Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

New Mac Pros are here!

BASSAM MSSALATIE said:
I CAN not under..stand anything it look's Helogrephic :ranting2: (first typing letters belongs to old egyptians).to me will anyone simplify it please
Click to expand...

"Buy even numbered sets of identical ram. Preferablly in sets of 4 or 8 when possible."
 
BASSAM MSSALATIE said:
hope it will gonna be helpfull thing.did you nitice anything now ? Hard's
performance improvement ? or after building your raid..
Click to expand...

I haven't used my new drives yet. Still waiting for my machine to get here on Thurs.

I usually zero out all my new drives and have never had one fail. The only time I've ever had a failure is with one of my old Medea VideoRAID RT's. But that had a bunch of drives in hardware RAID, and I was able to recover most of my data.
 
Does anybody have an idea when some real 8cores would come out. Giving us a total of 16 cores in one machine???? I did some research and I remember some roadmaps, but I suppose they changed after intel even did a delay on some of their chips.
 
Forget more cores. Why are we still stuck at about 3Ghz? My first G5 many years ago was a dual 2.5Ghz. Now I just bought a brand new 3.0Ghz machine? Only 1/2 Ghz faster 4 years later! That's just not right. I remember reading a Motorola processor road map in 2002 that said we'd be at 12Ghz+ by now. Not to mention, weren't there 4Ghz Pentiums like 3-4 years ago?

It seems like the industry has hit a Mhz wall and is compensating by tacking on more and more processors. I personally don't like this trend as it requires software to be rewritten to take advantage, plus there are a lot of tasks that are just too serialized to multiprocess well.

It seems that Moore's Law is coming to an end.
 
jeremyn said:
Forget more cores. Why are we still stuck at about 3Ghz? My first G5 many years ago was a dual 2.5Ghz. Now I just bought a brand new 3.0Ghz machine? Only 1/2 Ghz faster 4 years later! That's just not right. I remember reading a Motorola processor road map in 2002 that said we'd be at 12Ghz+ by now. Not to mention, weren't there 4Ghz Pentiums like 3-4 years ago?
Click to expand...

Physical limitations of our technology. Silicon / Copper / Aluminum chip construction, seems to have an upper speed limit of about 4GHz. And it becomes harder and harder to manage this as complexity increases. With circuitry sizes shrinking as tech advances, (now at a 45nm process), this reduces the energy consumption and heat dissipation, which theoretically allow for faster operational speeds. However, the smaller circuitry increases complexity and it becomes harder to manage power or noise transfer between the smaller circuits. We will probably see 45nm tech progress to 3.4~3.6GHz, but that's about it for current CPU designs.

The "MHz wall" was one of the factors which finally pushed chipmakers into multi-core designs. It's still going to be some time before software truly takes advantage of the multi-core paradigm, but some of that is also due to shortcomings with current designs. Most software designs are very linear or serial in nature. Tasks must be parallelized or broken up into cellular blocks to be distributed out on multiple processors. With only 4 to 8 cores in current systems and relatively low-speed bus architectures, it's hard to realize a truly cellular approach. Increasing MHz was the brute-force approach to increasing performance. We're now at the technological limit where all the chipmaking materials that people have come up with to use in practical manufacturing just won't allow us to push more electrons any faster.

It seems like the industry has hit a Mhz wall and is compensating by tacking on more and more processors. I personally don't like this trend as it requires software to be rewritten to take advantage, plus there are a lot of tasks that are just too serialized to multiprocess well.

It seems that Moore's Law is coming to an end.
Click to expand...

Yep. And don't worry about Moore's Law. Moore's Law is nothing more than marketing FUD. As it was originally written, it never applied -- only as a hypothetical possibility of what could be. Moore originally stated that over the next decade (at the time he wrote it) microcircuit manufacturing processes should allow for the number of transistors on a silicon wafer to be doubled every 18 months. It had nothing to do with processing power or transistors per cm, Hertz or anything of that sort. "Moore's Law" as been spun and re-spun countless times over the years to make it sound like Moore and the rest of his companion geniuses at Intel predicted the future with 100% accuracy. Only 2 years ago, Intel had a Moore's Law page on their site talking about how Moore was wrong and they had exceeded Moore's Law by huge margins and AMD sucks and all that sort of stuff. Now they're back to spouting about Moore being a genius and through their persistence of excellence, they have managed to up-hold what Moore envisioned or whatever the current FUD is. I actually haven't read it in the past 4 or 5 months, but it's funny to read new versions of their Moore's Law page every now and then.
 
You must log in or register to reply here.
Back
Top