Omfg... one Ci7 are nearly ~ 2 X5482. :eek:
Dual socket gainstown will simply destroy everything in dual socket segment... hell i bet a dual socket setup could keep up with a quadsocket one... imagine Beckton... :shocked:
Printable View
in the past, with Core 2, you could not down grade, so, when you was increasing the FSB, after a while, mem was becoming a problem on non extreme edition parts.
Here , you can increase the Bclock, while lowering the mem ratio, that will help in many case.
Got the point?
If you increase the bclock from 133 to 250, you would have need 2000Mhz DDRIII modules, and i am sure that you guy very quickly will pass 250 Bclock.
at this bclock, 800 ratio will be handy.
of course, all of this is at your own risk ...
Edit: ok, got your pm. so no memory overclock without raise the BCLK :(
Thanks
Chris
Are workstation/server segment (Gainestown) parts to be launched this month as well?
It seems that Core i7 addresses the deficiencies in the previous generation in this segment, but I haven't seen any published reviews in the frenzy yesterday (only desktop systems)
Hasn't Intel historically launched workstation/server parts first (along with the desktop "extreme" edition)? - which makes sense given the lucrative markup and nice earnings bump before the end of the quarter.
if 0 out of 64 and 5 out 80 FPS is 'scaling' with 1/2 the bus speed, then your level of comprehension of analysis is severely lacking, assume you mean the lost planet numbers.
Second, you clearly do not understand when GPU serves the limitation, are you a sock puppet? you sound a lot like Gosh aka Kassler.
your numbers are done with a SINGLE 8800GTX judging from those screens....
btw, 200MHz | 400 MHz | Increase in performance
average: 55.5 68.1 22.7 %
Snow: 64 64.1 0.1 %
Cave: 80.1 85.7 6.99 %
and this for a single 8800GTX is quite impressive if you ask me
what do you mean by sock puppet? this is my only XS account dont be silly...
btw, as you can see on my anand link, 2 GPU's scale quite the same on qx9770 and x58, not a lot of difference, but 3 GPU's act really weird
the FSB is just my guess, if your guess is more computational power because of nehalem, fine, then i just search some 5.5ghz qx9770 screens with 3-way sli...
Tom's Hardware clarifies things today and shows that as long as you have the Overspeed Protection turned off in the bios, the 130 watt rev limiter won't get in your way of overclocking ANY Core i7 CPU.
http://www.tomshardware.com/reviews/...e-i7,2063.html
Thanks Francois for clearing up this confusion, yesterday. :up:
There is no cheap DDR3. Also, if I pay $800 for the CPU + mobo, I'm not going to save $100 on a "cheap" RAM. Most likely I'll get 6 or more gigabytes of RAM.
Tell me the apps which you are using and that have benefit and can utilize 8 threads.Quote:
First yes it is overclocked to 3.4 and I do need this machine for the simple reason it has 8 threads which will increase the productivity
That motherboard is overpriced, although it offers 40$ MIB. You can find decent mainboards with huge OC-ing ability for only $100. For example the Asus P5Q/P5Q-E, the Gigabyte EP45-DS3L, etc. Or if you prefer MSI, you can get the P45 Neo-F for $94Quote:
Sorry, but that RAM is rated 1.8v. Also 2x1GB DDR3-1333 is ridiculous to put on a X58.Quote:
You can get 8GB of DDR2-800 for less money. (GSKILL 2GB DDR2-800 CL5 1.8v module = $22.99
I disagree. You can get Q6600 for $190.Quote:
So only the motherboard got a + price. What do you think about it? Is that worth?
So, here is the math:
any SB750 mainboard = $120
Phenom X4 9950 = $165
8GB DDR2-800 = $92
------------------------------------
total = $377
MSI P45 Neo-F = $95
Q6600 = $190
8GB DDR2-800 = $92
------------------------------------
total = $377
MSI P45 Neo-F = $95
Q9550 = $319
8GB DDR2-800 = $92
------------------------------------
total = $511
MSI Eclipse = $400
i7 920 = $320(which is discussable since it doesn't exist on market)
6GB DDR3-1333 = $250
------------------------------------
total = $970
Sorry, but i7 is not a feasible investment at this point of time.
A) Average is real time, the final numbers are updated when the scene finishes .... the Snow and Cave only have real meaning. 'Average' will change depending on when I press screen shot. So your first number is bogus... download Lost Planet demo to see what I mean. (Yes, it was an 8800GTX, 4870X2's were installed later)
B) Oddly, Anand's tech explanation counters your argument completely, did you comprehend what you actually linked?? In multiple GPU setups nVidia utilizes the broadcast feature of the PCIe, meaning all 3 cards gets the command set in one broadcast:
What ever traffic the CPU will send over the bus will be the same for one card as for 3 ... this is typical in broadcast networks where each bus agent is sent the same data. Also, if FSB BW is so critical, why does not cutting it in half have closer to 50% impact? You did not answer that question.Quote:
Broadcast technology allows only one message to be sent by the CPU where it is then received, replicated, and broadcasted to all GPUs, eliminating the need for multiple, near-identical transfers over the FSB.
Texture data is stored on card, and the only time bus transfers are used is when the texture required is not cached on the card. This is why memory keeps going up on video cards. When a texture is required for a new scene or object that is not in memory you will know -- that game stutters horribly (FSB, HT or even QPI) there is no bus currently available that matches the BW of VRAM to GPU...
http://http.download.nvidia.com/deve...erformance.pdfQuote:
Texture bandwidth is consumed any time a texture fetch request goes out to memory. Although modern GPUs have texture caches designed to minimize extraneous memory requests, they obviously still occur and consume a fair amount of memory bandwidth. Modifying texture formats can be trickier than modifying frame-buffer formats as we did when inspecting the ROP; instead, we recommend changing the effective texture size by using a large amount of positive mipmap level-of-detail (LOD) bias. This makes texture fetches access very coarse levels of the mipmap pyramid, which effectively reduces the texture size. If this modification causes performance to improve significantly,you are bound by texture bandwidth.
Texture bandwidth is also a function of GPU memory clock.
EDIT: Ok, so no puppet -- I wondered because you level of understanding of the concepts is on the order of Gosh's.
Tom had enough time to review its written words before the publishing date deadline, however it was concise to hear something from them about its misunderstanding.
If I could say thanks a million of times I would say to Francois who is our best source of helping us to understand the Nehalem Project as a whole including Triple channel, X58 and many more things.
lol quite fair to use a top highend board for Nehalem and el cheapo boards for all the others, also you can get 6gb ddr3 for 140$
Let me correct your nehalem rig:
MSI Platinum 295$
i7 920 320$
6GB DDR3-1066 140$
-------------------------------
755$ 20% Vat incl.
Now wait till the stuff actually gets available, and im sure you could come close to 700$. :yepp:
Isn't the comparison a bit misleading ?
Why not use a Asus Crosshair II Formula for the AMD system ? We're talking about high end enthusiast stuff.Nehalem isn't targeted at the value segment.
http://www.newegg.com/Product/Produc...82E16813131292
And you can put it in this way : no matter how much you spend on the AMD system , it won't reach the performance of the Nehalem configuration.
For what you want no, 6 Gb will not be cheap and I intend to buy when the prices drop.
I use 3D rendering programs to work which that is an undefinable multi-threading program.Quote:
Tell me the apps which you are using and that have benefit and can utilize 8 threads.
When not working I could use 1 to 4 cores to game 4 cores folding when I'm gaming and more things like group cores to each program and how would best suite it, winamp, P2P programs, server, videos, decode, encoder, I could use all at once 30 Firefox windows and so on and believe I could use up to 20 cores right now if I had a processor like that.
It is a platinum version so no, is not overpriced as I agree the platinum version does not offer much of a benefit compared to the $94 in a short term however I would not buy it for a long term, so reliability is my first step so yes the platinum version would suite me the best.Quote:
That motherboard is overpriced, although it offers 40$ MIB. You can find decent mainboards with huge OC-ing ability for only $100. For example the Asus P5Q/P5Q-E, the Gigabyte EP45-DS3L, etc. Or if you prefer MSI, you can get the P45 Neo-F for $94
Is not ridiculous as 2Gb is good enough for a starting system until DDR3 prices drop and as you already said 6GB will be expensive and I agree so no I will not pay an extra for something I will not use.Quote:
Sorry, but that RAM is rated 1.8v. Also 2x1GB DDR3-1333 is ridiculous to put on a X58.
You can get 8GB of DDR2-800 for less money. (GSKILL 2GB DDR2-800 CL5 1.8v module = $22.99
I have to say again, again and again this memory is rated at 1.8 CL6 which if you lower the voltage to 1.65 you can get 1333 CL7 cheap enough and again, again and again this was tested by many.
Yes I can get a cheaper quadcore which will not serve me well as my job is 3D rendering and for that matter I7 is my best choice and also contains + 4 logical cores which could be used by this program or something else I would like to and I7 will give me an insane 3D rendering as you can see this graph below.Quote:
I disagree. You can get Q6600 for $190.
So, here is the math:
any SB750 mainboard = $120
Phenom X4 9950 = $165
8GB DDR2-800 = $92
------------------------------------
total = $377
MSI P45 Neo-F = $95
Q6600 = $190
8GB DDR2-800 = $92
------------------------------------
total = $377
MSI P45 Neo-F = $95
Q9550 = $319
8GB DDR2-800 = $92
------------------------------------
total = $511
MSI Eclipse = $400
i7 920 = $320(which is discussable since it doesn't exist on market)
6GB DDR3-1333 = $250
------------------------------------
total = $970
http://images.anandtech.com/graphs/n...4358/17763.png
http://images.anandtech.com/graphs/n...4358/17766.png
You forgot something important and it was. "Sorry, but for me i7 is not a feasible investment at this point of time."Quote:
Sorry, but i7 is not a feasible investment at this point of time.
How could i know, i dont run such benchmarks.
So you admit those results are done with a single card much sower card then a GTX280, splendid, so you longer have no proof of anything
Have you read what i typed? I know damn well what that links contains. The broadcast feature helps reduce cpu-gpu communication over the FSB. Why do you think Nvidia has invented that feature? Because the FSB is a perfect something? Because the FSB is perfect for mutli-gpu scaling and is far from saturated?
Great minds think alike? ;)
no really, you always seems so intelligent, but right now, you are misenterpreting everything i type, i dont know if you do this on porpuse, but hey, like i care
you still havent proved your point and still havent contradicted my point
so Intel just basically copied AMD's architecture and called it their own
best quote by Intel:
"It's a bad idea until we copy it, then it's a great idea!!"
Nope. It is only you making such an assumption becouse you understanding close to nothing about AMD and Intel architectures.
@metroid
In one of your previous posts you said that by upgrading your Q9450 to i7 920 you are going "to get 10% more performance".
Quote:
Originally Posted by metroid
Err...high end enthusiast stuff with 2x1GB "el cheapo" DDR3?:rolleyes: Tell me more about it. ;)Quote:
Originally Posted by savantu
intel is on a roll, doesnt seemt to be getting slower at all, since 2006 intel is doing some amazing things for us!!
You misread my post, I do not have a Q9450 and I had not said that I have or had one, I gave 2 upgrade options, based on your reply it showed me you understood the question well, now I can not understand you at all.
Yes like I said going from Q9450 to I7 920 gives me 10% overall but do not misinterpret me as I said detailed, going from E6600 to Q9450 gives me 10% and from Q9450 to i7 920 gives me another 10% overall + 10% 3D rendering = 30% from an E6600 which means going from an E6600 to I7 920 gives me 20 to 30% overall.
So my claims were very clear since the beginning when I said I7 920 is my best option to upgrade and the motherboard is the only expensive part of this new system which is not a problem at all as if I buy a good motherboard it might last for 3 years more, until the next Intel architecture comes out again. So the price of the motherboard is the less to care as I think X58 is mature enough to drive me for 3 more years and If Westmere proves to be as good as Core 2 and Nehalem adding another 10 to 20% more performance then why not to upgrade that is generally for my workplace anyway.
I guess this is a joke, isn't it?
Yes, that's true on a superficial "let's compare the core flowcharts and call it a day" level or if you just compare the strategies of the 2 companies. Honestly, though, do you think a 731M transistor monster-chip, that neded at least 4 years engineering effort by the smartest IC-designers on earth, supported by a mutli-billion dollar company a. can be explained by some simple flow-charts b. is merely a copy of another (inferior) design?
If you think either a. or b. is true, go away troll and never come back to this board again and think before insulting some of the best engineers on the planet. :D
To paraphrase some of the smartest people that contribute to various hardware-boards: the similarities are superficial (JJ), Intel was the first company to introduce a memory controller to x86 chips anyway and in the end only the result (=performance&co) counts (linus torvalds).
On a brighter note, techreport's benchmarks show ~18% better performance including everything they tested and ~22% excluding gaming. I really like to use their datasets for comparing various systems...
Chip level power consumption is extremely low in idle, much lower than anything else by Intel (as per Tom's and lostcircuits), yet chip level power draw under load is higher than equally clocked penryn chips, but nehalem is still superior in performance/W for most applications.
On a system level, though, the power draw doesn't look that good anymore, but that was to be expected from an enthusiast level board with 1-2 more DIMMs.
here.
There is improvement, but it doesnt seem to be as near dramatic as sli's bump. Which is why the driver theory seems to hold weight, as we know in raw computational power, ie teraflops, the x2 has more.