Sign in with
Sign up | Sign in
Your question

OBRovsky is a troll - with regards to 1100t vs 8150 FX "benchmarks"

Last response: in CPUs
Share
September 17, 2011 8:03:33 PM


Im hjh the "kiddo" he keeps referring to and discrediting over and over on his website.

http://obrovsky.blogspot.com/

I have asked him many good questions that he promised to respond to through email yet this was in fact a lie .. he never did.

Most of my concerns are as follows: "comments on wccftech post with regards to his 1100t vs gx 8150 clock per clock comparisons


Comment 1:

Haha, people need to face the facts. If a fx 6110 costs more than an 1100t , 188 vs 179.. It will be an upgrade . If these results are representative of actual core per core basis, ie ipc, then. The fx 6110 wuld cost closer to 90 usd. Based on the leaked pricing people shuld take this with a grain of salt, and understand that amd is in the game for price per performance so they would never allow this to be true, PERIOD. Obr is a troll, for all we know these benchmarks are 100 percent fake. The fx 6110 is not 50 percent slower than an 1100t, amd is not retarded, don't believe these results one bit.

Comment 2:


Also if llano is 6 percent better ipc than phenom II as seen from official benchmarks, and bulldozer is 20 to 50 percent worse ipc, then why would amd bother releasing trinity ( a 4 core bulldozer apu) as their next gen solution to replace llano.

None of this adds up. I take amd's prowess over obr's trolling any day. They wouldn't dig their own grave by releasing new worse products for more money, period. He needs to grow up and admit that instead of acting like he knows everything when it comes to bulldozer. He needs to face these facts infront of him. Its common sense.

Comment 3:

Not to mention these benches imply an 1100t beats an 8150 fx when all cores are used in cinibench.. You know that's bullshit. No company would work on a highly multithreaded chip for years and redesign their architecture to get worse performance, and severley worse performance per dollar. Amd always states that bulldozer will be extremley well priced for its performance.

Obr is a troll, I don't even believe these results are credible at all. Based on the two official benchmarks amd did release, where amd beats the 2500k and 980x in video compression and gaming, I can already see these results are faked or biasly tested. How the hell would something with 40% less IPC than AMD 1100t beat a 980x in gaming tests which NEVER utilize more than 4 cores.. ie HALF the cores of the FX 8150. (it should be noted that obr's chip results show bulldozer performing worse than a 100 dollar phenom x4 in gaming tests)

Source: http://vr-zone.com/articles/amd-benchmarks-fx-against-i...

It should be noted that this CHIP he used is over 2 months old and is not even close to the final retail chip. (although he claims it to be b2 stepping he only hit 6.5 ghz on l2n when near production samples easily hit near 8 ghz)




a b à CPUs
September 17, 2011 8:37:46 PM

obviously we all knew that he may have had a prototype chip but its never representative of a retail product
September 18, 2011 7:54:47 PM

Except for the fact that he is 100 percent sure that the final retail b2 chip is no different than the chip he has , referring to unsourced claims that the retail chips are never really fundamentally different.

Its known amd is working its hardest to make sure zambezis launch is a strong one, and the stocks show everyone is anticipating it.

It makes no econommical sense for amd to release a chip on par with the one tested by obr ( if the results are even legit or unbias which is highly doubt ) at the prices listed and that's something everyone should agree on.
Related resources
Can't find your answer ? Ask !
a c 96 à CPUs
September 18, 2011 8:38:38 PM


... simply wasting bandwidth, giving him recognition and sending him page hits
September 18, 2011 11:23:48 PM

Well what he's saying at least makes sense somewhat. Core for Core I highly doubt it's faster that Sandy Bridge or even Phenom. But maybe I'll eat my words when it releases.
a c 79 à CPUs
September 18, 2011 11:55:19 PM

you doubt its faster than phenom core for core? why would they release a slower CPU? If that turns out to be true on release, noone will buy them, except for people that for some reason must have an 8 core cpu.
September 19, 2011 2:47:20 AM

iam2thecrowe said:
you doubt its faster than phenom core for core? why would they release a slower CPU? If that turns out to be true on release, noone will buy them, except for people that for some reason must have an 8 core cpu.


You just answered your own question. I don't know, I'm not AMD. But everything is moving to parallel processing where more cores = better.

But it makes you wonder when a 8 core FX is only 19% faster than a 4-core i5 that's running 400MHz slower in a multi-threaded application.
a c 79 à CPUs
September 19, 2011 7:04:52 AM

i wonder what would happen if someone just made a single core CPU with one BRUTALLY fast core that can handle lots of instructions per clock and lots of cache and clock speed? do you think it would keep up with dual or quad cores?
September 19, 2011 4:19:44 PM

iam2thecrowe said:
i wonder what would happen if someone just made a single core CPU with one BRUTALLY fast core that can handle lots of instructions per clock and lots of cache and clock speed? do you think it would keep up with dual or quad cores?


I don't know, is the application single threaded or heavily multi-threaded? I think in today's software having a mix is best. So 4 fast cores over 8 slower cores. The processor with the 8 cores may be faster overall when all cores are used, but what's the point if something only takes advantage of only 2, 3, or 4 cores such as most video games?
a b à CPUs
September 19, 2011 5:46:15 PM

Quote:
i wonder what would happen if someone just made a single core CPU with one BRUTALLY fast core that can handle lots of instructions per clock and lots of cache and clock speed? do you think it would keep up with dual or quad cores?


Taking heat and power out of the equation, and assuming memory access isn't a bottleneck, then in theory, a highly clocked, high performance single core CPU could be faster then any multi-core CPU, regardless of scaling. But the heat and power loads would make this near impossible to accomplish, and memory access times becomes a major limiting factor. But from a purely technical standpoint, there is no reason why a very fast single core could not be faster then a multi-core.
a b à CPUs
September 19, 2011 5:51:19 PM

Quote:
It should be noted that this CHIP he used is over 2 months old and is not even close to the final retail chip. (although he claims it to be b2 stepping he only hit 6.5 ghz on l2n when near production samples easily hit near 8 ghz)


One of my friends managed to clock his Q6600 @ 4.6GHz. The other could only get it to 3.2GHz. Chips vary, even within the same batch.

Quote:
Haha, people need to face the facts. If a fx 6110 costs more than an 1100t , 188 vs 179.. It will be an upgrade . If these results are representative of actual core per core basis, ie ipc, then. The fx 6110 wuld cost closer to 90 usd.


You assume pricing is representative of performance. It has more cores, it is running faster, so therefore, it should cost more. And if consumers don't recognize its a slower product, thats their problem.

So lets not conclude that because its more expensive, its better, kay?

Quote:
Not to mention these benches imply an 1100t beats an 8150 fx when all cores are used in cinibench.. You know that's bullshit. No company would work on a highly multithreaded chip for years and redesign their architecture to get worse performance, and severley worse performance per dollar.


Remember the AMD K5? Great ideas, but the final implementation was lacking to say the least. Better to release something and get some money back then simply abandon all the R&D Costs.


Note: I am NOT agreeing with the blogger, I'm simply pointing out that there is some logic you can apply to his numbers.
a b à CPUs
September 19, 2011 5:58:19 PM

My thoughts is this: Until a RELIABLE site releases benchmarks, everything right now to me is just rumors.
a c 134 à CPUs
September 19, 2011 10:05:03 PM

My thought was: Who cares if OBR is a troll.. its his blogspot and he can say what he wants. Speaking of trolls...Why bring this garbage thread here?
a c 79 à CPUs
September 19, 2011 11:10:15 PM

popatim said:
My thought was: Who cares if OBR is a troll.. its his blogspot and he can say what he wants. Speaking of trolls...Why bring this garbage thread here?

because now we have ANOTHER thread to speculate about AMD bulldozer performance in........YAY..........
September 20, 2011 2:10:11 AM

iam2thecrowe said:
because now we have ANOTHER thread to speculate about AMD bulldozer performance in........YAY..........


No need to speculate. What he's saying sounds plausible. Doesn't mean it's true, but he could be right.
a c 123 à CPUs
September 20, 2011 2:57:11 AM

I think it would be plausable. Remember they are not "full" cores, or say like Phenom II was.

As for AMD releasing slower parts, they did it with Phenom I. Phenom I on a per core and per clock level was slower than the 65nm Athlon 64 X2s. The original Pentium 4 (started at about 1.4GHz) was slower than a 1GHz Pentium III Coppermine.

It happens. Sometimes its better to release the new product and go from there, I doubt AMD would have survived if they delayed Phenom I and instead released Phenom II 18 months later, it would have financially killed them.
September 20, 2011 6:01:49 AM

jimmysmitty said:
I think it would be plausable. Remember they are not "full" cores, or say like Phenom II was.

As for AMD releasing slower parts, they did it with Phenom I. Phenom I on a per core and per clock level was slower than the 65nm Athlon 64 X2s. The original Pentium 4 (started at about 1.4GHz) was slower than a 1GHz Pentium III Coppermine.

It happens. Sometimes its better to release the new product and go from there, I doubt AMD would have survived if they delayed Phenom I and instead released Phenom II 18 months later, it would have financially killed them.


Which is exactly why I have the feeling that their single-threaded performance is going to stink unless they have a new technology we don't know about.
a c 79 à CPUs
September 20, 2011 7:02:51 AM

jimmysmitty said:
I think it would be plausable. Remember they are not "full" cores, or say like Phenom II was.

As for AMD releasing slower parts, they did it with Phenom I. Phenom I on a per core and per clock level was slower than the 65nm Athlon 64 X2s. The original Pentium 4 (started at about 1.4GHz) was slower than a 1GHz Pentium III Coppermine.

It happens. Sometimes its better to release the new product and go from there, I doubt AMD would have survived if they delayed Phenom I and instead released Phenom II 18 months later, it would have financially killed them.

But to AMD's credit, at least if you did buy something you were able to get the next gen CPU to work without upgrading the motherboard, and sometimes even another gen beyond that. My core 2 duo is stuck with no economically viable upgrade option for the board. That is my one true hate for intel. Thank god i didnt go for the 1st gen i5/i7 or i would have had another dead end motherboard. On a side note, what happened to the days where you could put CPU's from multiple different manufacturers in the same socket? I remember the good old Socket 7, you could install an Intel, AMD, Cyrix/IBM or IDT Winchip in the same freakin motherboard!
September 20, 2011 8:07:55 PM

iam2thecrowe said:
But to AMD's credit, at least if you did buy something you were able to get the next gen CPU to work without upgrading the motherboard, and sometimes even another gen beyond that. My core 2 duo is stuck with no economically viable upgrade option for the board. That is my one true hate for intel. Thank god i didnt go for the 1st gen i5/i7 or i would have had another dead end motherboard. On a side note, what happened to the days where you could put CPU's from multiple different manufacturers in the same socket? I remember the good old Socket 7, you could install an Intel, AMD, Cyrix/IBM or IDT Winchip in the same freakin motherboard!


Lack of socket changes is exactly why i3 > 955.
!