Sign in with
Sign up | Sign in
Your question
Closed

Adata's DDR4 Memory Modules Are Finally Arriving

Tags:
  • RAM
  • ADATA
  • Components
  • ddr4
  • Memory
Last response: in News comments
Share
a b } Memory
August 5, 2014 12:40:30 PM

2133-15 DDR4 when there is 2133-9 DDR3 @ 1.5V on the market... not convinced those DDR4 DIMMs will be a hit with the X99 crowd that is not exactly known for caring about saving 3-5W - particularly if it comes with any sort of performance penalty like the 66% higher latency on DDR4 would.

Well, the DDR4 transition has to start somewhere... even if the starting point seems to make little to no sense.
Score
8
August 5, 2014 12:44:03 PM

Those timings are really bad, I think we'll have to wait for high frequency dimms with 16Gb each until DDR4 is worth it
Score
1
August 5, 2014 12:55:07 PM

has it been confirmed or denied whether HaswellE CPU's will fit into the x79 socket? Or was ivy end of the line for x79. About a year ago, it was rumored the haswell-e would have a ddr3/4 controller and fit x79?
Score
1
August 5, 2014 1:00:38 PM

Well, it's about damn time! I won't be buying until DDR4 3200 is out though.
Score
2
a b } Memory
August 5, 2014 1:14:16 PM

delellod123 said:
has it been confirmed or denied whether HaswellE CPU's will fit into the x79 socket?

From what I have read, it seems like LGA2011-3 is effectively a new socket which requires x99, meaning IB-E is end-of-line for current LGA2011 motherboards.
Score
1
August 5, 2014 2:55:32 PM

the lowering of voltages is just to keep heat and power consumption in check, but if you take a DDR133 module it will consume 4 times less than a DDR4 module at 2133. Given the performance increase it still looks like a worthy compromise and in the desktop market a 20W difference is nothing.
We will have to address heat dissipation because of TSV stacking NANDs and so higher heat density. I really hope we abandon that aluminum heresy going on this days to address the problem with copper plating like Patriot Viper Xtreme 2000 9-10-9 have and they are regular height. We will probably see LP banks later and with lower capacity than the maximum 128GB unless in the meantime 20nm NANDS go down another two steps.
Score
0
August 5, 2014 2:58:57 PM

I'm considering getting the X99 with an I7-5930k (seems to be the best pick for gaming, better pcie connectivity than then 5920k and 500mhz faster than the 5960x, also more likely to overclock better) along with some descent DDR4 (As for the ddr4 comments above, there are other benefits to DDR4 than just basic ras/cas and power savings)

What do you guys think? Aim for the per core slower octa or the quicker hexa core for gaming!?
Score
0
a b } Memory
August 5, 2014 3:07:25 PM

X79 will not suppourt haswell-e only x99
Score
0
a b } Memory
August 5, 2014 3:16:57 PM

Drejeck said:
I really hope we abandon that aluminum heresy going on this days to address the problem with copper plating

Regular DIMMs already have built-in copper heatsinks: the ground and power planes in the PCB itself. The DRAM chips' packaging is made of plastic or epoxy, which are far worse thermal conductors than aluminum so the bulk of the DRAM's heat goes in the PCB through the solder balls anyway. Extend the PCB by 5-10mm, fill the extra area with copper planes and you can ditch the silly heat-spreaders altogether.
Score
4
August 5, 2014 3:36:32 PM

Quote:
2133-15 DDR4 when there is 2133-9 DDR3 @ 1.5V on the market... not convinced those DDR4 DIMMs will be a hit with the X99 crowd that is not exactly known for caring about saving 3-5W - particularly if it comes with any sort of performance penalty like the 66% higher latency on DDR4 would.

Well, the DDR4 transition has to start somewhere... even if the starting point seems to make little to no sense.


I hope you do realize that 2133-15 DDR4 will not have a 66% performance penalty over 2133-9 DDR3.
Score
4
a b } Memory
August 5, 2014 3:50:07 PM

zfreak280 said:
I hope you do realize that 2133-15 DDR4 is faster than 2133-9 DDR3.

15 cycles latency / 1066MHz clock = 14ns latency for DDR4
9 cycles / 1066MHz clock = 8.4ns latency for DDR3

That makes DDR3 40% faster on first-word latency. Bandwidth is the same since both are the same width and same data rate.
Score
4
August 5, 2014 3:52:47 PM

The DIMM problems is that most producers sports uselessly high heatsink making difficult to mount a high end air cooler and making people think that liquid AIOs are a good option. They are in the moment you don't have enough room for a high end cooler but the low level AIOs cannot compete silence, maintenance, price, performance and warranty. Only high level custom loops.
Back in topic: If you think an i7 5930k is the best pick for gaming you have been really living under a rock. They compared an Athlon 5350 with a 750 Ti with a 4960X, a 25W TDP cpu against a 130 or 150W TDP cpu. One is 4 cores 4 threads the other 6 core 12 threads and the results were pretty impressive giving the Athlon only have 4 PCI lanes and the 4960X has 10 times more. There are a few games that like to have more processing power, but almost all depends on GPU power. Now the options available are 256 bit mid-low to high level GPUs, 128 bit low level GPUs and 384 and 512bit extreme level GPUs. Well the bits that should enable us to the 4K gaming world simply are not working as expected, and the 384 bit aswell. While you cannot call yourself a gamer at that level of money spent but just a jocker who spent loads of money on things that will get old really soon it's better to address what I and some wise people call "diminishing return". You are really going to spend loads of money on a processor that will initially be starved by the ram, giving it will have a legendary good controller, you will get initially the same amount of ram as nowadays PCs and lower performance (it depends on the ram controller but the latency is really huge).
You are going to get a CPU fast and powerful like the Q9550 was, even the Q6600 do as an example, overclockable and one of the most powerful in the market just for a few fps more, if any, and get obsolete by the day that many threads will be used. Well, the wallet is yours, I cannot tell you anything else than this two words:

DIMINISHING RETURN
Score
0
August 5, 2014 4:21:25 PM

Quote:
zfreak280 said:
I hope you do realize that 2133-15 DDR4 is faster than 2133-9 DDR3.

15 cycles latency / 1066MHz clock = 14ns latency for DDR4
9 cycles / 1066MHz clock = 8.4ns latency for DDR3

That makes DDR3 40% faster on first-word latency. Bandwidth is the same since both are the same width and same data rate.

Quote:
Quote:
2133-15 DDR4 when there is 2133-9 DDR3 @ 1.5V on the market... not convinced those DDR4 DIMMs will be a hit with the X99 crowd that is not exactly known for caring about saving 3-5W - particularly if it comes with any sort of performance penalty like the 66% higher latency on DDR4 would.

Well, the DDR4 transition has to start somewhere... even if the starting point seems to make little to no sense.


I hope you do realize that 2133-15 DDR4 will not have a 66% performance penalty over 2133-9 DDR3.

You cannot calculate the real performance loss (bacause nowadays there is no gain) because even if it's an Haswell architecture the ram controller may behave differently. I think the performance penalty is more mathematical than real world experience, but it's there. We have not yet seen what Vpp boost does so the first word access latency could be addressed in this way. So we cannot really compare DDR3 with DDR4 because they work differently and hopefully much better with the latest having an advantage even with NAND chips apparently weaker.
Note that even with lower voltage, the DDR4 are going to consume more power than the same DDR3 memory.
We all hope that the semiconductor problem gets addressed by smart logic.
Score
-1
August 5, 2014 4:21:49 PM

Quote:
zfreak280 said:
I hope you do realize that 2133-15 DDR4 is faster than 2133-9 DDR3.

15 cycles latency / 1066MHz clock = 14ns latency for DDR4
9 cycles / 1066MHz clock = 8.4ns latency for DDR3

That makes DDR3 40% faster on first-word latency. Bandwidth is the same since both are the same width and same data rate.

Quote:
Quote:
2133-15 DDR4 when there is 2133-9 DDR3 @ 1.5V on the market... not convinced those DDR4 DIMMs will be a hit with the X99 crowd that is not exactly known for caring about saving 3-5W - particularly if it comes with any sort of performance penalty like the 66% higher latency on DDR4 would.

Well, the DDR4 transition has to start somewhere... even if the starting point seems to make little to no sense.


I hope you do realize that 2133-15 DDR4 will not have a 66% performance penalty over 2133-9 DDR3.

You cannot calculate the real performance loss (bacause nowadays there is no gain) because even if it's an Haswell architecture the ram controller may behave differently. I think the performance penalty is more mathematical than real world experience, but it's there. We have not yet seen what Vpp boost does so the first word access latency could be addressed in this way. So we cannot really compare DDR3 with DDR4 because they work differently and hopefully much better with the latest having an advantage even with NAND chips apparently weaker.
Note that even with lower voltage, the DDR4 are going to consume more power than the same DDR3 memory.
We all hope that the semiconductor problem gets addressed by smart logic.
Score
-1
a b } Memory
August 5, 2014 5:13:00 PM

Drejeck said:
So we cannot really compare DDR3 with DDR4 because they work differently

Differently? The basic operating principles and signaling scheme for DRAM has remained nearly unchanged since capacitor-based DRAM chip was invented. With custom PCBs and firmware, it would probably be possible to use LPDDR3 interchangeably with DDR4.

The CAS, RAS, RD, Precharge, command, etc. latencies mean exactly the same thing on any memory from the ancient Fast Page Mode asynchronous memory to DDR4; the only difference is that it is expressed in terms of clock cycles on synchronous memory.

On a fundamental basis, all mainstream SDRAMs are awfully similar.
Score
1
August 5, 2014 6:04:12 PM

come on you guys with a move to new ddr we will always have higher latencies dont you remember the move from ddr2 to ddr3
Score
1
August 5, 2014 6:08:53 PM

Nice to know they're working on speed AND power consumption. As much as I really couldn't care less in my desktop, if DDR4 is ever going to make it to laptops, tablets, and phones, then they need to optimize for that from the get go.
Score
1
a b } Memory
August 5, 2014 6:25:56 PM

daekar said:
if DDR4 is ever going to make it to laptops, tablets, and phones, then they need to optimize for that from the get go.

DDR4 getting everywhere is not an if; it is merely a question of when: DDR3 production will start ramping down after Skylake comes out, ramp down even faster once AMD announces their own DDR4 chips, add maybe two more years and DDR3 will be almost history.

The transition will take about four years... just like the DDR1-DDR2 and DDR2-DDR3 transitions before it.
Score
0
August 5, 2014 6:46:39 PM

Intel's DDR4 systems are also quad channel while DDR3 is dual channel or triple channel so while the latency is higher currently the bandwidth is higher and more channels can be accessible simultaneously.

We will have to just wait and see how it plays out in actual real world test comparisons and benchmark results. Obviously these DDR4 timings will change a lot in quick fashion anyway and the voltages are quite a bit lower. They'll likely lower the timings and bump up the voltages a bit over time and probably boost the clock speeds as well as it matures.
Score
0
August 5, 2014 6:49:15 PM

I don't think the mainstream market will have anything worth looking at until 2016.
The increase of latency is history repeating and as a matter of fact, in each jump the first new modules were slower than the older and they consume more. DDR4 2133 consume 4 times more a DDR 133, and only the same capacity modules will consume less, but DDR4 introduces Vpp at 2,5V and its logic seems to improve memory usage so without a benchmark to tell us how much I think DDR4 consume more than DDR3 and is going to consume a lot more with higher density memory. The latency is always a concern because we never broke the 7.5 ns wall. Data moves a lot faster on higher frequencies ram and controllers have been designed to take advantage more of frequencies than latencies so while I consider both for quality in the end only benchmarks will tell us how the new logics uses DDR4.
Score
0
August 5, 2014 6:58:24 PM

DDR3 is dual, triple and quad channel.
The triple channel is dead by quite some years...
We have i7 E processors sporting quad channel controller, some motherboard with dual memory per channel. While theoretical bandwidth should be higher there's no gain over the dual channel DDR3 when you consider the same architecture. DDR4 doesn't support dual memory per channel, that's why it has high density. It support switching which isn't the same thing and doesn't improve performance, it's made for servers who have 16 ram banks for 4 controllers but you can only use one memory per channel at a time.
Score
1
a b } Memory
August 5, 2014 7:01:15 PM

knowom said:
Intel's DDR4 systems are also quad channel while DDR3 is dual channel or triple channel so while the latency is higher currently the bandwidth is higher and more channels can be accessible simultaneously.

LGA2011 is quad-channel DDR3 and LGA2011-3 is quad-channel DDR4.

The first mainstream Intel CPU with DDR4 support is Skylake and will still only support dual-channel on LGA1151.

So there are no changes in the number of channels going from DDR3 to DDR4... mainstream remains dual and enthusiast remains quad.

The triple-channel memory was on the ancient LGA1366 socket for first-gen Core i7 XXX series.
Score
0
August 5, 2014 10:06:08 PM

Hopefully better timings and clockspeed will occur in the future. I have seen a 128gb ddr4 somewhere. Since its the beginning of ddr4, we may except more down the line.
Score
-1
August 5, 2014 11:08:45 PM

can anyone explain why we are on ddr4 when graphics cards already use ddr5? why not just jump straight to 5? Cheers.
Score
1
August 6, 2014 12:31:45 AM

Quote:
can anyone explain why we are on ddr4 when graphics cards already use ddr5? why not just jump straight to 5? Cheers.


Are different things. Graphic cards use GDDR5, not DDR
Score
1
August 6, 2014 1:24:07 AM

Quote:
I'm considering getting the X99 with an I7-5930k (seems to be the best pick for gaming, better pcie connectivity than then 5920k and 500mhz faster than the 5960x, also more likely to overclock better) along with some descent DDR4 (As for the ddr4 comments above, there are other benefits to DDR4 than just basic ras/cas and power savings)
Personally I am aiming for octo core and do a mild/average OC on it, Reason is Direct X 12 next year will finally be utilizing more cores for games instead of typically loading most of the work onto just 1 of the cores so it should balance out. Either one(5930 or 5960) you cant go wrong but I would skip the 5920 if you will be considering multiple gpu's and maybe adding an M.2 SSD drive onto those pci-e lanes to share that bus.
Bit of a shame on the DDR 4 speeds/latency but it has to start somewhere, Also next gen gpu's will be another round of 28nm, Seems like next year will be when all the good shit gets sorted out,Most likely have slightly updated/faster Haswell-E's by then as well.

What do you guys think? Aim for the per core slower octa or the quicker hexa core for gaming!?

Score
0
August 6, 2014 1:23:42 AM

Quote:
Quote:
can anyone explain why we are on ddr4 when graphics cards already use ddr5? why not just jump straight to 5? Cheers.


Are different things. Graphic cards use GDDR5, not DDR


For clarification, Graphics memory is "Dual ported". What this means is that data in GDDR can be simultaneously written to by the GPU and read by the video driver. This is different than normal DRAM in which a block can only be read or written to at a given time.
Score
1
August 6, 2014 1:27:00 AM

-
Score
0
a b } Memory
August 6, 2014 5:31:24 AM

jgs5093 said:
For clarification, Graphics memory is "Dual ported". What this means is that data in GDDR can be simultaneously written to by the GPU and read by the video driver.

GDDR is not dual-ported. There are two major differences between DDR and GDDR:
- GDDR uses Quad-Data-Rate signaling on the data pins, which is made possible mainly because GDDR signals do not need to go through socket interfaces
- GDDR has separate read and write data strobe signals to enable faster turn-around when switching data bus direction while DDR has bidirectional strobe pins which introduce one cycle of latency to switch signal driver between ends

Conventional dual-ported memory went practically extinct over 15 years ago because dual-ported memory itself was much more expensive (more expensive to produce due to the extra complexity and also a niche product so much fewer sales to spread development costs on) and having dedicated read and write data ports also doubled the pin cost on the video chip.

The concept of dual-ported access is still alive and well but mostly in (semi-)custom form... like Intel's Crystalwell 128MB L4 cache for Haswell with GT3e, Broadwell and Skylake.
Score
0
a b } Memory
August 6, 2014 6:55:30 AM

If the timings were so much important then why DDR5 memory is being used in Graphics card.

Ofcourse when 1600 CL6 RAM is compared with 1600 CL9 RAM then the former will be winner but if 1600 CL6 is compared with 2133 CL9 then the latter one will be faster attributable to its higher frequency.

DDR4 is something new & better and should be taken positively.
Score
0
August 6, 2014 6:59:58 AM

Does anyone knows if DDR4 have the same form factor than DDR3?
Score
0
August 6, 2014 7:03:08 AM

my 2 year old Hyperx Predator is 2400-11 DDR3, so there is literally no need to jump the ship, this ram will be slower. not massively, but slower
Score
0
a b } Memory
August 6, 2014 7:54:25 AM

maktovic said:
If the timings were so much important then why DDR5 memory is being used in Graphics card.

DDR5 does not exist. GPUs use GDDR5.

The reason why GPUs can work with GDDR5's 20+ cycles latency is because streaming texture data from memory is far more bandwidth-intensive than latency-sensitive. The order in which data will be requested for most GPU-optimized tasks is known well in advance so the GPU can easily hide the latency with prefetching.

For a main CPU on the other hand, guessing which pieces of data will be needed next is far more difficult since the code running on CPUs is so much more diverse and traditionally much less linear.
Score
0
August 6, 2014 9:00:47 AM

Hoping for 16GB NON-REG-ECC ram sticks. Need for ramdisk of course.
Score
1
August 6, 2014 3:13:20 PM

wonder what the prices on those could be !!
I'll wait till G-Skill release their kits on the market and then maybe I'll buy a decent x99 rig ...say 5930k ...64 GB Ram ... Rampage V if released .... and a GTX 880 if it comes before Christmas ..
Score
-1
a b } Memory
August 6, 2014 4:28:28 PM

hardcore_player said:
wonder what the prices on those could be !!

Previous generations started with a 50-100% premium over mainstream memory during most of their first year... so I would not be surprised if the first generation of 16GB DDR4 kits started around $200-250.
Score
1
August 28, 2014 2:57:10 AM

Quote:
2133-15 DDR4 when there is 2133-9 DDR3 @ 1.5V on the market... not convinced those DDR4 DIMMs will be a hit with the X99 crowd that is not exactly known for caring about saving 3-5W - particularly if it comes with any sort of performance penalty like the 66% higher latency on DDR4 would.

Well, the DDR4 transition has to start somewhere... even if the starting point seems to make little to no sense.
I agree i have OC'ed my memory to 1866 at Cas 9 and is still at 1.5 V DDR4 has no point right now
Score
0
!