Intel Demos DDR4 Reference Test Platform

[RedundantInk]

Intel's Haswell-E powered "DDR4 Reference Test Platform" made a brief appearance at this year's Intel Developer Forum.

Intel Demos DDR4 Reference Test Platform : Read more

 

[bombebomb]

So, a DDR4 Platform exists, and is running fine. Great.

 

[Amdlova]

so no more updates... DDR4 is on the way, maybe pci x 4.0 too

 

[CaedenV]

I am really excited about DDR4 coming down the line for quite a few reasons.
1) In mobile it is going to allow for super high densities with extremely low power, making 4GB of ram in a cell phone and 8GB in a tablet very possible. This will fix a lot of the multi-tasking and productivity issues that these devices suffer in. The performance of DDR4lp will probably be no faster than what we have today in low end desktops, but the power savings will allow for much better active-idle power use to vastly extend up-time on a charge.
2) DDR3 in the desktop space has seen horrible scaling from 800 to 1866, mostly due to Intel's memory controller architecture. Hopefully DDR4 will force them to rethink their architecture in order to gain the performance necessary to entice people to upgrade. Hardly anyone upgrades for added CPU horsepower anymore, these days upgrades are all about fixing system bottlenecks, be they in memory, SATA, PCIe, or USB. DDR3 has been a known bottleneck for quite some time, so hopefully DDR4 breaks that barrier.
3) DDR4 may (and I understand that this is wishful thinking) be fast enough to use as an extension for graphics memory. Next gen games and 4K monitors are going to dictate that high end GPUs have 8+GB of graphics memory... which would make these devices prohibitively expensive! The hope is that system memory would finally be fast enough, and that next gen controllers would be intelligent enough, to allow the GPU to have direct access to system memory as an extension to their own internal GDDR, which would allow next gen GPUs to continue to use 2-4GB of GDDR, while having fast access to the 8+GB that they really need. I only replace system memory with each core upgrade every 5-6 years, while I replace my GPU every 2-3 years. It is much more justifiable to buy a metric crap ton of system memory every 6 years than it would be to buy 8GB of very expensive GPU memory every 2.

Anywho, good to hear that DDR4 is coming down the pipe! Late 2014 it will make its way into high end desktops, 2015 will be right on track to upgrade my desktop, and then we will see it in phones within a year after that.

 

[thundervore]

I waited for USB3, PCIe3 and DDR3 to all be incorapted on the same Z77 motherboard before i upgraded from my IP35 motherboard with USB2, PCIe2 and DDR2. I will wait until there is USB4, PCIe4, and DDR4 all on the same motherboard before i spend another penny to upgrade my system. I do not even care if i have to wait until 2020

I'm sure i am not the only one with these thoughts.

 

[eodeo]

What are the benefits of ddr4? From what I understand, ddr3 did ever so little to improve anything other than benchmark numbers over ddr2.

 

[InvalidError]

I am really excited about DDR4 coming down the line for quite a few reasons.

There aren't that many reasons to be excited about DDR4...
1- lower power and higher densities are possible with DDR3 too with minimal specification bending but with DDR4 just around the corner, the industry is choosing to wait for DDR4 instead of add extra rungs to the DDR3 ladder.
2- the reason Intel CPU performance scaling does not scale much with massive bandwidth increase is simply because SB/IB/Haswell do not need anywhere near that much bandwidth to max out their processing resources - if you look at memory controller benchmarks, Intel's MC actually kicks ass on both latency and bandwidth.
3- system DDR4 on a 128bits bus and 3GT/s is still only 1/4th as fast as 2GB GDDR5 with 256bits and 6GT/s on the GPU card or 1/8th as much as high-end cards with 512bits memory bus, which is nowhere near as fast so you would expect a 75-90% penalty from transferring large chunks of data from system RAM instead of local GPU RAM. Also, accessing system RAM as an extension of the GPU's own means a ton of latency going through the PCIe bus to reach the CPU's memory controller which would further degrade performance.

The thing that will benefit most from DDR4 is IGPs across the whole spectrum from mobile SoCs to desktops CPUs. Low-end GPUs using DDR4 would also get modest boost from it but not as much since they already push DDR3 clocks close to DDR4 territory.

 

[nolarrow]

@Thundervore

I did the same exact thing. I plan to do the same as you... unless I need more juice to run star citizen at 120hz 1080p maxed out with an occulus rift, at least the game wont be out for another year or two.

 

[blubbey]

The thing I'm looking forward to is the performance with APUs.

 

[ingtar33]

intel will wait for AMD's AM4/FM3 platform to be right around the corner then release their DDR4 platforms.

Right now Intel isn't pushing anything in the chip market, they'll wait for AMD to be in the game then release first. The longer they can sandbag tech the better for their bottom line.

 

[InvalidError]

The longer they can sandbag tech the better for their bottom line.

There is nothing to "sandbag" here.

DDR4 is a minor technology improvement and since Intel's current CPUs come nowhere close to bottlenecking on their DDR3 controller under typical workloads, the only reasons they would have to bother with DDR4 are reducing platform power and increasing bandwidth for their IGPs. Two things Intel will do sooner or later regardless of what AMD does.

 

[ingtar33]

The longer they can sandbag tech the better for their bottom line.

There is nothing to "sandbag" here.

intel has a multi billion dollage R&D budget, yet they've come up with no significant improvement in their platform/chip design and performance in 4 years. There is a reason for this. As long as AMD exists they are not a monopoly and free from anti-trust laws, and have no fear of being broken up. Intel has nursed AMD AND OTHERS along for most of it's existence for the expressed purpose of avoiding those anti-trust laws.

Right now AMD can't compete with 4 year old intel chips, if they get too far ahead of AMD and kill it in the CPU business (something they're very close to doing) they'll be open to anti-trust hearings and potential breakup. So they HAVE to sandbag any tech which might put them too far ahead of AMD.

You think it's coincidence they're artificially handicapping their haswell/ivybridge cpus with poor design on their heat-spreader/thermal paste? They don't use that thermal paste on their extreme edition chips... so it's not some limit of die size, it's because they need to keep the pressure off AMD.

They do the same with other tech as well. They need AMD to remain a viable alternative, so they will continue to work on energy efficiency and shrinking the size of their cpus while keeping performance the same, because their real competition is ARM right now. So they'll re-position themselves so that their desktop and mobile products are one and the same, so that they can knock off AMD and still point to ARM as their competition, but until there is no difference between their desktop and mobile lineup they won't kill AMD.

Side-note, it's because of AMD and ARM's continued existence that we can pay affordable prices for cpus. as long as intel remains in indirect (or in ARM's case, direct) competition with them the consumer will win. when AMD is no longer in the picture, the desktop landscape will shift drastically from what is good for us to what is good for intel's bottom line.

Remember that.

 

[InvalidError]

intel has a multi billion dollage R&D budget, yet they've come up with no significant improvement in their platform/chip design and performance in 4 years.

And neither has AMD for the most part. If improving performance was so easy, AMD should be closing the gap with Intel much more quickly. The real reason that justifies the stagnation is the lack of any mainstream killer app to force people to demand faster CPUs. There isn't much motivation to depreciate high-end parts into mainstream segments when 90-95% of PC demand can be met with relatively low-end parts - Intel sells lots of dual-core i5 and i7 in the low-power and laptop markets.

As far as DDR4 giving Intel a "performance advantage", Intel's chips have shown only marginally significant performance scaling from DDR3-1066 to DDR3-2133 so it is extremely unlikely that DDR4-3200 would yield any significant gains CPU-wise. Actually, the higher latency on initial DDR4 may actually be a handicap just like the higher launch latencies on DDR2 and DDR3 were.

DDR4 is the wrong tree to bark at.

 

[ingtar33]

And neither has AMD for the most part. If improving performance was so easy, AMD should be closing the gap with Intel much more quickly. The real reason that justifies the stagnation is the lack of any mainstream killer app to force people to demand faster CPUs. There isn't much motivation to depreciate high-end parts into mainstream segments when 90-95% of PC demand can be met with relatively low-end parts - Intel sells lots of dual-core i5 and i7 in the low-power and laptop markets.

As far as DDR4 giving Intel a "performance advantage", Intel's chips have shown only marginally significant performance scaling from DDR3-1066 to DDR3-2133 so it is extremely unlikely that DDR4-3200 would yield any significant gains CPU-wise. Actually, the higher latency on initial DDR4 may actually be a handicap just like the higher launch latencies on DDR2 and DDR3 were.

DDR4 is the wrong tree to bark at.

AMD has plenty of motivation to innovate and improve. 10% market share & bleeding money. They're sinking fast, and need a hit. Intel hasn't really improved much of anything since their last gen core2duo chips except for the incredible memory controller breakthrough on nehalam for the first gen core i, and some moderate ipc improvements every cycle as the architecture reaches it's peak maturity.

 

[SuperAxilla]

"I waited for USB3, PCIe 3 and DDR3 to all be incorapted on the same board before i upgraded from my Q6600 board USB2, PCIe2 and DDR2 board. I will wait until there is USB4, PCIe 4, and DDR4 all on the same board before i spend another penny to upgrade my system.

I am sure i am not the only one with these thoughts."

Same here. Although USB 4.0 isn't such a factor for me personally, I just really want to see where Intel (and AMD) will go with DDR4. A mobo capable of 64GB of DDR4 would be my "holdout", and whether or not Intel will go past 6 cores on the 2011 socket.

 

[tului]

Can't wait to see a dual socket Haswell EP Xeon. DDR4, AVX2, and maybe some new chipset goodness. Wouldn't mind seeing SAS 12. Regardless, it'll be time to build a new video rig soon.

 

[InvalidError]

AMD has plenty of motivation to innovate and improve.

They do.

But at the same time, they have almost no margin they can afford to sacrifice to claw market share back so unless software catches up with CPUs to push demand for much faster mid-range CPUs to compensate thinner margins with volume (this used to be the driving force behind price wars until Intel crushed AMD with Core 2 and Core i5/i7), AMD is stuck at trying to catch up on performance.

If a couple of mainstream killer apps came along, AMD would be able to make a decent profit selling ~$150 CPUs instead of $50-100 ones and Intel would likely need to rearrange their performance and pricing ladders to accommodate a major shift in mainstream performance requirements and competitive price points just like they often used to do 10 years ago when AMD was often leading benchmarks and many everyday tasks were still slow enough for most people to desperately desire faster PCs.

Most of today's mainstream software grossly under-uses modern CPUs so we have a large and growing chunk of the market that simply does not care how much faster new desktop and laptop CPUs are anymore. We need new killer apps to reset that.

 

[ingtar33]

If a couple of mainstream killer apps came along, AMD would be able to make a decent profit selling ~$150 CPUs instead of $50-100 ones and Intel would likely need to rearrange their performance and pricing ladders to accommodate a major shift in mainstream performance requirements and competitive price points just like they often used to do 10 years ago when AMD was often leading benchmarks and many everyday tasks were still slow enough for most people to desperately desire faster PCs.

Most of today's mainstream software grossly under-uses modern CPUs so we have a large and growing chunk of the market that simply does not care how much faster new desktop and laptop CPUs are anymore. We need new killer apps to reset that.

i agree 100% generally speaking i can't tell the difference between a late core2duo, a phenom II x4 965be (overclocked none the less) and an i5-3570k. they all feel basically the same in 95% of your day to day stuff. the limiting factor for all of them is the hard drive. put an ssd in ANY of those computers and it would feel like the superior cpu, frankly, i rarely advise people with core2duos or later cpus to update their systems... i just suggest they get an SSD for the biggest day to day performance boost of them all. If all 3 of those example systems had an ssd, and i had a stopwatch, then yes, i could tell them apart. otherwise unless we're gaming something that is single threaded and heavy on the cpu you won't be able to tell the difference.

 

[ojas]

Just a note, Scot Wasson of TR found this booth too, it was actually Kingston showing off DDR4, not Intel, and they didn't tell him which platform, though it was a dual socket design with quad-memory channels.

 

[elsom23]

Have fun trying to buy laptop or gaming pc what support ddr4 in 2014 because thats coming in 2015 or later retards

 

[InvalidError]

Have fun trying to buy laptop or gaming pc what support ddr4 in 2014 because thats coming in 2015 or later retards

Mainstream PCs are only one part of the global market for RAM. Servers like Xeons that use FBDIMMs can be converted to DDR4 by simply creating an FBDIMM to DDR4 bridge chip to replace the existing FBDIMM to DDR3 bridge chips.

Smartphones, tablets and other low-power SoC applications will likely be next on the DDR4 train since the lower power draw and bandwidth will be worth the extra few bucks for mid-range devices and higher.