than 20TFLOP and is rated at about 300W. Future supercomputers will use only APUs. I already commented about this twice, no sure why I spend time repeating stuff.
 


Add that to your prediction that Kaveri will be a 37% slower than Bulldozer and Kaveri will be the worst launch in the history. Have you planned more FUD for this week?
 
Juaranga, we shall see in about a year or so. I am just pretty confident in how I feel about AMD's plans because I at the very least, have some shady evidence to back it up.

The profit margins for AMD are too massive to pass up in professional workstations. A platform which revolves around 4m/8c Opteron at $400 and FirePro at $1000+ with HSA would destroy anything Intel could release, no matter how many cores Intel throws on a die or MCM.

http://www-03.ibm.com/press/us/en/pressrelease/41684.wss

I guess we will see if IBM and Nvidia show up with a Power APU or if they get it working across dCPU and dGPU. From what I'm getting out of this is that they want to basically copy HSA, but instead of AMD CPU + AMD GPU, it'd be IBM CPU + Nvidia GPU. Correct me if I'm wrong, but it looks to me like IBM and Nvidia want to go after exactly what I'm describing (as well as HPC) in the same exact way I'm talking about it. So at this point the only one really left out of this type of HSA ecosystem is Intel, who at best as Xeon which is a system on a board (with full Linux OS).
 
Seems to me that all cpu's today are above and beyond more than enough
it is the igp that seems to be the next big thing
with intel frantically playing catch up in that department and Haswell being delayed
The steamroller APU's seems to be the ticket
Especially in mobile and midrange systems

especially now that the economic situation does not favor high end machines and the upgrade industry is feeling the pinch.


DO you think HSA will eventually become a big deal? and if so would that put Intel a decade behind or more
 


Not really... What Intel is doing is making bad calls on OTHER things but the way they're shrinking the transistors. Their BULK is state of the art IMO and GF is just getting into their footsteps.

This is some proof:

cpuid_i7_4.6ghz.jpg


I don't recall having 1.4v at 4.4Ghz, but I haven't had the time to fine tune the OC (fat lies again; just laziness).

Anyway, point is credits due where credits owed: Intel's BULK is top of the crop. I'm really skeptical GF in BULK will reach Intel's 32nm in 28nm even. Much less 22nm. I'd even say TSMC's 28nm will be better than GFs (if I got all the BULK thing for Kaveri right, lol).

Cheers!

EDIT: Right... I'm on Air.
 
@juan

I said your stupid cosmology ES benchmark was 37% slower if in fact that was kaveri.

Are you still mad about that because you actually think that ES is kaveri?

200-250w apu.... rofl. What is the cooling for that, ln2?

@yuka

Id suspect that even UMC is more refined than GF at this point.
 


You still fail to tell me how come your saying 17% on average instead of 30%
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom