4850x2 review

[W1zzard]

so you claim bottleneck even though there are cards on the same rig that perform better? please do your homework first. a bottleneck looks like tf2 on 1024x768. all cards over a wide range score pretty much the same.

my philosophy is to test on rigs that people can afford. you do know that most titles don't benefit from quad core or 4 gb, a psu has zero effect on performance.
if you dont like the test configuration, feel free to read other reviews. maybe we should adopt hardocp's method of banning all persons who criticize reviews.

 

[Solaris17]

(ok this last part. I looked and found UT3 and HL2, no comments - good games - but again no comments)

Yes i see that, there are some tests a 9800gtx can do better than other high end cards, but this is because the games graphics engine is not always perfect, the UT3 engine is really light and optimized a card like a 9800GTX can sure take it to the max. I don't care if a 9800GTX can beat my card in say FAR CRY 1 for god sakes! I always look for vantage and new game releases, instead of old games. If i was interested in old games i wouldn't had upgraded from a 8800GT

And about wiz review, look at his test setup, 2GB? 3.6Ghz? 700WPSU? I smell bottleneck. And to say that a 4850X2 is better than a 70 even if the results are well within a margin of error?
Just look at benchmarkreviews testsetup, that a rig of respect

And in other subject, nice Q9650 over, and on air... me wants one now

not only is that rig based on hardware most people can afford but all hardware and card reveiws are beched on that same system so even if their was a bottleneck the scores and performance are STILL COMPARABLE because every card will sufer the same.....imho you seem to be basing all your doubt (as youve mentioned it in multiple posts) not on the test system or hardware/arch but you keep mentioning 50 and 70 ...i have one honest question for you and im not trying to be an ass or call you out but are you one of those people that judge hardware being better than another by the model name? like just to be clear you know for example that a 7900GX2 will beat an 8600 regardless of the 8600 having a bigger number right?

 

[wolf2009]

maybe we should adopt hardocp's method of banning all persons who criticize reviews.

woah , thats some serious stuff !

 

[douglatins]

so you claim bottleneck even though there are cards on the same rig that perform better? please do your homework first. a bottleneck looks like tf2 on 1024x768. all cards over a wide range score pretty much the same.

my philosophy is to test on rigs that people can afford. you do know that most titles don't benefit from quad core or 4 gb, a psu has zero effect on performance.
if you dont like the test configuration, feel free to read other reviews. maybe we should adopt hardocp's method of banning all persons who criticize reviews.

Yes i read them, and yours is the only one that states as the 4850X2 better that a 4870X2, absurd.

 

[btarunr]

And about wiz review, look at his test setup, 2GB? 3.6Ghz? 700WPSU? I smell bottleneck. And to say that a 4850X2 is better than a 70 even if the results are well within a margin of error?
Just look at benchmarkreviews testsetup, that a rig of respect

...you lost me there

For Crysis, a E8400 at stock speeds performs on par with a QX6850. Perhaps Supreme Commander is the only test we have, that can benefit from 4 CPUs.

2 GB for a Windows XP environment, and for high-end video card reviews is more than sufficient. 700W PSU? Here's a surprise: The VX450W PSU Corsair sells, can run a Core 2 Quad + HD 4870 X2 + 2 HDD according to Corsair. If a single HD 4870 X2 bench genuinely requires a PSU better than the 700W one TPU uses, it's ATI epic fail, going by your own logic.

Well, W1z did his best to show you how the HD 4850 X2 performs, which is very much part of his job. It seriously isn't his job to explain/justify how a HD 4850 X2 outperforms a HD 4870 X2. It's Sapphire/AMD's job, so go bother the people really behind this.

 

[douglatins]

woah , thats some serious stuff !

Great Axed in free speech right?

And people do check other reviews umkay? And to wiz, sorry if i have been disrespectful in any way, its just my honest criticism. And if a PSU cant supply enough power to a card they sure have an impact on performance, and pls no smart comments about my PSU, its on my upgrade list a 1000W corsair

 

[Solaris17]

Great Axed in free speech right?

i honestly dont think youve been around enough or you would know he isnt being serious...imho he posts better reveiws than other sites i mean seriously look at other sites 3 way sli on 280's 8GB 2000mhz DDR3 128GB SSD like seriously wtf kinda test bench is that can you afford that? that and w1zz tests retests and verifys his reveiws i honestly think alot of these other sites post what you want to see or rig it to do so then they play with the numbers a bit and throw off the tech world.

 

[douglatins]

i honestly dont think youve been around enough or you would know he isnt being serious...imho he posts better reveiws than other sites i mean seriously look at other sites 3 way sli on 280's 8GB 2000mhz DDR3 128GB SSD like seriously wtf kinda test bench is that can you afford that? that and w1zz tests retests and verifys his reveiws i honestly think alot of these other sites post what you want to see or rig it to do so then they play with the numbers a bit and throw off the tech world.

Great by that thought, a Ferrari is bad, because of traffic and speed limits. Comparing VGAs is complicated even more high-end/price ones, you have to have the best hardware so that just the VGAs matter, im not saying is unimportant to test in a "real" pc, but its incomplete.
Again you wouldnt review a ferrari, if you were into speed and all that, just by taking a drive in the park.

 

[Solaris17]

Great by that thought, a Ferrari is bad, because of traffic and speed limits.

what im saying is because of traffic and speed limits is exactly why traffic and speed limits dont matter a ferrari and porsche are going to perform relatively the same it doesnt matter if you on the autobahn or city streets as long as both cars are in the same element their speeds excel and MPG are comparable

 

[btarunr]

Comparing VGAs is complicated even more high-end/price ones, you have to have the best hardware so that just the VGAs matter, im not saying is unimportant to test in a "real" pc, but its incomplete.

The point is, the "Ferrari" we're reviewing is already on the Autobahn. You're basically asking for us to drive it on a road made of gold, which doesn't make sense.

 

[douglatins]

what im saying is because of traffic and speed limits is exactly why traffic and speed limits dont matter a ferrari and porsche are going to perform relatively the same it doesnt matter if you on the autobahn or city streets as long as both cars are in the same element their speeds excel and MPG are comparable

Ok, in the streets, a ferrari gets an average speed than a civic, so that, is always true, regardless, even on a track. (see how this doesnt work)?

The point is, the "Ferrari" we're reviewing is already on the Autobahn. You're basically asking for us to drive it on a road made of gold, which doesn't make sense.

I respectfully disagree

 

[Solaris17]

The point is, the "Ferrari" we're reviewing is already on the Autobahn. You're basically asking for us to drive it on a road made of gold, which doesn't make sense.

because benching it on an incredably high end system doesnt make sense...if you bench on a midrange gaming machine people have a better idea of how well it will run on their rig and on a higher end system if you bench it on a $5000 gamng machine with something like the specs i said above you will be unable to compare to a lower end machne because the part change is to drastic the bench is scewed and one sided

 

[mullered07]

Techpower has to be sponsored by nvidia, this site posts the worsts X2 reviews i've seen, and what the hell, a 50 cant be better that a 70, something doesnt smell right

because benching it on an incredably high end system doesnt make sense...if you bench on a midrange gaming machine people have a better idea of how well it will run on their rig and on a higher end system if you bench it on a $5000 gamng machine with something like the specs i said above you will be unable to compare to a lower end machne because the part change is to drastic the bench is scewed and one sided

maybe he could spend the $5000 on a rig to bench them himself so we cnn all find out what really is the best, afterall what does wizz's word mean on tpu ?

 

[btarunr]

because benching it on an incredably high end system doesnt make sense...if you bench on a midrange gaming machine people have a better idea of how well it will run on their rig and on a higher end system if you bench it on a $5000 gamng machine with something like the specs i said above you will be unable to compare to a lower end machne because the part change is to drastic the bench is scewed and one sided

Erm did we differ in our views

I respectfully disagree

Back your disagreement with something respectable.

 

[Solaris17]

Erm did we differ in our views

no i think i consentrated one one sentance in yours and that would be benching on a golden road the rest was more of a respons or expansion to earlier posts and ideas i just didnt break it up or direct quote because if i did that id be no longer known on tpu for having horrible typing habits and i wouldnt want to let down a community of thousands of people

 

[douglatins]

because benching it on an incredably high end system doesnt make sense...if you bench on a midrange gaming machine people have a better idea of how well it will run on their rig and on a higher end system if you bench it on a $5000 gamng machine with something like the specs i said above you will be unable to compare to a lower end machne because the part change is to drastic the bench is scewed and one sided

I stated that both are important, but to say one is better that the other, i only agree with review using really high-end rigs, c'mon look at Guru3D Multi-GPU, nehalem test. Nehalem puts the core2s to shame. But SLI still suck, because people only have E8400? Theres a bigger picture here

 

[douglatins]

Erm did we differ in our views

Back your disagreement with something respectable.

I already spent to much time here in the first place, wont reply something "respectable", because i dont think is worth doing it. In this case.

 

[btarunr]

I already spent to much time here in the first place, wont reply something "respectable", because i dont think is worth doing it. In this case.

So long

 

[douglatins]

So long

that doent mean i'm leaving, just that i wont reply to narrow minded people.

 

[btarunr]

that doent mean i'm leaving, just that i wont reply to narrow minded people.

Ah...

Validate your broadmindedness by backing your argument with something concrete.

 

[Binge]

I don't mean to be crazy or obtuse to anyone but the proc is not the issue here. In doing more research I've found a lot of people complaining about performance hits on p35/some p45 boards when using 4870x2. That still wouldn't make sense because the 4850x2 should have the same pci-E config as the 4870x2. What's interesting is that in all the reviews with the 4870x2 still coming out on top there is DDR3 memory in the mobo. Just food for thought.

As for our pal the W1z I believe it was a fine review because afterward I felt that with that system I would want to go with a gtx260 216 or a 4870 1gb. If you read the review, you may be thinking it's going to tell you what will work best in your configuration when you know they're completely different-- think again.

 

[Solaris17]

I stated that both are important, but to say one is better that the other, i only agree with review using really high-end rigs, c'mon look at Guru3D Multi-GPU, nehalem test. Nehalem puts the cores to shame. But SLI still suck, because people only have E8400? Theres a bigger picture here

i understand man im not one that doesnt undrstand the "bigger picture" im just saying that i and many others find this way of benching better comparison purposes anyone can project performance onto a better system however you cannot do the same in reverse.

 

[douglatins]

Ah...
Validate your broadmindedness by backing your argument with something concrete.

How do i state a concrete argument if we are in a hypothetical scenario. At least about the hole road discussion. A same level response would be a counter statement. An autobahn should be a Nehalem setup, a E8400 should be a good highway.

 

[btarunr]

How do i state a concrete argument if we are in a hypothetical scenario. At least about the hole road discussion. A same level response would be a counter statement. An autobahn should be a Nehalem setup, a E8400 should be a good highway.

Only that you still don't seem to understand, that all the tests run are tailored for a dual-core setup (barring SupCom, methinks). What you don't understand, is that both HD 4850 X2 and HD 4870 X2 were tested on the very same setup, so the platform becomes abstract, and the performance numbers are all that's left to see.

The machine used is the reflective of that kind of hardware really goes into gaming PCs.

 

[douglatins]

How do i state a concrete argument if we are in a hypothetical scenario. At least about the hole road discussion. A same level response would be a counter statement. An autobahn should be a Nehalem setup, a E8400 should be a good highway.

And Solaris its okay to do it, just dont state as being surely better regardless. People would read this and think "OMG, OMG, the cheaper 4850X2 ownzz 4870X2", but this is only true in this particular scenario.