![]() |
|
|
#176 |
|
Senior Moderator
Join Date: Jan 2006
Location: Worcestershire, UK
Posts: 12,731 (4.75/day)
Thanks: 1,093
Thanked 1,931 Times in 1,435 Posts
|
He probably means that an overclocked Q9650 @ 4.5gig is near to an overcocked i7 920 @ 4.2 - 4.3gig, I went 920 from 9650 and found a decent increase in performance across the board and thats with my old 9650 @ 4.5gig.
__________________
|
|
|
|
|
|
#177 |
![]() Join Date: Sep 2009
Posts: 160 (0.12/day)
Thanks: 68
Thanked 65 Times in 30 Posts
|
__________________
“I would like for him to point out and quote specifically the portion of both pics that shows a bart as a 6800 series...” -EastCoasthandle
|
|
|
|
|
|
#178 |
|
Senior Moderator
Join Date: Jan 2006
Location: Worcestershire, UK
Posts: 12,731 (4.75/day)
Thanks: 1,093
Thanked 1,931 Times in 1,435 Posts
|
Cant read/see any of them lol..... maybe just me, old eyes and that!
__________________
|
|
|
|
|
|
#179 | |
![]() Join Date: Sep 2009
Posts: 160 (0.12/day)
Thanks: 68
Thanked 65 Times in 30 Posts
|
Quote:
http://www.hardware.fr/articles/842-...rma-ii-oa.html
__________________
“I would like for him to point out and quote specifically the portion of both pics that shows a bart as a 6800 series...” -EastCoasthandle
|
|
|
|
|
| The Following User Says Thank You to jaredpace For This Useful Post: |
|
|
#180 |
![]() Join Date: Mar 2005
Location: USA
Posts: 2,100 (0.70/day)
Thanks: 131
Thanked 517 Times in 383 Posts
|
Great Review. Makes me happy to see my i7-930 is still a great performer at the resolution I run which is 2560x1600. Almost no difference at least not enough to force me to upgrade. Here's to holding out for i9.
__________________
i7 930 @ 4.2Ghz GIGABYTE x58 UD7 Kingston HyperX T1 Series 12GB GIGABYTE 7970 (1280/1800) OCZ Vertex 2 100GB SSD & VelociRaptor 300GB HEATWARE |
|
|
|
|
|
#181 |
![]() |
Thanks
Nice review thanks, but there is one thing i don't understand.
The FX matched the 2500K in the highest test, that's the FX bottlenecking the GPU more then the 2500K, ok fine.. But if you look here http://amdfx.blogspot.com/2012/02/di...y-request.html They have set the AA to 8x where here it is 4x. The FPS are about a continuation from what they are in the test here, and suddenly clock for clock the FX overtakes the 2500K by 20% (avr-FPS) What is going on with that? |
|
|
|
|
|
#182 |
![]() Join Date: Aug 2011
Location: Perth, West Australia
Posts: 373 (0.57/day)
Thanks: 32
Thanked 68 Times in 59 Posts
|
Its Actually very good to see the i7 920 still holding up with the big guns. It deserves credit. How old is it now and what speed is it running at! Im going to thank w1zzard 4 this one. Very informative review.
|
|
|
|
|
|
#183 | |
![]() Join Date: Mar 2011
Location: Burlington, VT
Posts: 1,706 (2.17/day)
Thanks: 228
Thanked 404 Times in 297 Posts
|
Quote:
![]() If Nahalem won there, I assure you SB would have done even better. You cannot compare data from multiple data sets since everyone uses different setups. That post is clearly bias since it's from, you know, AMDFX.blogspot.
__________________
“If BF3 doesn't turn out to be the biggest selling PC shooter of all time, then I will eat my graphics card.” -MatTheCat
|
|
|
|
|
|
|
#184 | |
|
Senior Moderator
Join Date: Jul 2006
Location: Milwaukee, WI.
Posts: 31,880 (12.77/day)
Thanks: 2,779
Thanked 12,253 Times in 7,806 Posts
|
Quote:
|
|
|
|
|
|
|
#185 | |
![]() Join Date: Jan 2012
Location: Dover, New Hampshire, USA
Posts: 4,233 (8.87/day)
Thanks: 1,259
Thanked 1,313 Times in 975 Posts
|
Quote:
Seriously though, it's a review worth checking out if you haven't already. http://www.guru3d.com/article/radeon...rmance-review/ Bulldozer isn't bad, it's just not as good as Sandy Bridge. The sooner everyone realizes that, the less flaming there will be. Also you can only improve a processors IPC so much before you hit a top-end limit. As software develops for post anno-2012, you will notice a lot more software utilizing more threads, and for bulldozer, an architecture that scales almost linearly with the amount of cores (unlike Intel's HyperThreading,) you will see that in the future more cores will result in more speed. Similar to video encoding on the FX-8150, something that it keeps up with the 2600k very well. Reference for video encoding: http://www.guru3d.com/article/amd-fx...ssor-review/14 (Yes, I know that the FX-8150 doesn't do as well on other benchmarks, but this is a practical application where the FX-8150 is more than adequate when all cores are being used.) Also keep in mind that when you're playing a game that can't use all 8 cores, you have that extra power to do other things. Virtual machines? Folding? Video encoding? Yeah, it can do that. :| Last edited by Aquinus; Feb 26, 2012 at 10:47 AM. |
|
|
|
|
|
|
#186 |
![]() Join Date: Jan 2012
Location: Dover, New Hampshire, USA
Posts: 4,233 (8.87/day)
Thanks: 1,259
Thanked 1,313 Times in 975 Posts
|
Since no one really wants to make a comparison, I grabbed CPU benchmark numbers from Guru3d and plugged them into excel. I calculated Performance of FX-8150 vs i7 2600k and i5 2500k.
If you take a look at the attached document, you will see bulldozer (at stock speeds,) excels over the i7 2600k in the following applications: FlyRender Espresso Transcode Video Transcoding (H.264 (DTS5.1) to x.264 AC3 5.1) CPU Hash Bench using SHA1 Sandra Memory Index (matches i7,) When you overclock Bulldozer to 4.6ghz, it takes the lead in the following applications: SiSoft Sandra FPU GFLOPS Zlib Compression VP8 Video Compression Notice a trend? All the heavily multi-threaded applications seem to chug right on through with a bulldozer. It only suffers on severely single-threaded applications, and even if it does it runs adequately fast for them anyways, but if you're getting a BD, it is not for single-threaded performance, that is for sure. Also keep in mind the price of the 2600k and the FX-8150, so anything that is not less than 25% slower than the i7 2600k gives you what you paid for in 75% of all cases. Bulldozer delivers for the price that it is set at very well and for any instance where the 8150 overtakes the 2600k is simply a great value for what you're getting. You can all bash Bulldozer as much as you want, but numbers don't lie. For a brand new, first revision platform, it does pretty damn well. Information provided is courtesy of Guru3d. http://www.guru3d.com/article/amd-fx...essor-review/1 Edit: I would also like to add that all of these benchmarks use DDR3-1600, and not the FX-8150's native 1866. Guru3d also didn't try overclocking the 8150's FSB, which doesn't help the memory controller or cache latency. Last edited by Aquinus; Feb 26, 2012 at 12:41 PM. Reason: Spelling and grammatical corrections. |
|
|
|
| The Following User Says Thank You to Aquinus For This Useful Post: |
|
|
#187 | |
![]() |
Quote:
Crank the resolution up, the GPU needs to do more, and more to keep up, enable sufficiently demanding anti-aliasing and anisotropic filtering the GPU will no longer be able to keep up with the CPU, just because this test shows the 2500K running 200 FPS compared with 120 FPS with the FX is not to say the 2500K is the more powerful CPU for THIS game. If you look at this test more closely you can see where the ceiling vs the ceiling of the FX meet, or not... as at the very least they match in the highest test, if not the FX starts to overtake,- it is computing more FPS in the last test (all be it very slight) the FX FPS has barely moved from where is was at low resolution, where the 2500K has dropped by about 50%. If you look at another review sites (mentioning no names here) test of BF3 it shows an FX-4100 matching a 2500K and as near as makes no difference an FX-8150 and 2600K with an i3 2100 not far behind at high graphics settings / resolution. I would not write the FX Blog off, had a higher test been published here we MIGHT be having a very different conversation. @ Aquinus, i have a 6 core Thuban, clock for clock it beats a 2500K in multi-threaded benchmarks, an FX-8150 beats my Thuban in the same benchmarks, you compare them in a Linux OS where thread allocation is far more efficient than in Windows and AMD monster Intel. CRAY are using 16 thread Bulldozer CPU's on a Linux based OS for there new supper computer. Last edited by abundant threading; Feb 26, 2012 at 01:07 PM. |
|
|
|
|
|
|
#188 |
![]() Join Date: Feb 2008
Location: Southeastern Virginia
Posts: 3,731 (1.95/day)
Thanks: 367
Thanked 1,079 Times in 795 Posts
|
Guys, erocker has a FX8150 chip and a 7970 gfx card. He has a good idea of what he's talking about.
__________________
“Never underestimate the predictability of stupidity.”
“Amd....definitely a blond female.....not much on the cpu side, but always strutting about showing off her gpus......wow that was bad. I apologize.” -ensabrenoir
|
|
|
|
| The Following User Says Thank You to BarbaricSoul For This Useful Post: |
|
|
#189 | ||
![]() Join Date: Jan 2012
Location: Dover, New Hampshire, USA
Posts: 4,233 (8.87/day)
Thanks: 1,259
Thanked 1,313 Times in 975 Posts
|
Quote:
![]() (For those of you who don't know, at higher resolutions there is more stress on the GPU, not the CPU.) You can't compare the Linux kernel and Windows NT (yes, ever since Windows NT, all versions of Windows has been based on the NT kernel,) because they're completely different animals with different goals in mind... also the average user won't care about *nix performance on their platform... unlike us. ![]() Quote:
![]() Edit: If you have time, I don't want to be pushy because I know how valuable time is and how long benchmarks can take. |
||
|
|
|
|
|
#190 |
![]() |
Yes i know, as i said in my post an i3 doesn't do to badly either...
although it is the one where its ability to keep up has faltered, an i3 2100 is an i5 2500 with half as many cores... and a 45nm Athlon is basically a Phenom without L3. (what do you make of that)For me, in this test we simply don't know whats better for this game, there comes a point where the CPU can no longer keep up, or we could run these games on Pentium II's, this test does not show where the 2500K vs the FX gives up first. it shows they are both as capable as eachother at setting and REZ this game will be played at, the 2500K is not the winner in this Dirt3 or BF3... test as it stands, more data is needed. Last edited by abundant threading; Feb 26, 2012 at 04:59 PM. |
|
|
|
|
|
#191 |
|
My name is Dave
Join Date: Apr 2006
Location: The Great White North
Posts: 10,766 (4.15/day)
Thanks: 4,472
Thanked 5,185 Times in 3,196 Posts
|
I'm not sure why though? I mean, it's a given that the 8 threads of the 8150, in some instances, will beat out the four threads of the 2500k.
But the 2500K costs less than the 8150, and in gaming, the actual number of titles that benefit from the extra threads are few and far between. the FX chips aren't bad..they are just badly priced in order to compete. When most people are concerned about cost, that completely kills the FX's chance of success on a wide scale. That doesn't mean it's a failure...it's just not going to be the popular choice. A few benchmarks isn't going to change that.
__________________
Gadgets, Phones, Tablets, Cameras, TVs, HiFi...NextPowerUp -Only real men play games THIS way. |
|
|
|
| The Following User Says Thank You to cadaveca For This Useful Post: |
|
|
#192 | |
![]() |
Quote:
|
|
|
|
|
|
|
#193 | |
|
My name is Dave
Join Date: Apr 2006
Location: The Great White North
Posts: 10,766 (4.15/day)
Thanks: 4,472
Thanked 5,185 Times in 3,196 Posts
|
Quote:
When electronic retailers sell "upgrade" warranties for considerable value, consumers on a whole put real monetary value on their warranties, and having made just one successful RMA, really bolsters that value. Enthusiasts know the real story already. However, stock performance of an 8150 isn't what enthusiasts are after. They are after overclocked performance, and on Intel, you can purchase a warranty to cover your OC, nevermind that cost of cooling that OC'ed system is much less, thanks to Intel's lower power consumption, so the average consumer and enthusiast alike is more likely to OC with Intel, provided sales people do their jobs properly. Either way, you can go with either camp and have a fairly decent system, each with it's own strength and weaknesses. With ALL things considered, to me, Intel is better with current generation stuff. For the same power on the high-end, you get x79, not a 2500k, so for me, that's where the real comparison is, and at that, Intel to me still offers more. After all, A75 is AMD's entry platform right now, while 1155 is Intel's entry. Both come with onboard GPUs. AMD is mroe cost-effective, but lacks Intel's CPU processing grunt. X79 and FX...well, you get the picture.
__________________
Gadgets, Phones, Tablets, Cameras, TVs, HiFi...NextPowerUp -Only real men play games THIS way. |
|
|
|
|
|
|
#194 |
![]() |
Oh no argument from me, the FX is a disappointment to me but i don't think its as bad as some like to portray it, as you say in a round about way.
AMD have taken a big gamble here designing a completely new architecture, its not paying diffident yet, but it is new and may pay off in the longer term, Intel had the same problems with P4, you may remember? A more interesting discussion is the future Piledriver chips which will use a resonant clock mesh based on that new architecture. That's exiting stuff, but perhaps not for this thread. |
|
|
|
|
|
#195 | |
![]() Join Date: Jan 2012
Location: Dover, New Hampshire, USA
Posts: 4,233 (8.87/day)
Thanks: 1,259
Thanked 1,313 Times in 975 Posts
|
Quote:
As for resident clock mesh, I don't think it will be enough to lower power consumption as much as Intel's 3d (3 sources, 3 drains per transistor,) transistor. Like I said before, it is a new architecture and for the amount of die space per module, you can cram more cores on the same die area. The future is multi-threading. This really has been running on but I will say this. Intel has the CPU market figured out for now. AMD has the CPU market figured out for down the road. It really is as simple as that. (Bulldozer makes a very nice server chip, and for workstations, dual 8-core Valencias at 3.3ghz is just plain sexy and cost effective.) |
|
|
|
|
|
|
#196 | |
![]() |
Quote:
![]() PD's will start with a clock speed of around 4Ghz and moving up from there with 8 or maybe 10 cores with better power consumption and less heat. it might not be as efficient as Ivy but a whole lot more cores and better OC potential than BD..... that's something worth an interest in. Last edited by abundant threading; Feb 27, 2012 at 12:07 AM. |
|
|
|
|
|
|
#197 |
![]() Join Date: Mar 2011
Location: Burlington, VT
Posts: 1,706 (2.17/day)
Thanks: 228
Thanked 404 Times in 297 Posts
|
Your argument falls flat on its face when you consider that, who cares how threaded the CPU's of today are when the future is in multi-threading? Intel scales their Core Count with what the average consumer uses, they could probably release 6 and 8 core iterations if they wanted, but why release a product only 0.1% of your userbase could benefit from? By the time 8 threads is the norm for all applications, BD will be long gone. People have been praising multi-threaded apps since like 2005, but software to accomedate that is just too slow.
__________________
“If BF3 doesn't turn out to be the biggest selling PC shooter of all time, then I will eat my graphics card.” -MatTheCat
|
|
|
|
| The Following User Says Thank You to xenocide For This Useful Post: |
|
|
#198 |
|
OH, I have such a headache
Join Date: Dec 2004
Location: Planet Earth.
Posts: 6,080 (1.97/day)
Thanks: 1,805
Thanked 872 Times in 734 Posts
|
Looks like Sandy E Bridge is killing every thing! Man Intel is still on top and looks like it will be there for some time to come! Intel FTW!!!
__________________
Asus P5Q pro turbo MB , Q9650 @ 4.0GHz water cooled , 8gb Hyper-X RAM @1066MHz , VisionTek HD5870 , Corsair TX850w PSU , Barracuda HDD SATA RAID0 1Tb , Windows 7 pro 64 bit . HeatWare |
|
|
|
|
|
#199 |
![]() |
|
|
|
|
|
|
#200 | |
|
OH, I have such a headache
Join Date: Dec 2004
Location: Planet Earth.
Posts: 6,080 (1.97/day)
Thanks: 1,805
Thanked 872 Times in 734 Posts
|
Quote:
__________________
Asus P5Q pro turbo MB , Q9650 @ 4.0GHz water cooled , 8gb Hyper-X RAM @1066MHz , VisionTek HD5870 , Corsair TX850w PSU , Barracuda HDD SATA RAID0 1Tb , Windows 7 pro 64 bit . HeatWare |
|
|
|
|
![]() |
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | |
|
|
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Ivy Bridge Official Benchmarks – Markedly Better Performance Than Sandy Bridge | qubit | News | 66 | Dec 2, 2011 05:05 AM |
| Ivy Bridge Early Sneak Performance Peek: Any Faster Than Sandy Bridge? | qubit | News | 34 | Nov 28, 2011 05:42 PM |
| Sandy Bridge E or Ivy Bridge | the54thvoid | System Builder's Advice | 5 | Oct 18, 2011 05:20 PM |
| Sandy Bridge-E 2011 Launch Put on Slide, Ivy Bridge in March-April 2012 | btarunr | News | 20 | Aug 9, 2011 05:12 PM |
| Intel Ivy Bridge Dual-Core Put Through Clock-to-Clock Benches Against Sandy Bridge | btarunr | News | 70 | Jul 7, 2011 11:53 PM |