Slashdot Mirror
Intel's Haswell-E Desktop CPU Debuts With Eight Cores, DDR4 Memory
crookedvulture writes: Intel has updated its high-end desktop platform with a new CPU-and-chipset combo. The Haswell-E processor has up to eight cores, 20MB of cache, and 40 lanes of PCI Express 3.0. It also sports a quad-channel memory controller primed for next-gen DDR4 modules. The companion X99 chipset adds a boatload of I/O, including 10 SATA ports, native USB 3.0 support, and provisions for M.2 and SATA Express storage devices. Thanks to the extra CPU cores, performance is much improved in multithreaded applications. Legacy comparisons, which include dozens of CPUs dating back to 2011, provide some interesting context for just how fast the new Core i7-5960X really is. Intel had to dial back the chip's clock speeds to accommodate the extra cores, though, and that concession can translate to slower gaming performance than Haswell CPUs with fewer, faster cores. Haswell-E looks like a clear win for applications that can exploit its prodigious CPU horsepower and I/O bandwidth, but it's clearly not the best CPU for everything.
Reviews also available from Hot Hardware, PC Perspective, AnandTech, Tom's Hardware, and HardOCP.
*drool*
'nuff said.
I'm still clunking along on a P4 3.8 GHz. I'd love a new box that fast!
I do not fail; I succeed at finding out what does not work.
At least the one review I looked at said it was only $1K.
Troll is not a replacement for I disagree.
I recently (less than a year ago), bought an i7, four core, 8 thread machine. I use it a lot for chess analysis, and it is amazing how quickly it can get to a 24 ply deep analysis. Even with a slightly slower clock, 8 cores would be so much quicker.
See my journal for slashdot ID's by year. Mine created in 2005. http://slashdot.org/journal/289875/slashdot-ids-by-year
But does it run Linux?
Tubby or not tubby. Fat is the question
until next year. 14nm shrink should be a huge boost in both efficiency and performance.
The x99 is an "enthusiast" platform and has pricing along those lines.
DDR4 is also extremely new. Expect it to get faster/better timing specs as time progresses.
The 5820K is packing 6 cores and an unlocked multiplier for less than $400. If you don't absolutely need the full 8-core 5960X, then the 5820K is going to be a very powerful part at a reasonable price for the level of performance it delivers.
AntiFA: An abbreviation for Anti First Amendment.
Lets hope so. DDR3 has always been a joke, since it gained speed over DDR2 when configured in 3 channel banks. Except it is almost never configured that way, and thus resulted in faster clocked DDR2. Hopefully DDR4 works appropriately when configured in a 4 DIMM bank.
But will it run Crysis?
Jesus was all right but his disciples were thick and ordinary. -John Lennon
That is the problem with Apple's obsession with small and sexy. Of course if Apple updates the Pro this year all is good but given their history I would not bet on it.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
if you can wait then you should always wait for new tech
That's just nonsense. Just because there are taller buildings doesn't make the Empire State Building any smaller.
Tubby or not tubby. Fat is the question
No one is talking about the elephant in the room: RAM prices are so high that you'd have to spend $700 to hit 64GB RAM (the max the board supports). That is just outrageous.
These prices are going to lead to a severe drop in demand.
If you are a rich mofo, you don't use Intel at all!
Oh, what are the rich folk buying instead?
For 'how many lanes of NSA Bulldozer 7.0" (is that the latest?) we'll just have to wait for the next Vanunu or Snowden.
CAS latency hasn't been measured directly in nanoseconds for some time now. It is now measured in clock cycles. The shorter your clock cycles (the higher your frequency) the shorter in absolute time your CAS latency is for the same number. CAS 10 at 2133 is about the same as CAS 5 on 1066.
CAS latency on Wikipedia
Memory timing on Hardware Secrets
FAQ on RAM timings from Kingston
DDR is not about the number of channels. You could design a system with 8 channels DDR1 or single channel DDR4 if you want to. New generation DDR RAM is always about lower voltage and higher clock speed. Usually at the cost of higher latency (800 MHz DDR3 is a bit slower than DDR2)
Just to put "some time now" the time frame into perspective, the last mainstream PC memory form-factor to use asynchronous DRAM was 72 pin SIMMs.
When PCs went from 72 pin SIMMs to the first 168 pin DIMMs, in the mid-1990s, the interface changed to (non-DDR) synchronous clocking.
PJRC: Electronic Projects, 8051 Microcontroller Tools
They buy TWO Intels.
But if I'm trying to game on an old i5-750, wouldn't this be a good time to upgrade to one of the cheaper 4-core Haswells that are running 3.8mhz instead of 2.7? Maybe a Haswell i5 (I guess, I'd need a new mobo then, right?) And the latest PCI-E for a new graphics card.
I don't like to buy the newest and best, but when the second newest becomes cheap. I've got a really nice case, but I'm not sure if I could put a new processor into my old motherboard or if it would even be worth it.
I'd like to do something before the fall games come out. Would I be better off just upgrading my old Radeon HD6850 to a nvidia 760 or a Radeon R9 285 or something?
And did I fall through a wormhole and end up at Tom's Hardware?
You are welcome on my lawn.
Come on. The Mac Pro requires you to spend big bucks. It's not too much to ask Apple follow Intel's roadmap with the Mac Pro.
8 of 13 people found this answer helpful. Did you?
Mac Pro uses Xeon E5 v2. It's much more likely that a (small) upgrade is going to use Xeon E5 v3. But hey, don't let me stop you from celebrating the upcoming death of Apple. Parties that go on for decades must be great ;-)
Perl Programmer for hire
That's right. The Empire State Building is Majestic, and Quaint and stuff. Like the Mac Pro.
As a gamer, I have been on a 3 year PC build cycle since 1992. Every three years (more or less), I build a new PC. Since 2008, I've only upgraded once and my current build (now 2 years old, a 2500k overclocked to 4.4 Ghz) feels pretty darnn fast still. It's weird because 15 years ago, I'd be itching for new hardware with a 2 year old system. Since my Core2 Duo build in 2008, I haven't really seen any noticeable performance jumps other than the move to SSDs and the bigger IPS panels. My old Core 2 system, which is now 6+ years old and on my workbench in the garage can still hold it's own with new games without a problem.
Obviously there is a lot more to PC/game perfomrnace than just the CPU. But what I'm getting at, is ever since the move to multi-core platforms in the mid-2000s and offloading more work to the GPU(s), hardware development for PCs seems to have slowed down quite a bit. Maybe a combination of the newer hardware not being utilizaed as well as stuff back in the ewarly 2000s and smaller bumps up in the various chip/processor design? Perhaps the greater focus on power savings and mobile development? The trend towards netbooks, and now tablets? Maybe the fact that most kids these days only play on consoles?
I dunno. It just thinks like the rate of change has slowed down a lot over the last decade. Or maybe I'm just getting old.
Can anyone please tell us why is there a need to slow down the CPU speed in order to put in more cores?
Thermals. More CPUs generate more heat, more heat with the same thermal envelope means you can't run each CPU as fast. Of course in a different environment (say with a liquid nitrogen cooling rig vs an air cooled rig), you could probably clock those CPUs higher.
Just because you can put in more CPUs doesn't mean you should. It used to be the limiting engineering factors were area vs chip yield. Now days thermals are arguably the most important consideration because often you are limited both thermally (and sometimes even electrically) to the amount of power you can deliver to a square millimeter of a computer chip.
I use -- and write -- image processing software. Correct use of multiple cores results in *significant* increases in performance, far more than single digits. I have a dual 4-core, 3 GHz mac pro, and I can control the threading of my algorithms on a per-core basis, and every core adds more speed when the algorithms are designed such that a region stays with one core and so remains in-cache for the duration of the hard work.
The key there is to keep main memory from becoming the bottleneck, which it immediately will do if you just sweep along through your data top to bottom (presuming your data is bigger than the cache, which is typoically the case with DSLRs today.) Now, if they ever get main memory to us that runs as fast as the actual CPU, that'll be a different matter, but we're not even close at this point in time.
So it really depends on what you're doing, and how *well* you're doing it. Understanding the limitations of memory and cache is critical to effective use of multicore resources. You're not going to find a lot of code that does that sort of thing outside of very large data processing, and many individuals don't do that kind of data processing at all, or only do it so rarely that speed is not the key issue, only results matter. But there are certainly common use cases where keeping a machine for ten years would use up valuable time in an unacceptable manner. As a user, I am constantly editing my own images with global effects, and so multiple fast cores make a real difference for me. A single core machine is crippled by comparison.
I've fallen off your lawn, and I can't get up.
Interesting essentially how little benefit they get.
The X99 mobo and platform is nice, I like a lot of what they're doing there, and all of the system components matter a lot to user experience. But unless you have a very specific requirement any user would be just as well served with a quad core or a octa core, if not better served with the devil's canyon quad core given the single threaded performance. That's probably a bad place for intel to be positioning these, as the target audience for these processors is looking for blazing fast and lots of cores. And it only delivers one of the two.
I think if I was buying a system this week or next (which... I am) I'd be a bit disappointed that I can't put a devil's canyon quad core on an X99 mobo, and then upgrade the CPU later if they manage to refresh the E series into something more attractive.
for non-multithreaded games?
> Failure to full exploit SMP in 2014 is a fine reason to avoid a given game as far as I'm concerned.
That's a crappy reason. You'll miss out on Path of Exile, Minecraft, and Terraria, all which are excellent games.
When will the newer gen of Intel chips go DDR4? also what about AMD?
broadwell is not going to have DDR4.
Also why can't the DMI link be better in other cpu's
Why do have to now get an 6 core to use Hasswell-e to get more then 16 pci-e 3.0 + X4 pci-e 2.0 (DMI)
Most people may only need 1 Video card but with pci-e SSD's coming out more pci-e is needed.
8-core processors?
Over clocked POWER chips in liquid nitrogen.
Cwm, fjord-bank glyphs vext quiz
I do a lot of FPGA programming and It takes me 15-20 minutes to synthesize a design on a modern fast computer. As more of the part is being used, synthesis takes more and more time, as the chip becomes harder to rout.. I'm a user that is primarily CPU bound. I hope that Intel will continue to push on the raw performance. For the past few years, as we've only seen marginal improvements in CPU performance.
There is also the issue that FPGAs keep getting cheaper/bigger, so no matter how fast your rig, it always takes a long time to synthesize. I'd be curious about what other FPGA developers use to boot performance.. Overclocking/water cooling does seem to help, as does using faster ram.
Have you looked at RAM prices? 32GB of DDR3 RAM is about $300-400 for a 4x8GB set, depending on speed and company. So $600-800 for 64GB. Ok well how about server memory, since you can get servers with 6TB of RAM if you like (really, check HP or Dell). For a 16GB DIMM, which is the largest you can get before the price per GB skyrockets, it is about $160-200. fo $640-800 for 64GB.
So hmmm, looks like DDR4 is right in what other ram costs, plus a bit of a premium since it is brand new tech. What a shock! Who would have every thought it would cost about what RAM costs!
Get off it. Also it is stupid to act like everyone would need to buy the max amount of RAM. That the system SUPPORTS 64GB doesn't mean you have to BUY 64GB. It means that if you need that much, you can have it. If you need less, get less. Most desktops sold today support 32GB in the form of 4 sticks of 8GB DDR3 RAM. Most systems ship with only 4-8GB of RAM, in 1 or 2 sticks. There is nothing stopping you from using less.
You see this even more on the server market. We like Dell R720XDs at work. They support 768GB of RAM. However 0 out of 5 that we have purchased have that much RAM. It is exceedingly expensive, since it needs 32GB DIMMS. However it also means that getting 384GB is much cheaper, since it has the ability to do that on 16GB DIMMS. That said, we have only one system that needs that much RAM. The rest? Between 128-256GB. The rest of the slots sit empty, ready to be filled as our needs grow. Two of the 128GB servers will probably be getting more memory soon.
So seriously, get off it. DDR4 really isn't much more expensive than DDR3, much less than I thought, and memory is cheaper than ever. All these boards mean is if you need a lot of RAM, you can have it.
(citation needed)
I have never seen RAM as cheap as it is now. When you can buy a 16GB ECC DIMM for less than $200, it is rather wonderful. Our researchers that use big amounts of memory are extremely happy with how much memory they can stuff in desktops and servers for a reasonably price.
Now I'll admit, I don't have a chart of RAM prices, so I suppose I could be wrong, but then I've worked in IT for the last, oh, 20ish years on a continuous basis and spec'ing and buying hardware is a fairly common part of my job.
So please, show me some evidence from two years ago when RAM was half its current price. Right now I see a 16GB 1600MHz 2R ECC DIMM as running about $170, and a 4x4GB 1600MHz unbuffered set running about $150. So please show me some proof that two years ago I could get those for about $70-90 each.
They run the same benchmarks. It's a lot of copypasta from Intel's marketing material. Boring. How many enthusiasts are helped by a photo of the chip with the cores labelled?
From the reviews I could not figure out whether vPro or the virtualisation bits were turned on.
vPro: no
VT-x: yes
VT-d: yes
Source: http://ark.intel.com/products/...
In computer technology, there is ALWAYS something new next year. Yes, there'll be a 14nm shrink next year (or maybe later this year)... but then just a year away will be a technology update, a new core design that is more capable, and of course they'll have more experience on the 14nm process and it'll be better... however only like a year after that 10nm will be online and that'll be more efficient.
And so on and so forth.
With computers, you buy what you need when you need it. Playing the "Oh something better is coming," game is stupid because it is always happening, generally very quickly.
So if you want a 6 or 8 core system, this is what to buy (it's cheaper than their Xeon setups). Will there be a better ones later? For sure. However sitting in neutral waiting for "the next big thing" is silly. Get a system, keep it as long as it is useful, get a new one when you need a new one.
Also hating on this for being enthusiast is silly. Ya it is expensive. So don't get it if you don't need it. However for what it does, it isn't bad. Maybe you need that kind of power. Maybe you need more. Not long ago we had a faculty member purchase workstation with 2x 12 core CPUs. These things cost about $2600 PER CPU, never mind the other hardware to support it. System was over $10,000. However, for the simulations he was doing, it was worth it. I'd never buy that for home, my workloads are much lighter, but I'm not going to hate on him needing it.
Same shit here. Do most users need this? No. Heck most users don't need a quad core. But there are uses for it.
AMD, and IBM have both been talking about stacked designs for cache memory, Intel has been a big player in HBM/FCRAM development, and AMD, ARM, and others are throwing a lot of weight behind HSA, even Intel is bringing in some of the idea's of HSA at least as far as unified cpu/gpu virtual memory address spaces are concerned. The next 2-3 years is going to be transformative for computing, languages and software libraries will need to catch up with not just with macro threaded concurrency, but also with micro threading concepts. The convergence of "large enough" caches something like Iris Pro but with real cache memory instead of edram, HSA making igpu a first class citizen(think if opencl had access to the programs heap/stack, aka being able to call virtual functions, checking type information, accessing arbitrary objects not directly passed in the functions parameter list), and hopefully HBM/FCRAM will finely catch memory speed up at least for a year or 2(it'ill never last but here's hope'n lol).
So I'm only going to have the ability to use two GPUs in SLI?
Nope. Fuck that.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
"If it's for gaming, then the CPU isn't really a bottleneck like it's made out to be and there is no gain in going i7 over an i5 unless you are going to be streaming."
Man you're so full of shit I can smell you through the internet. First, the i7 has more PCI-E lanes, which translates over to "I can drop in more GPUs if desired."
Streaming shit is all dependent upon the framebuffer access now days - GPU. Not CPU.
It's like people don't understand how hardware acceleration works.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
"4-core Haswells that are running 3.8mhz"
Only 3.8 MHz? Wow, that's fucking slow. Glad I went AMD instead of intel!
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Maybe you could make an FPGA to synthesize faster than a CPU?
https://mojang.com/2014/08/minecraft-1-8-pre-release-the-bountiful-update/
These changes consist of both new features, and large game structure changes such as replacing the hard-coded "block renderer" with a system that is able to read block shapes from data files, or performance enhancements such as multi-threading the client-side chunk rendering. We hope you will enjoy it!
Considering Minecraft has been out for 3 years and they have sold 54+ million copies it is about time they addressed some of the performance issues!
CAS latency hasn't been measured directly in nanoseconds for some time now. It is now measured in clock cycles.
Yah, so to compare two different sticks of RAM you have to multiply the time/cycle by the number of cycles. Which gives you (wait for it....) time!
Which the parent did, to point out that all these "new" memory technologies haven't been decreasing the RAM latency much at all. RAM latency is still a _VERY_ important part of overall execution performance. Particularly for single threaded operations reading RAM in unpredictable manners. Cache misses are overwhelmingly the single largest optimization variable for modern applications.
Haswell-e does not use the same socket or chipset as Ivy-Bridge E does. It will take an completely new motherboard for the Pro to move to Haswell-E. Look now long Apple let the the old macpro sit after Intel moved to Sandy Bridge and Ivy Bridge.
As to Apple dying I really hope not. I love OSX and my macbook. Apples continued obsession with thin and small in markets that do not really need it is what I hate. Had Apple come out with a nice tower and simply used a standard EATX or ATX motherboard for the Pro they would have had a better product.
Of course you still have the problem with Thunderbolt. Making it both storage and a video connector is dumb. It would have been better as just a storage connector IMHO so standard video cards could be used.
BTW A thunderbolt connection is no where near as fast as a PCI3 3.0 8x or 16x slot so do not even bother with the external card argument.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
And the indy car that won the 1974 Indy 500 is just as fast but new indy cars are faster. The Boeing 707 can still fly passengers across the atlantic it is just that the 767 can fly more people cheaper.
The Mac Pro was a good deal when it came out compared to other workstations in it's class. So you will have the opertunity to buy machines for the same price that are faster. The Apple way is okay for consumer products. "Honestly I hate the imac because they traded the ability to easily upgrade the SSD for thin". It does not work IMHO for a workstation class machine.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
They did not with the old Mac Pro.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
5820K should be enough for anyone?
Well, I keep seeing clock rates go up and high-end DIMMs keep having CAS numbers like 9 and 10 despite the rate going from 1333 to 2400 the past few years. (1/2400000000)*10 is 4.16666666666667e-10 while (1/1333000_00)*10 is 7.50187546886722e-10 which looks like a relatively major difference to me.
If you are a rich mofo, you don't use Intel at all!
Oh, what are the rich folk buying instead?
Processors hand-made by artisans from individual valves/vacuum tubes.
Of course, you need a rather large house to hold the 1.4 billion valves required to match something like the Core i7. Well, actually you need a rather large estate with enough room to build a large number of very large buildings, and a literal army of support staff to replace the failed valves.
Trust me though, it's worth it for the additional warmth the use of valves lends to playing back your Nicki Minaj MP3s.
Then again, that warmth might just be coming from the hundreds of megawatts of waste heat given off...
"Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).