Vishera Review Up - Anandtech

[Abwx]

Your way off dude... i had a i7 920 at 5ghz and i was touching high 7 scores in cinebench back in the Day..

Please quit trying to Hype a complete failure of another Amd Cpu.. Your still 3 Generations behind Intel and it is quite Sad..

The 8350 do about 7 at stock frequency , that is at 4Ghz
for multithreaded softs as CB....

Yeah , talk about living in the past....:biggrin:

 

[alexruiz]

TechReports overclocking results:

4.8GHz, 1.475V - reboot
4.7GHz, 1.4875V - lock
4.6GHz, 1.525V - errors on multiple threads
4.6GHz, 1.5375V - errors with temps ~55C
4.6GHZ, 1.5375V, Turbo fan - stable with temps ~53.5C, eventually locked
4.6GHZ, 1.5375V, manual fan, 100% duty cycle at 50C - lock
4.6GHZ, 1.55V, manual fan, 100% duty cycle at 50C - crashes, temps ~54.6C
4.4GHz, 1.55V - ok
4.5GHz, 1.55V - ok, ~57C, 305W
4.5GHz, 1.475V - errors
4.5GHz, 1.525V - errors
4.5GHz, 1.5375V - OK, ~56C

"There's a 48W gap between the TDP ratings of the Core i5 parts and the FX-8350, but in our tests, the actual difference at the wall socket between two similarly configured systems under load was over 100W. That gap is large enough to force the potential buyer to think deeply about the class of power supply, case, and CPU cooler he needs for his build. One could definitely get away with less expensive components for a Core i5 system."

Okay, so let's use your figure anyway. So the undervolted 8350 uses 18 watts more than the 3770K and is 25% slower in highly threaded applications. Is that more to your liking?
(We don't even have to talk about lightly threaded applications)

Other sites were able to get much closer to the 5GHz mark. Different binning, or the "witch hunter" didn't even try hard enough?

 

[Phynaz]

See by yourself , keep in mind the 400Mhz higher frequency of the 3770K
with highly threaded loads.

I will admit where I made an error. In some of the charts it's a 3960 on top, not the 3770. I didn't see that.

Now how about the rest of the review...Care to address that? Or don't I won't be here to read it, going to be away for a few days.

 

[Abwx]

Other sites were able to get much closer to the 5GHz mark. Different binning, or the "witch hunter" didn't even try hard enough?

Not very surprising for a first batch that was surely released in a hurry..

We ll see in three months with the wafers that will enter production
next month with the APM that will have integrated this product
manufacturing dispersions.

 

[Phynaz]

Other sites were able to get much closer to the 5GHz mark. Different binning, or the "witch hunter" didn't even try hard enough?

No idea. If somebody wants to use Techreport's results, then we use the results. I certainly didn't tell him to use Techreport as reference material, and he seems to have a distaste for Anand's benchmarking.

 

[hokies83]

Bout everything summed up here.

http://youtu.be/u7wFKewAbgs

This is the best part lol.. Click picture to go to video..



Then he goes on to say not only does the 3570k destroy the 8350 clock for clock but it does it and uses 27% less power while doing it...

 

[Abwx]

I will admit where I made an error. In some of the charts it's a 3960 on top, not the 3770. I didn't see that.

Now how about the rest of the review...Care to address that? Or don't I won't be here to read it, going to be away for a few days.

I was also caught off guard at first look , this forcibly happen when
you browse in several sites at the same time to check the results
consistency....

As for the rest of the review , all is said in the few pictures above ,
in the multithreaded workloads it do very well but the lightly threaded
workloads are still lacking considering the competition and unfortunately
for them the higher frequency do not help as much as they expected.

As a hope for their future it looks like BD scale almost linearly
with frequency but the current 32nm process is already stretched
close to its limits and i doubt that there will be CPUs with significantly
higher frequency , perhaps 4.2/4.5T at most with the last batches ,
with reduced power comsumption as some sites managed to reduce
TDP by 30W at full load just by undervolting , AMD being large with
voltages as a mean to qualify as much chips as possible.

They will have less difficulties in the server market for wich this chip
seems tailored from the start.

 

[hokies83]

I was also caught off guard at first look , this forcibly happen when
you browse in several sites at the same time to check the results
consistency....

As for the rest of the review , all is said in the few pictures above ,
in the multithreaded workloads it do very well but the lightly threaded
workloads are still lacking considering the competition and unfortunately
for them the higher frequency do not help as much as they expected.

As a hope for their future it looks like BD scale almost linearly
with frequency but the current 32nm process is already stretched
close to its limits and i doubt that there will be CPUs with significantly
higher frequency , perhaps 4.2/4.5T at most with the last batches ,
with reduced power comsumption as some sites managed to reduce
TDP by 30W at full load just by undervolting , AMD being large with
voltages as a mean to qualify as much chips as possible.

They will have less difficulties in the server market for wich this chip
seems tailored from the start.

I see a place for it in Linux / server world..

But as far as a Everyday desktop cpu is concerned you have to admit it is another...

 

[Abwx]

Bout everything summed up here.

http://youtu.be/u7wFKewAbgs

This is the best part lol.. Click picture to go to video..

Then he goes on to say not only does the 3570k destroy the 8350 clock for clock but it does it and uses 27% less power while doing it...

Lol , he has an ugly face , i understand better the psychism of PC adorators ,they compensate some lacks or frustration by projecting their PCs as being a part of themselves , hence their tendancy to refute anything that would put them under question , notably their PCs supposed greatness...

 

[ElFenix]

if much of what you're doing is video encoding (and you don't care about your electric bill) it looks like a nice processor. AMD desperately needs a die-shrink.

"There's a 48W gap between the TDP ratings of the Core i5 parts and the FX-8350, but in our tests, the actual difference at the wall socket between two similarly configured systems under load was over 100W. That gap is large enough to force the potential buyer to think deeply about the class of power supply, case, and CPU cooler he needs for his build. One could definitely get away with less expensive components for a Core i5 system."

yeah because a 650 watt power supply and hyper 212 cooler are SOOOOOO much more expensive than what commonly goes into 3570 boxes.

 

[oceanside]

Bout everything summed up here.

http://youtu.be/u7wFKewAbgs

This is the best part lol.. Click picture to go to video..

Then he goes on to say not only does the 3570k destroy the 8350 clock for clock but it does it and uses 27% less power while doing it...

What he said in 38 minutes could easily be communicated in 3 minutes or less by your average human being. Is he getting paid by bandwidth?

 

[hokies83]

What he said in 38 minutes could easily be communicated in 3 minutes or less by your average human being. Is he getting paid by bandwidth?

Why i made the 1 min video lol He likes to talk.. His Cosmos 2 review is like an Hour long..

 

[AtenRa]

You make no sense 8350 is equal in performance to a i7 950 clock for clock that is about it...

And Sandybridge over Lynnfield was 10% faster on average clock to clock, so what ???

FX8350 obliterates the Core i7 950 in every metric, FX8350 is even faster in many multithreaded applications even than Sandybridge 2600K (both at 32nm).

 

[AtenRa]

Then he goes on to say not only does the 3570k destroy the 8350 clock for clock but it does it and uses 27% less power while doing it...

From his review

http://www.overclock3d.net/reviews/cpu_mainboard/amd_vishera_fx8350_piledriver_review/9

Cinebench 11.5

FX8350 @ 4.8GHz = 8,28
Core i5 3750K @ 4.8GHz = 8,02

POV-Ray

FX8350 @ 4.8GHz = 1702
Core i5 3750K @ 4.8GHz = 1492

Now, Is he or is he not BIASED ????
There is no hope for the 3750K, the FX8350 is faster when it comes to Multithreaded apps.

 

[inf64]

Yes Atenra,but they don't know better therefore


Another review I've found ,this time from Russia(it has 4.7Ghz BD vs 4.7Ghz PD numbers):
http://www.overclockers.ru/lab/50228/Obzor_i_testirovanie_processora_AMD_FX-8350_Vishera.html

 

[Olikan]

Another review I've found ,this time from Russia(it has 4.7Ghz BD vs 4.7Ghz PD numbers):
http://www.overclockers.ru/lab/50228/Obzor_i_testirovanie_processora_AMD_FX-8350_Vishera.html

WHOOT... nice, find

didn't done the math, but seems more than 20% ipc/scaling D:

 

[know of fence]

The trolling in this topic is insufferable, posting old internet memes and rambling, sweatpants wearing chavs now?

That Russian website though has some very clever mouse-over bar charts. The i7@3.4Ghz(no HT) appears to represent the i5-2570K price competitor.

 

[MisterMac]

Get over it - AMD improved PD alot more than people expected.

And priced it accordingly - it's a nice budget offering that despite having 8 cores - will only last until the day the intel mainstream segment decides it's 6\8 core time.

That or WQHD becomes a standard along with some cray cray performance increases on the GPU side requiring more CPU juice.

 

[hokies83]

And Sandybridge over Lynnfield was 10% faster on average clock to clock, so what ???

FX8350 obliterates the Core i7 950 in every metric, FX8350 is even faster in many multithreaded applications even than Sandybridge 2600K (both at 32nm).

Yes it beats the i7 950 @ 3ghz you did not notice me saying clock for clock? that means 4ghz i7 950 vs 4ghz 8350?

Wow u point out a cpl of token aps.. BD could do that to .. Heck i could post afew token aps of a duel core beating a 2600k to...

The review sums it up.. quit straying from the truth..

The 8350 is a True failure just like the 8150 is..

And is a nice little upgrade for somebody with AM3+

Anyone building there own NEW System would have to out of there mind to go with Amd right now

Which is, at the end of it all, about the size of it. We used the very best possible setup to give the FX8350 every chance of proving its worth. If you've already got an ageing AM3 setup, then the FX8350 is the best option available as a decent upgrade for a decent price. If you are looking to upgrade a full system then it's impossible to recommend. It's too slow, it draws too much power, it's too hot. It's just not worth it.
We so wanted this to be a return to form for AMD. This is the best they have to offer, and they are still a mile behind the competition. If you've already got an AMD setup and can only afford a CPU upgrade then this is just worthy of our OC3D Bronze award. For everybody else, steer well clear. For AMD, start with a fresh sheet of paper.

And if you say anything Different then you are just an AMD fan boy and Out of touch with "reality"
http://www.youtube.com/watch?v=u7wFKewAbgs&feature=youtu.be

 

[Rvenger]

Yes it beats the i7 950 @ 3ghz you did not notice me saying clock for clock? that means 4ghz i7 950 vs 4ghz 8350?

Wow u point out a cpl of token aps.. BD could do that to .. Heck i could post afew token aps of a duel core beating a 2600k to...

The review sums it up.. quit straying from the truth..

The 8350 is a True failure just like the 8150 is..

And is a nice little upgrade for somebody with AM3+

Anyone building there own NEW System would have to out of there mind to go with Amd right now

Which is, at the end of it all, about the size of it. We used the very best possible setup to give the FX8350 every chance of proving its worth. If you've already got an ageing AM3 setup, then the FX8350 is the best option available as a decent upgrade for a decent price. If you are looking to upgrade a full system then it's impossible to recommend. It's too slow, it draws too much power, it's too hot. It's just not worth it.
We so wanted this to be a return to form for AMD. This is the best they have to offer, and they are still a mile behind the competition. If you've already got an AMD setup and can only afford a CPU upgrade then this is just worthy of our OC3D Bronze award. For everybody else, steer well clear. For AMD, start with a fresh sheet of paper.

And if you say anything Different then you are just an AMD fan boy and Out of touch with "reality"
http://www.youtube.com/watch?v=u7wFKewAbgs&feature=youtu.be

Spreading around AMD fail contradicts how you are behaving in VC&G, let me quote you.

If you say so..

I still do not see how sharing updated Driver information is Marketing..

When there is like some Amd driver thread here every week or so..

You do not see me trolling in there.. I could careless what Amd is doing.... there for i do not even belong in the Amd driver thread nor do i want to be.

 

[Idontcare]

I like what I see in the 8350.

If being the best was all that mattered when it came to making money and selling CPUs then Intel would only have one SKU and just sell the best of the best CPU, but the fact that they have an entire portfolio of SKU's - and still actively produce/sell 32nm SKUs alongside the 22nm ones - is proof-of-the-pudding-is-in-the-eating type evidence that the TAM is large enough for an entire buffet of chips to be sold into.

I'm seriously considering buying an 8350 to use as my main work machine and repurpose my 3770k as another node in the farm.

I don't do strenuous gaming (dragon age of origins is the most hardware-challenging game I play) but I do a fair amount of transcoding of DVD's and home video clips for the HTPC, combined with the forex algorithm prototyping.

First I need to find the ~$400 required to get into a piledriver setup while avoiding the craptastic mobos that are out there. I like my MIVEZ, awesome investigatory tool for analyzing the finer qualities of the underlying CPU's silicon IMO. Not sure if there is an equivalently high quality mobo made for the AM3+ market.

Also, has anyone determined if the 8350 can be delidded? And has anyone done any indepth temperature-dependence studies to see how much of piledrivers high power usage when overclocking is because of skyrocketing temperatures?

 

[NTMBK]

Also, has anyone determined if the 8350 can be delidded? And has anyone done any indepth temperature-dependence studies to see how much of piledrivers high power usage when overclocking is because of skyrocketing temperatures?

Nah, we're all waiting for you to do it IDC

 

[Idontcare]

Nah, we're all waiting for you to do it IDC

I'd do it today if I had one in my hand.

I posted the following info in another thread, even with its uber-3D xtors there is a huge impact in power-consumption and required operating voltage as a function of temperature for Ivy Bridge.

^ all told, by reducing the temperature I was able to reduce the voltage needed for stable operation by some 5%, netting a 21% reduction in power consumption by the CPU all because I reduced the operating temperature by some 40°C through a combination of replacing the stock HSF with a lapped H100 and by delidding the IB chip.

This makes me wonder if Piledriver offers the same opportunity to enthusiasts if we delid it and get aggressive with the cooling and voltage optimization.

These reports of 400+W for 5GHz piledrivers may be true but entirely needless if the enthusiast puts in a little elbow grease like we do with our Ivy bridge chips.

 

[hokies83]

Spreading around AMD fail contradicts how you are behaving in VC&G, let me quote you.

Comparing Intel to Amd...
Amd fails hard Almost all the reviewers are saying it.

This has nothing to do with Me being a Galaxy Rep With Nvidia.. this is CPU'S
Not GPU'S.

There for i will not stick my Nose in an AMD Gpu Thread...

But Cpu is wide open.

 

[pelov]

Comparing Intel to Amd...
Amd fails hard Almost all the reviewers are saying it.

This has nothing to do with Me being a Galaxy Rep With Nvidia.. this is CPU'S
Not GPU'S.

There for i will not stick my Nose in an AMD Gpu Thread...

But Cpu is wide open.

True, but it also makes it harder to believe you won't provide a heavily biased opinion :awe:

It would be like asking Chris, our resident Intel rep, what he'd say regarding the discrete GPU market. At least from Chris I'd expect an honest and unbiased answer even if I did ask him, but you...

well...

^ all told, by reducing the temperature I was able to reduce the voltage needed for stable operation by some 5%, netting a 21% reduction in power consumption by the CPU all because I reduced the operating temperature by some 40°C through a combination of replacing the stock HSF with a lapped H100 and by delidding the IB chip.

This makes me wonder if Piledriver offers the same opportunity to enthusiasts if we delid it and get aggressive with the cooling and voltage optimization.

These reports of 400+W for 5GHz piledrivers may be true but entirely needless if the enthusiast puts in a little elbow grease like we do with our Ivy bridge chips.

Judging from the reviews I've read, they need some very aggressive cooling, and even with it the chips hover around that 70C mark at a relatively modest overclock of ~4.6ghz.

The temperature kicks up quite quickly and it wouldn't surprise me if it were from the lack of some dark silicon. Maybe AMD should make some 3-module APUs

-- The Trinity chips might even fair better with some heavy overclocking, so long as that on-die GPU is flipped off.