so basically the saem advise we've been giving people on the forums for the past 4-5 months.. if you need a system now go for it, if you can wait sandybridge is right around the corner (as is bulldozer) and i'll be wating till both are out and benchmarked to buy one
I'm still unsure how Sandy Bridge will outpace X58 systems. From what I've read (and maybe misunderstood) it seems like the two CPU families are very similar other than a puny GPU built into the CPU. For enthusiasts, is this really that big of a deal?
IMHO - Sandy bridge may be good for sub 800$ systems, since you won't need a separate GPU, but anyone looking to push the limits wouldn't benefit.
Can't they say that about just every new generation of procs? I mean... in theory, each get is about double the one before it. I know this isn't as true as it once was. My point is that "P2 represents the largest increase in computing performance in our history"... "P3 represents the largest increase in computing performance in our history"... "P4 represents the largest increase in computing performance in our history"... etc, etc... should hold true in general.
[citation][nom]chiral[/nom]I'm still unsure how Sandy Bridge will outpace X58 systems. From what I've read (and maybe misunderstood) it seems like the two CPU families are very similar other than a puny GPU built into the CPU. For enthusiasts, is this really that big of a deal?IMHO - Sandy bridge may be good for sub 800$ systems, since you won't need a separate GPU, but anyone looking to push the limits wouldn't benefit.[/citation]
There are going to be SB based systems that will not have a GPU built onto the die, specifically the K series that has a unlocked multiplier.
In terms of performance, Anand had a review with some SB CPUs in a sort of Alpha stage and it was pushing 10-20% overall better performance than a equally clocked Core i7. With driver improvements and final silicon it could be higher than that. The SB to truly replace X58 based Nehalems though will come in late 2011 in X68 based systems on LGA2011.
[citation][nom]dylansaliba[/nom]Ramp shmamp, just keep trickling the tech down, milking us for every bit that we are worth...[/citation]
Yea if thats how you feel, you should not be a PC enthusiast. Technology always gets better. Should I be mad that LGs Ally is such a nice phone over the EnV Touch? Nah. I am glad. But you can't get everything for free.
hmm... I remember when they added the FPU to the CPU with the 486... on graphics and specific apps, it made a HUGE difference or the first Pentium chips... if this is bigger, then it alone may not give us a huge leap, but future versions will.
This is complete BS. There will be only a minor improvement (10-20% at best) with Sandy Bridge. Think about it...It uses the same 32nm process and has the same number of cores, so the improvement will be lack luster.
I think that the real jump in performance will come when Ivy Bridge is released on the 22nm process. Notebooks will have 4 cores with 8 threads, and desktops will have 8 Cores with 16 Threads. Wait until Ivy Bridge or you will be very, very sorry.
I switched to AMD for the Phenom II. While I have no doubts that Sandy Bridge will be a marked improvement, what I really want is AMD to get their stuff together to compete with Intel. Sandy Bridge will be big in the $200 - $300 (from what I've been led to believe) range and AMD isn't ready to compete with their current lineup. Which is sad. I'm agnostic as to AMD vs Intel, but if AMD can't compete that's bad for everyone.
I plan to build my next rig in the Spring, around the time I get my extortion refund. If Sandy Bridge performance is only 20% better, I won't care. I will care that it will lower the prices of perfectly acceptably-performing i5 and i7 CPUs.
Then there's Bulldozer. I'm optimistic that Bulldozer will at least match i5/i7; since it will undoubtedly cost less, very likely that's the way I'll go, even if Sandy Bridge is 20% faster.
Of course, I could be pleasantly surprised and Bulldozer will equal or beat Sandy Bridge, while still costing less. That would be sweet indeed.
If you keep stacking 5% increases then each increase is bigger than the last. "The biggest performance leap ever" is a marketing scheme.
Now, I still look forward to it but the pricing is always a concern with Intel. If AMD can even at least match i7 with Bulldozer but with their usual prices of considerably lower than Intel I would still be interested. I'd rather not spend 50% more for 10% more performance and end up with an unbalanced system.
Not all of Intel promises are good..I wait on my money anyway.Even if it does enter the market it would cost at least 1000USD.Making an expensive market price at first always gets the consumers attention.Then they slowly make it lower.Such as Intel i7 950.Last year it was 500USD now its around 270USD in one year.
Sandy Bridge is a completely redesigned chip, not a Pentium Pro derivative, so it's going to have a nice improvement in single-threaded apps, for sure.
But, his claim is absolutely ludicrous. The 286 was an incredible improvement over the 8086 (and yes, it was a successor, the 186 was made in parallel with the 286 and had different design goals). Not only was it over three times faster per clock cycle, it increased addressable memory by 16 times, and added virtual memory and allowed multitasking operating systems to be effectively created for it.
Call me crazy, but I just don't think the Sandy Bridge is going to be over three times faster per clock cycle. This is not to downplay the significance of the chip, since Intel finally left the Pentium Pro behind (again), but his statement is just not correct, or even close to it.
[citation][nom]the_krasno[/nom]If this is as big as a leap like the one from Pentium to Core, I might have to give them my money. I still use an old Core 2.[/citation]
Actually, I wouldn't bet on it, although ironically, they used a lot of technology from the Pentium 4 in this chip.
The reason the increase was so big from the Pentium 4 to the Core 2 was because the Pentium 4 was so bad, as well as the Core 2 being excellent. Evidence of that is the relative performance to AMD processors, as well as IBM's. The Nehalem is a very good processor, which makes that type of improvement much more difficult to improve to the same extent.
Still, Sandy Bridge is a full redesign, so will be good. I hope you're right, but I think it might be a bit too much to expect considering the excellent performance of Nehalem based processors.
I need a new machine for january, but I don't think Intel will have anything SB ready then. On the other hand 3.4 Ghz with turbo at 3.8 with new architecture improvements looks really good, but never 2x-3x. I read somewhere clock for clock SB without hyperthreading to be as powerful as Nehalem with HT. It will be something like future SB i5 family outperforming current i7, and of course i7 being even faster.