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 :)

5k Retina iMac

Sergio Perez said:
Jeff, how about versus the current top of he line Macbook Pro? What performance difference can we see between the top of the line iMac 5K and the Macbook Pro? I have a first generation Retina, and was tempted to upgrade to the TB2 version as a creative computer for my Dragon (with Rocket X in Magma enclosure). Portability aside, maybe this could well be it, a stop gap machine until a 2016 Macbook Pro upgrade?
Click to expand...

I've got a loaded mbp... based on specs alone this iMac should be substantially faster... faster processor, more memory, much better gpu. Get it with the 1TB flash storage and you should have comparable storage performance. I'm tempted to get one for home even though $4k is twice what I normally spend on home machines. My justification is that the display alone is worth a good chunk of the price (as well as 1TB of fast flash storage)... not sure my wife will buy my story though. :)
 
According to one of my contacts at Intel, Apple made the right choice here. Broadwell CPUs for the iMac range of desktops are, if all goes well from this point, still 7 to 9 months out. :-(

I've known for some time about Apple's procurement of these 5K displays. They made their initial order over a year ago -- not joking. So it was difficult to say exactly which product and when it would happen. Manufacturing of these panels didn't go into full swing until just a few months ago. But largely because the market wasn't ready or primary purchasers were not ready for delivery. I'm betting Apple would have launched this 5K iMac 6 months ago if Intel's schedule wouldn't have fallen to pieces. Our current desktop, workstation and server platforms based on Haswell-E CPUs, the X99 and Xeon E5-v3 chips, were not really supposed to happen. These came about to fill market space for "new" products and also offered a small means of transition as far as DDR4 memory is concerned. If Broadwell would have went to fab on time, Haswell-E wouldn't exist and X99 wouldn't exist, or it might, but we would be driving 15nm process CPUs on our desktop right now and that hot new i7 would have 12 cores instead of 8... Looks like well into next year at this point now.

Thunderbolt 3 will not launch until 2016. Perhaps an early preview or special product jumping the gun at the very end of 2015 might happen if all the stars align.

Of course, This could all change inside of 6 months if Intel gets things rolling again or feels that Thunderbolt needs to be pushed ahead of schedule with the upcoming USB changes or whatever...
 
Jeff, I was never told from any Intel Channel people that Haswell E was off the books or shelved. The timelines changed some but as far as release it was never a question that I was told. Intel has been off cycle with Desktop/ Mobile and Workstation/Server since I7 Gen 2. I pretty much accepted that was on purpose due to management in house. It didn't always make sense because often times desktop would release making workstation server far less desirable but they have maintained that time frame regardless. I also havent heard of any Broadwell delays and Haswell E was accelerated enough that I wonder if there is going to be a delay.
 
Haswell-E is continued evolution of the Haswell platform and probably would have launched regardless of Broadwell status. As for Broadwell, Intel has never promised any dates, nor made any roadmap public outside of IDC 2013 when they stated Broadwell would launch for the desktop in late 2014 and for mobile platforms in Q3 2014. That is not happening. Their leaked roadmap from back in January placed the mobile Broadwell platform as launching in October, as in now. Not happening. There are prototype Broadwell ultra books out in the wild that are pretty amazing. Noticeable boosts in battery life, on the order of 15% or so and that's pretty good for a total system just based on a change of CPU. Thinner designs are possible too with the reduction in cooling needs. Delays, if you can call them that, are internal. It's not that the CPUs are "delayed", they're just not ready. And it's not an issue with the chips themselves, it's manufacture yield and refining the 15nm process and I think they're still tweaking the final CPU designs as well. As I understand it, they have quite a bit more they can still offer with Haswell if the market is up for another round or Broadwell continues to take more time.

To say that Haswell-E only happened because Broadwell is behind schedule is probably over reaching by quite a bit. OTOH, Intel has teased and promised 15nm Broadwell for two years now and outside of closely guarded prototype chips making the rounds and some prototype products from key partners, there's nothing out there for people to see from it yet.
 
Based on GPU alone, especially the BTO upgrade 4GB version of the R9 M295X, the 5K iMac would crush the 15" MBPr in pure performance. Add in faster CPU clocks and it's a rout.

Like most of us, I'm waiting for Broadwell to build out a new primary system - which, quixotically, makes this 5K iMac even more enticing. It would give me a closed loop 5K playback system that avoids the weak sauce interfaces currently available (HDMI 1.4, DP 1.2). As Mark and Jeff note, it should be very popular with creatives thanks to the hi-res screen and form factor. Moreover, any workgroup with GigE LAN and a proper IT person should be able to layoff rendering/transcoding to other machines so the iMac isn't a big bottleneck. If Intel can get TB networking sorted then it's even more viable to think of the 5K iMac as a middleweight client that can take advantage of external compute power without transit times cutting too deeply into the benefits.

Like Jeff, I prefer to think of computer resources as an amalgam of multiple machines sharing duties rather than some mind (and wallet) blowing super system that can do everything. From that perspective, the 5K iMac is attractive for the things it does very well (mostly the 5K display) rather then a fail because it's compute power is middle of the road in 2014. Prior to TB2 it could be argued that an iMac was just too limited in terms of I/O alone to be taken seriously - with TB2 and USB3 that's old news.

Cheers - #19
 
Hey guys, I know this question is not geared towards RED cameras in particular...so I apologize for that. Do you think the new iMac Retina can easily chew through non RAW 4k images....say from cameras like the new Sony FS7?
 
Old news I guess -- I hope it's not true...

"Apple won’t ship a standalone 5K display until at least 2016, and it won’t work with any of today’s Macs, including the 2013 Mac Pro."

http://www.macrumors.com/2014/10/17/retina-5k-imac-external-display-no/
 
As a creative type this works for me. An all in one unit with external storage connected via thunderbolt. I can feel an upgrade coming on.

Yes I will most probably pay more than a "Hackingtosh" or some PC based system. But I don't care, this game is all about time management and stuffing around with temperamental tech. is simply not efficient, for me anyway. Having certainty is and apple proves this time and again. Heavy rendering /outputting / workflow , that's why god invented post houses.

What this new IMac does is change my decision on upgrading to a 15" fully spec'd MP retina . Something with less horsepower for out in the field. Suddenly the 13 inch MP retina version becomes an option.

p.s thanks to everyone answers to an early post on this thread.
 
FrankMcPartland said:
Old news I guess -- I hope it's not true...

"Apple won’t ship a standalone 5K display until at least 2016, and it won’t work with any of today’s Macs, including the 2013 Mac Pro."

http://www.macrumors.com/2014/10/17/retina-5k-imac-external-display-no/
Click to expand...

Running a single 5K display is the same as running FOUR 2560x1440 displays. That is going to need 40Gbps of bandwidth -- that is TWO Thunderbolt 2 ports, or TWO DisplayPort v1.2 ports as the case may be.

The 2013 Mac Pro will in fact run the Dell 5K monitor that should start shipping next month. I'm getting conflicting information though about TB2 and DP1.2 data allocation... It takes two TB2 ports out of commission for sure, but it may actually cancel out 4 of them depending on when and where the DP bandwidth enters and exits the pipeline. There's 6 ports, but driven by 3 headers, each header capable of 20Gbps. Technically less in terms of PCIe bandwidth, but that's another line of discussion.

That said, I don't expect to see a stand-alone 5K display from Apple until they can provide a single cable connection via DisplayPort v1.3 or Thunderbolt 3.
 
We can barely run these 4K monitors at 60 hertz as it is....
 
L. Langer said:
I think Nvidia and Apple have parted ways. Some of that might be due to Nvidia clinging to it's proprietary tech, in both hardware and software, in order to drive sales of their hardware. Nvidia is very attached to CUDA and other proprietary hardware and software tech that is built into their own hardware while AMD embraces open standards and tech, despite their questionable record of doing so with Linux. Given that, it makes sense for AMD to be inside the Apple hardware.
Click to expand...

Maybe Apple is preparing to buy AMD. Face it, Apple has so much money they could do an imitation of Richie Rich's money vault. AMD has struggled for many years, but they offer a treasure trove of technology.

Maybe AMD is going to make a leap so far ahead, or they are simply keeping competition alive by swapping so they always have two suppliers to bid against each other, and hopefully not because they get a better deal on a lesser product.

Blair Thornton said:
My guess TB3 is 9 months away. Perhaps a new MacPro with DP 1.3 via USB 3.1 would be an interum solution? Who knows what Apple is planning.
Click to expand...

I wish somebody would stack out the TB port as a TB PRO version, and make DP 2 a simple version of TB (one protocol port), so you can plug your display into any new TB port and use the maximum bandwidth, or into a simple one protocole TB port (this is for OEM to make simpler hardware, not that usefull I know).

The other section of this stacking out the TB port was a standard where TB ports line up alongside each other in parrallel, so that devices can be made that plug into multiple ports in parallel. Something I had planned years ago with USB for my designs (along with a few extra tricks). This would have been a good solution for external PCIE card boxes.
 
Michael Jarvis said:
Yes....but what do you reeeeallly think. Kidding. :)
Click to expand...

They should have had the 5k touch version out years ago. Apple used to be a very desirable machine because they brought things out first, ahead of the competition, and they certainly have the money to do this for the premium models, that's how you sell them. Now they are a consumer style and convenience brand, but style is cheap, competitors can do it, substance rules. Few competitors have the money to drop down to secure an exclusive 5k run of panels (3 years ago) which would have been gold in marketing terms. They could have also had an external GPU port for an optional box to go on the back of the IMac for professionals (and bluray Mount box (both of which could be shared with the mini)). An under IMac slim storage box would have allowed people to house drives and devices etc, without all the clutter. Simple improvements.

Instead too little, too late. They could have done 4k touch 3D displays in 1998, instead they concentrated on IOS products, which now have run their past course (time to change the style) and Android is current King, though still struggling to look like old royality. I've edited several years for Apple computers to be first again. They should have released the new top model with user upgrade options on things like the CPU and interfaces (TB3, hdmi, dp1.3) so that when the shift to broadwell comes the present models can be optionally upgraded to closely match it. This is a marketing ploy to preseserve customer interest and sales.

Jeff, a number of foundaries are seeming to be finally about ready to get node processes more competitive to Intel. We maybe about to see large shifts in processing to power ratio in various products.
 
Wayne Morellini said:
The other section of this stacking out the TB port was a standard where TB ports line up alongside each other in parrallel, so that devices can be made that plug into multiple ports in parallel. Something I had planned years ago with USB for my designs (along with a few extra tricks). This would have been a good solution for external PCIE card boxes.
Click to expand...

Funny thing is that TB was originally pitched with the ability to stack and bond connections into higher aggregate bandwidth. They did it again with TB2, but have never enabled production software/ firmware to allow it. Thunderbolt 3 is supposed to use a different connector, moving away from the mini-DisplayPort connector. Even rumors saying that they may take another look at the universal, reversible USB interface. Connection of DisplayPort and legacy TB devices can be made with simple, inexpensive passive adaptors.


As far as display connections go, I would really like to see an interface spec that is future-thinking. There is no reason why HDMI 2.0 could not have been this. Go ahead and design an interface that is capable of supporting 360Gbps for 18K resolution in full 3D at 240Hz paired with 64 audio channels. Of course display and component manufacturers don't have to utilize all that bandwidth and all those features, but make it so they all play within the designated sandbox of that specification. Then in 15 years from now when 8K TV's are normal, no one has to swap out cables and matrix switching hardware and AV receivers may not offer all the new bells and whistles, but they would still be able to route and pass compatible signals.

IMO, DisplayPort has most of it right in terms of being able to carry multiple video and audio streams over one interface. The problem is they only beef up their spec once display and component hardware demand it. Instead, they should be designing a spec to aspire to. Go the extra mile and design active electronics like Thunderbolt or RedMere that can be incorporated as optional to provide long-run cabling or transport over alternative infrastructure like UTP cable installed everywhere.

Using interfaces and cabling as a means to pace market development, which is what has been going on for some time now, holds back innovation and competition. It becomes a structured advance that is dictated by the big few who always seem to be in control of the new interface specs. If that factor is removed from the equation, then competition should increase and more disruptive technology will emerge.

And don't get me started on remote control and automation systems. That whole industry and sub culture is an insult to human intelligence and common decency. With a business model that still reeks of 1970's photocopier salesmen and custom setups that should have died along with the 8 foot satellite dishes.
 
Michael, about Apple I've been long been planning to partner on a million core processing chip. A million instructions in a 20th billionth of a second maybe possible for a thousand watts, maybe less. Basic mathematics density maybe a number of times more than current processing units.

I've come up with new intra chip communications techniques. New mathematics design. I'm looking into new processing technology (not included on this design except I'm interested in an idea to virtually eliminate leakage) and a host if either things.

On the low energy side, I have an alternative design using a new way of processing. Not likely to be fast, but the present technology the others have can be made to leave ARM in the dust in power consumption, and my techniques further lower that.

I came up with design proposals for a number of personal rack mount systems, laptop, tablet and phone module systems in recent decades, and another that allowed hundreds of mini computers to be stacked up in a small space (I'm being obscure to protect IP here). Another that just allowed you to whack together a portable desktop system with hundreds and hundreds of possible communications lines (future TB might be good).

These are a fraction of the investigation and research I've done over the decades into a whole range of revolutionary paradigm shifts. All I lack is the money and a team (on my terms).

Now, if I of all people can do this (and my physician mentioned something like Wilson's disease), instead of companies earning hundreds a billions a year, I'm not the one being, stupid and lazy.
 
You must log in or register to reply here.
Back
Top