Vishera Review Up - Anandtech

[SlowSpyder]

In all honesty I do not think the 8350 (or PD in general) is as good as some posters want to make it out to be. But even more so I think it isn't as bad as others want to make it out to be. The 6300 hex core chip, assuming it can clock well, could be a $130 monster.

 

[AtenRa]

You should know that your manipulation doesnt work on me. For gaming only BF3 (MP64) works too, right? ()

And basicly if you want to render, encoding or compress data. Thats where the FX8350 shines? Got it. Just a shame it will suck is basicly everything else you do. Now I dont know about you, but applications tends to be much more than those few cases you like to cherry pick.

Even this is a bad list due to the nature of wanting to benchmark:
http://www.anandtech.com/bench/Product/701?vs=697

But really...FX is just a joke for the wast majority. No wonder why AMD is giving up.

The frustration is evident in you little one, FX8350 is faster than 3570K. People don't buy multi-core CPUs just for single threaded applications.

And up until now AT benchmarks suite was fine, now that an AMD product is faster than your little Intel CPU is a bad list.

 

[Yuriman]

The frustration is evident in you little one, FX8350 is faster than 3570K. People don't buy multi-core CPUs just for single threaded applications.

And up until now AT benchmarks suite was fine, now that an AMD product is faster than your little Intel CPU is a bad list.

It's a respectable chip in many ways. Faster in some games, slower in others. It is unfortunately much slower in most of the games I play, but I imagine future games will fare better on AMD's architecture.

 

[hokies83]

The frustration is evident in you little one, FX8350 is faster than 3570K. People don't buy multi-core CPUs just for single threaded applications.

And up until now AT benchmarks suite was fine, now that an AMD product is faster than your little Intel CPU is a bad list.

LoL there you go Again.

The 8350 is faster then a i5 3570k in Heavy multi threaded Aps that is it..

This is a 3570 K were talking about here not a *3570*

That Stock 3.4ghz means NOTING Once you Over clock that 3570 K It destroys the 8350. Every body knows this And does it while using 27% less power. You know it to.. you just Love your AMD hey man good for you I love there APU'S there great But there Cpu's are not anywhere in the same class as Intel's latest.

 

[Vesku]

LoL there you go Again.

The 8350 is faster then a i5 3570k in Heavy multi threaded Aps that is it..

This is a 3570 K were talking about here not a *3570*

That Stock 3.4ghz means NOTING Once you Over clock that 3570 K It destroys the 8350. Every body knows this And does it while using 27% less power. You know it to.. you just Love your AMD hey man good for you I love there APU'S there great But there Cpu's are not anywhere in the same class as Intel's latest.

It destroys it in MT too? The 3770K does better at max OC vs max OC MT but I didn't see the 3570K doing so when I looked at reviews. Maximum fps in games isn't everyones goal for highend desktops. I thought the original quads embarassing the all you need is an OCed core 2 duo folks, now that time has clearly delivered for early quad adopters, would kill this kind of rhetoric. Seems it has not.

 

[AtenRa]

And you are not biased????

I have never linked a benchmark showing an AMD CPU being slower than an Intel CPU and said that the Intel CPU is slower, sucks, its a joke and more.Thats what the reviewer did in his review.
Actually, you will never see me use those words for any AMD, Intel or NVIDIA product.

When Intel is faster its faster, the same also applies to when AMD is faster and in the majority of the reviews then FX8350 is faster than the 3570K. Why people cant accept that is another story.

I have never seen a poster try as hard as you to find a reason to recommend AMD.

Have you ever wondered that this could just be because those AMD CPUs are more competitive than YOU believe ?? I cannot recommend an AMD CPU against a Core i7 39xx but i can sure recommend any AMD CPU at $200 mark and below.

Also, depending on the users needs, I can recommend an Intel CPU with the same ease as I do with AMDs CPUs. Not many do that here including you

 

[Vesku]

I could see recommending these x3xx series FX chips given the right circumstances. Not something I could say about the x1xx series.

 

[hokies83]

I could see recommending these x3xx series FX chips given the right circumstances. Not something I could say about the x1xx series.

Yes if your running Heavy Multi Threaded Apps on a very Tight Budget then go for it.. other then that look at Intel.

 

[Vesku]

Yes if your running Heavy Multi Threaded Apps on a very Tight Budget then go for it.. other then that look at Intel.

Or if the extra power draw isn't a concern then the FX 6300/8320 are very good at their price points. Unless you're going to now go on a philosophical journey trying to figure out why not everyone in the world is running 3770Ks or 3930Ks?

 

[hokies83]

Or if the extra power draw isn't a concern then the FX 6300/8320 are very good at their price points. Unless you're going to now go on a philosophical journey trying to figure out why not everyone in the world is running 3770Ks or 3930Ks?

My Htpc is a i5 760... all it does is stream Mickey Mouse club House on my 50 Inch Samsung TV... for my 3 yearold.

But it is Equal in Gaming performance with Amd's FlagShip Cpu...

I paid 155$ for Cpu Mb And ram.. Now that is gaming performance on a Budget lol.. But meh other then Mickey Mouse it gets no other uses @_@

Ive just got a Amd Apu Tablet.. and see if it can Replace it.. so i can sells it. :thumbsup:

 

[Vesku]

My Htpc is a i5 760... all it does is stream Mickey Mouse club House on my 50 Inch Samsung TV... for my 3 yearold.

But it is Equal in Gaming performance with Amd's FlagShip Cpu...

I paid 155$ for Cpu Mb And ram.. Now that is gaming performance on a Budget lol.. But meh other then Mickey Mouse it gets no other uses @_@

Ive just got a Amd Apu Tablet.. and see if it can Replace it.. so i can sells it. :thumbsup:

Oh, now we have used (or an amazing MC/Fry's once a year) pricing thrown in. $150 buys a retail i3-3240 no MB no RAM. You're starting to derail the thread at this point, fyi.

 

[hokies83]

Oh, now we have used (or an amazing MC/Fry's once a year) pricing thrown in. $150 buys a retail i3-3240 no MB no RAM. You're starting to derail the thread at this point, fyi.

HuH? i just said my i5 760 is equal in gaming performance to a 8350..

I bought used off OCN about 7 months ago.

It is good Budget gamer performance for how cheap it is lol.

 

[Vesku]

The value of purchasing used is not at all related to the topic. Here is how much of your post was on topic:

"
My Htpc is a i5 760...

But it is Equal in Gaming performance with Amd's FlagShip Cpu...
"

Which is not much of a post, imo.

 

[hokies83]

The value of purchasing used is not at all related to the topic. Here is how much of your post was on topic:

"
My Htpc is a i5 760...

But it is Equal in Gaming performance with Amd's FlagShip Cpu...
"

Which is not much of a post, imo.

General Chatter not every post has to be a Tech related one as long as it has something to do with the Topic.

 

[Vesku]

General Chatter not every post has to be a Tech related one as long as it has something to do with the Topic.

Oh hey I have an AM3+ board but I'm running a 1090T in it. Vishera is AM3+ compatible.

I was bored and browsing through Netflix and I stumbled upon a sitcom called "Raising Hope". It was actually pretty good for a sitcom and thankfully didn't use the crutch that is laugh tracking. After watching a few episodes I decided to check on my home server box which is running a Pentium G620 on a Tyan s5510gm3nr which let's me run it with ECC memory. I got the motherboard open box from Newegg for 50% off MSRP. The whole server cost about what a barebones atom server box goes for.

I can see no potential for this post to spin this thread woefully off-topic.

 

[ShintaiDK]

The frustration is evident in you little one, FX8350 is faster than 3570K. People don't buy multi-core CPUs just for single threaded applications.

And up until now AT benchmarks suite was fine, now that an AMD product is faster than your little Intel CPU is a bad list.

The AT bench list pretty much shows your FX8350 to be inferiour. Maybe you should check again instead of your usual cherrypicking to try and manipulate people to buy from your precious company.

But hey, FX8350, when utilities are free and you dont mind noise and only do rendering/encoding/compression all day

 

[AtenRa]

The AT bench list pretty much shows your FX8350 to be inferiour. Maybe you should check again instead of your usual cherrypicking to try and manipulate people to buy from your precious company.

Lets take every real life application from AT bench list and see what the final score will be,

http://www.anandtech.com/bench/Product/701?vs=697

x264 HD 5.0.1 First pass : FX8350
x264 HD 5.0.1 second pass : FX8350

POV-Ray 3.7 RC6 MT : FX8350

Visual Studio 2012 : Core i5 3570K

Mozilla Kraken Java script : Core i5 3570K

Photoshop CS4 : Core i5 3570K

DivX 6.8.5 : Core i5 3570K

x264 HD 0.59.819 first pass : Core i5 3570K
x264 HD 0.59.819 second pass : FX8350

Windows Media Encoder 9 64bit : Core i5 3570K

Cinebench 10 MT : FX8350

POV-Ray 3.7 Beta 23 : FX8350

Par 2 MT : FX8350

Blender 2.48a : Core i5 3570K

MS Excel 2007 SP1 Monte Carlo : FX8350

Sorenson Squeeze Pro 5 : FX8350

WinRAR 3.8 : Core i5 3570K

CineBench 11.5 MT : FX8350

x264 HD 3.03 first pass : Core i5 3570K
x264 HD 3.03 second pass : FX8350

7-zip : FX8350

Out of 21 applications in AT bench suite, FX8350 wins 12 of them and core i5 3570K wins 09.
I dont see how you came to the conclusion that the FX8350 being inferior to the Core i5

 

[ShintaiDK]

Lets take every real life application from AT bench list and see what the final score will be,

http://www.anandtech.com/bench/Product/701?vs=697

x264 HD 5.0.1 First pass : FX8350
x264 HD 5.0.1 second pass : FX8350

POV-Ray 3.7 RC6 MT : FX8350

Visual Studio 2012 : Core i5 3570K

Mozilla Kraken Java script : Core i5 3570K

Photoshop CS4 : Core i5 3570K

DivX 6.8.5 : Core i5 3570K

x264 HD 0.59.819 first pass : Core i5 3570K
x264 HD 0.59.819 second pass : FX8350

Windows Media Encoder 9 64bit : Core i5 3570K

Cinebench 10 MT : FX8350

POV-Ray 3.7 Beta 23 : FX8350

Par 2 MT : FX8350

Blender 2.48a : Core i5 3570K

MS Excel 2007 SP1 Monte Carlo : FX8350

Sorenson Squeeze Pro 5 : FX8350

WinRAR 3.8 : Core i5 3570K

CineBench 11.5 MT : FX8350

x264 HD 3.03 first pass : Core i5 3570K
x264 HD 3.03 second pass : FX8350

7-zip : FX8350

Out of 21 applications in AT bench suite, FX8350 wins 11 of them and core i5 3570K wins 10.
I dont see how you came to the conclusion that the FX8350 being inferior to the Core i5

You prove exactly my point again and why the FX8350 is a terrible buy. Try categorize those benches and you see why. You even have to show multiple versions of the same apps to try and manipulate your way for the crappy FX8350 to look better than it is. Just like you did before.

But hey, FX8350, when utilities are free and you dont mind noise and only do rendering/encoding/compression all day

Specially in realtime dependent applications, FX8350 comes up short...very very short.

 

[AtenRa]

You prove exactly my point again and why the FX8350 is a terrible buy. Try categorize those benches and you see why. You even have to show multiple versions of the same apps to try and manipulate your way.

You are starting to sweating and searching for straws, it must be unthinkable for you to even think that an AMD CPU could be faster than an Intel.

If i remove x264 first pass, your Core i5 3570K will loose two(2) wins, that will make the final score 12 for the FX vs 7 for the Core i5 3570K.

 

[Olikan]

an interesting info:

Clock mesh
Another new feature, namely the resonant clock mesh, which is used in Trinity is, according to statements from AMD not "Vishera" included. Nevertheless, the basic clock had increased to 400 MHz.

yes, google translated from here:
http://www.planet3dnow.de/vbulletin/showthread.php?t=408737&garpg=3

 

[ShintaiDK]

You are starting to sweating and searching for straws, it must be unthinkable for you to even think that an AMD CPU could be faster than an Intel.

If i remove x264 first pass, your Core i5 3570K will loose two(2) wins, that will make the final score 12 for the FX vs 7 for the Core i5 3570K.

Try again. 2 POV-Ray versions, 2 Cinebench version, x264? Hah...

Not to mention the same overlapping categories.

The only thing in those socalled "real life" apps that more than 1% might use is essentially WinRAR and 7zip. Yet you completely avoid gaming that atleast 20% do.

As icing on the cake we could add this since you are so focused on cost:

 

[NTMBK]

Try again. 2 POV-Ray versions, 2 Cinebench version, x264? Hah...

Not to mention the same overlapping categories.

The only thing in those socalled "real life" apps that more than 1% might use is essentially WinRAR and 7zip. Yet you completely avoid gaming that atleast 20% do.

As icing on the cake we could add this since you are so focused on cost:

Gosh. At full load it uses a light bulb's worth of electricity more. And your computer will spend very little time actually at full load (even less for the FX than for the i5, as it gets the job done quicker ). Forgetting to turn out the lights when you leave the house will cost you more money.

 

[Idontcare]

an interesting info:
yes, google translated from here:
http://www.planet3dnow.de/vbulletin/showthread.php?t=408737&garpg=3

If you want to learn more about resonant mesh, what it does accomplish and the tradeoffs that came in implementing it, you should check out AMD's whitepaper on the topic.

For example, they can't/don't use resonant mesh clocking below 2.9GHz because it actually increases power consumption relative to conventional clocking below 2.9GHz. (ironically this means it isn't really used in mobile products despite the fact that it was supposed to enable lower-power products)

At full load and optimal clockspeed, the average power reduction of resonant mesh versus conventional clocking is 4.5% per AMD (see their conclusion slide).

Even when operating at optimal clockspeed (the effect has a peak efficiency, it doesn't keep increasing with increasing clockspeed), temperature undoes the benefits. You have to keep the temps down.

Optimal clockspeed for peak efficiency benefits from resonant mesh is 3.3GHz (see slide 31). This is hardwired into the mesh layout and design. It is not a factor of process maturity or process variation.

Resonant mesh actually decreases the fmax (the peak clockspeed for the CPU). They iteratively optimized the resonant mesh for piledriver such that the impact was minimized to 0.2% lower max clockspeed. This decrease in max clockspeeds is the result of higher clock skew from the resonant mesh over that of conventional mesh (see slide 29).

Basically it is a solution that has a narrow window of utility. You have to operate the CPU within a pre-defined clockspeed window (not too slow and not too fast) and keep the temperatures low to avoid losing the benefits. And even then, when optimally operated, it yields 4.5% average power reduction under load.

4.5% power reduction is worth about 100MHz, not 400MHz, if you want to increase the clockspeed and burn up the power-savings with higher clocks.

 

[Olikan]

If you want to learn more about resonant mesh, what it does accomplish and the tradeoffs that came in implementing it, you should check out AMD's whitepaper on the topic.

For example, they can't/don't use resonant mesh clocking below 2.9GHz because it actually increases power consumption relative to conventional clocking below 2.9GHz. (ironically this means it isn't really used in mobile products despite the fact that it was supposed to enable lower-power products)

i think that you missed the part, that there is no Clock Mesh on Vishera...just on trinity

 

[Rvenger]

Its clear that the 8350 is faster than the 3570k and many reviewers pointed that out. I don't need anyone to convince me otherwise because I can read the graphs, thank you.



We are comparing a 4 core to an 8 core (With shared resources or HT, what ever you want to call it) Of course it will have higher power consumption.