Intel 9th Gen Review
anandtech.com
anandtech.com
It seems people are OK with Intel stating a 95W TDP when its max could be 225W. We knew TDP wasn't a reliable figure for a long time and had sort of expect 20 - 30% variant, but now we are looking at ~2.4x difference....
After all the lies Intel have told, after all the disappointment and anger that Intel has been sitting there not innovating, cost cutting in the simplest thermal measures, their near irresponsible sales pitch to enterprise, the cries for AMD to bring competition.
And they still buy an Intel Chip.
The lesson here is brand building and stickies. I could only wish AMD best of luck. I am waiting for Zen 2.
In the end, it seams to me like the intent is to deceive.
https://ark.intel.com/compare/186604,52214
Kinda though I would see a lot bigger differences. There was a time in my life when I expected I would have to build a new computer every two years. Weird.
The i9-9900k is way better. 3x 256-bit execution ports, 64-byte (512-bit) messages going around the Ringbus, DDR4 RAM (double memory bandwidth).
The raw clocks ignore the huge advancements in IPC that exist between Sandy Bridge and Coffee Lake. And even then, the 5GHz clock on the i9-9900k makes it completely incomparable to compare the ancient Sandy Bridge with the modern Coffee Lake.
Ancient? Yes. Incomparable? No.
Still, its reflective in the naming. Netburst, Willamette, Northwood, Prescott, and Tejas were all called "Pentium 4", and lasted from a time period from year 2000 to 2006.
Sure, there were a few additions: Hyperthreading I guess. But otherwise, the main difference in this timeframe were the 250nm to 90nm scaling (which got "free" GHz upgrades back then without any effort)
In short:
> what we used to call the 'annual upgrade.'
You're exaggerating. Progress has slowed but Willamette (year 2000 1.5GHz) vs Northwood (year 2002 2GHz) Pentium 4s weren't that dissimilar.
Prescott (2004) is even more amusing, as it hit 3.8GHz (Intel Pentium 4 570J). But boy, oh boy... you really can't compare Prescott 3.8GHz against Sandy Bridge (2011, 3.6GHz) or Coffee Lake (2018, 3.6GHz ), even if the base clocks are all close.
I've been sitting on an old i7 Sandy Bridge precisely because Intel hasn't been offering me a compelling reason to upgrade. I'm probably going to finally do it next year... but it won't be to another Intel system.
2. We have already proved should the SandyBridge Overclocked to 4GHz+ it would have perform faster in some cases than Skylake Uarch with same Clock Speed. Remember SandyBridge was 32nm.
3. "Netburst, Willamette, Northwood, Prescott, and Tejas " You made it sounds as if there were many generation. All of them are part of Netburst uArch and Willamette was the first Gen, Tejas was cancelled. So there is only "Willamette, Northwood, Prescott". Each with much improved Clock speed, better implementation of HT, better memory controller and cache read improvement, and finally AMD64, all translate to substantial performance improvement, regardless of its power usage.
Netburst may not be a good uArch, but it definitely brings many improvement, the writing was on the wall in 2004 and we got Core 2 in 2006. And we have been stuck with Quad Core Desktop CPU despite we have had over 4x transistor density improvement.
All those don't matter if they hardly make a 10-15% difference in everyday use, and require specialized code to take advantage of.
But my work (statistics) seems poorly represented in benchmarks and the community at large. Ryzen is in the later group, and gets rave reviews.
While most press I've seen talking about avx512 has been negative. Yet I get a lot more flops out of the 10-core i9 7900x than the similarly priced 16-core Threadripper 1950x in simulations.
A big part of that is probably doing most of my work in Julia, where because the code you run is compiled locally the first time a function/argument type combination is called (a) code will be generated for all your CPU's features, and (b) it is easy to encourage aggressive optimizations to vectorized code by the compiler, eg by parameterizing arrays by their bounds.
R benchmarks will not see benefits at all, as much of R's math libraries are terribly unoptimized -- the default BLAS doesn't even use any runtime dispatch! (Thankfully it's easy to change, but the benchmarks all reflect that reality -- and most people's personal experiences too).
Similar likely applies to many NumPy benchmarks, but depending on the distributor of those binaries, the situation likely isn't better. Either generic unoptimized binaries, or shipping with something like Intel MKL which, while the fastest linear algebra library for Intel processors, dispatches to slow code paths for AMD processors -- a very unfair choice for benchmarking. I'm less less familiar with NumPy though, so perhaps there's a booth fair and well optimized distribution.
I fear the future will move against my use cases.
But a CPU has latency benefits. Most noticeably, it takes only a few clock cycles for an Intel CPU to transfer data from its RAX register -> AVX vector registers, compute the solution, and transfer it back to the "scalar world".
This level of data-transfer is done within single-digit nanoseconds.
In contrast, any GPU memory transfer takes hundreds-of-nanoseconds to microseconds... an order of 100x to 10,000x slower than transferring between Scalar-code and AVX-code on a CPU.
----------
So GPUs are good for bulk compute (big matrix multiplies, Deep Learning, etc. etc.)... but I'm definitely interested in the low-latency uses of SIMD.
For example: Consider a SIMD-based Bloom Filters to augment a Hash Table (or any data-structure, really). Multi-hash systems like Cuckoo-hashes have also been successfully implemented with CPU-SIMD acceleration.
You can use CPU-accelerated SIMD to accelerate CPU-based algorithms. And CPUs certainly can benefit from fatter, and better designed, SIMD instruction sets.
unfortunately, it takes millions of cycles for the CPU to switch to the higher-power state necessary for AVX instructions, so your latency really isn't that much better.
the radfft guy did a good writeup https://gist.github.com/rygorous/32bc3ea8301dba09358fd2c64e0...
But I've found getting into GPU computing extremely difficult. Admittedly, part of that may be having bought an AMD Vega graphics card instead of NVidea (not a fan of vendor locking, and AMD is building an open source stack with ROCm/HIP).
My code does lots of different things, many small, but given a billion iterations, it can add up. If you're running a Monte Carlo simulation over millions of different data sets, fitting each iteration with Markov Chain Monte Carlo, with a model that has a few long for loops and requires (automatic) differentiation and some number of special functions... It's not hard to rack up slow execution times out of small pieces.
Part of the problem is I need to actually find a GPU project that's accessible for me, so I can gain some experience, and learn what's actually possible. How I could actually organize computation flow. It's easy with a CPU. Break up Monte Carlo simulations among processors. Depending on the model being fit, as well as the method, different computations can vary dramatically in width. But optimizing is as simple as breaking up all those computations into the appropriately sized AVX units (and libraries + compilers auto-vectorization will often handle most of that!), so wider units directly translate to faster performance.
Part of the problem is I don't really know how to think about a GPU. Can I think of the Vega GPU as a processor with 64 cores, a SIMD width of 64 (64 * 64 = advertised "4096), and more efficient gather loads/scatter store operations?
If there were a CPU like that, compiler + library calls and macros (including libraries you've written or wrapped yourself) go a long way so you can quickly write well optimized code.
I really need to dedicate the time to learn more. My 7900x processor is about 4x faster at gemm than my 1950x, but 10x slower than the Vega graphics card that cost less money. I see the potential.
To start, I need to find a GPU project that lets me really get experimenting and figure out how programs can even look.
Is there an Agner Fog of GPGPU?
From my limited experience, that's the wrong way of looking at things.
SIMD is good at looking at "converged" instruction streams, and bad at "divergent" instruction streams.
"Converged" is when you've got your bitmask (in AMD / OpenCL: the scalar bitmask register) executing all 64-threads as much as possible. This means all 64-threads take "if" statements together, they loop together, etc. etc.
"Diverged" is when you have a tree of if-statements, and different threads take different paths. The SIMD processor has to execute both the "Then" AND the "Else" statements. And if you have a tree of "if / else" statements, the threads have less-and-less in common.
-------
That's it. You try to group code together such that as much code converges as possible, and diverges as little as possible. It helps to know that the work-group size on AMD can be anywhere from 64 to 256 (so keep your "thread groups" together as large as 256 at a time).
The OpenCL compiler will automatically translate most code into conditional moves and do its best to avoid "real if/else branches". But as the programmer, you just have to realize that nested if-statements and nested-loops can cause thread-divergence.
---------
Matrix maths are the easiest for SIMD, because of how regular their structure is. When you get to "real" cases like... well...
> My code does lots of different things, many small, but given a billion iterations, it can add up. If you're running a Monte Carlo simulation over millions of different data sets, fitting each iteration with Markov Chain Monte Carlo, with a model that has a few long for loops and requires (automatic) differentiation and some number of special functions... It's not hard to rack up slow execution times out of small pieces.
Okay, well that's why its hard to program properly on a GPU. Because they're all diverging. Unless you figure out a way to "converge" these if statements, it won't run on a GPU correctly.
Chances are, if you have a million-wide Monte-carlo, a lot of those threads are going to be converging. Can you split up the steps and "regroup" tasks as appropriate?
Lets say your million threads run into a switch statement:
* 106,038 threads will take path A.
* 348,121 threads take path B.
* 764 threads take path C.
Etc. etc.
Can you "regroup" so that all your threads in pathA start to execute together again? To best take advantage of the SIMD-architecture?
Think about it: a gang of 64 may have thread #0 take PathA, thread#1 take PathB, thread #2 take PathA, etc. etc. So it all diverges and you lose all your SIMD.
But if you "rebatch" everything together... eventually you'll get hundreds of threads ready for PathA. At which point, you gang up your group of 64 threads and execute PathA all together again, taking full advantage of SIMD.
SIMD is an architecture that allows "similar" threads to run at the same time, in groups of 64 or more (up to 256 threads at once). And they truly execute simultaneously as long as they are all taking the same if/else branches and loops. The real hard part is designing your data-structures to maximize this "convergence" of threads.
Something like Chess would practically be impossible to SIMD on a GPU. Its just impossible to handle how divergent the typical chess analysis engine gets due to the precision of chess board positions. But something like Monte-Carlo surely will have hundreds, or thousands, of threads taking any "particular execution path". So I'd have hope for a SIMD-based execution on Monte-Carlo simulations.
You still need some understanding of gpus, but at least you can write your code in Julia rather than CUDA or OpenCL.
Concretely, I’ve used this for parameter sweeps of large systems of ODEs and SDEs.
Maybe. Sparsity of the matrix in question matters a lot. Matrixes with very little sparsity, like say an image, do well. Others may not. It’s more like GPUs do well on algorithms with predictable branching.
The future will probably include more avx512, but not yet.
Video Editing, Video Games, Graphics, 3d Modeling, Photoshop. Even Stockfish Chess uses new instructions (not SIMD: but the Bit-board popcnt and pext / pdep instructions) to accelerate chess computations.
At a bare minimum, AVX grossly accelerates memcpy and memset operations. (Setting 256-bits per assembly instruction instead of 64-bits per operation is a big improvement). And virtually every program can benefit from faster memcpy and faster memsets.
"Standard Software" isn't written to be very fast. But anything that's even close to CPU-bound is being upgraded to use more and more SIMD instructions.
How often are memcpy and memset CPU-bound, though?
L3-cache on Skylake architectures has 18-bytes-per-clock sustained, which is still greater than 128-bits per clock bandwidth to your L3 cache.
Soooo... I'd say roughly for any memset or memcpy smaller than 256kB or so... you're going to benefit to an AVX-based memset or memcpy.
At ~8MB or so, where you're hitting L3 cache, it probably is a benefit but not much of one. The important thing is Skylake only has one store unit, so you can only write once per clock cycle.
So... do you write one 64-bit value, or do you write one 256-bit (32-byte) value?
----------
To be fair: I'm pretty sure every compiler's default settings today outputs SSE-based memcpy and memsets (128-bit). So AVX doubles that to 256-bit.
Still, it seems like the documentation elsewhere says that AVX 128-bit doesn't cause any clocking issues. So AVX512 (applied to 128-bit) should still lead to good speeds without any clocking problems.
void memcopy(double* restrict a, double* restrict b, long N){
for (long n; n < N; n++){
a[n] = b[n];
}
}
gcc -march=skylake-avx512 -O2 -ftree-vectorize -mprefer-vector-width=128
-shared -fPIC -S memcopy.c -o memcopy128.s
gcc -march=skylake-avx512 -O2 -ftree-vectorize -mprefer-vector-width=512
-shared -fPIC -S memcopy.c -o memcopy512.s
I'm using gcc 8.2.1. I believe the option `-mprefer-vector-width` was added recently.
Now I compiled both into shared libraries (drop the `-S` and choose an appropriate file name), and benchmarked from within Julia.
julia> using BenchmarkTools, Random
julia> memcopy128!(a, b) = ccall((:memcopy,
"/home/chriselrod/Documents/progwork/C/libmemcopy128.so"),
Cvoid, (Ptr{Cdouble},Ptr{Cdouble},Clong), pointer(a), pointer(b), length(a));
julia> memcopy512!(a, b) = ccall((:memcopy,
"/home/chriselrod/Documents/progwork/C/libmemcopy512.so"),
Cvoid, (Ptr{Cdouble},Ptr{Cdouble},Clong), pointer(a), pointer(b), length(a));
julia> b = randn(32); a = similar(b);
julia> b = randn(32); a = similar(b);
julia> all(a == b)
false
julia> memcopy128!(a, b); all(a == b)
true
julia> randn!(a); all(a == b)
false
julia> memcopy512!(a, b); all(a == b)
true
julia> @btime memcopy128!($a, $b)
7.517 ns (0 allocations: 0 bytes)
julia> @btime memcopy512!($a, $b)
5.277 ns (0 allocations: 0 bytes)
julia> b = randn(64); a = similar(b);
julia> @btime memcopy128!($a, $b)
10.980 ns (0 allocations: 0 bytes)
julia> @btime memcopy512!($a, $b)
7.300 ns (0 allocations: 0 bytes)
julia> b = randn(100); a = similar(b);
julia> @btime memcopy128!($a, $b)
15.060 ns (0 allocations: 0 bytes)
julia> @btime memcopy512!($a, $b)
14.031 ns (0 allocations: 0 bytes)
julia> b = randn(200); a = similar(b);
julia> @btime memcopy128!($a, $b)
33.867 ns (0 allocations: 0 bytes)
julia> @btime memcopy512!($a, $b)
19.546 ns (0 allocations: 0 bytes)
julia> b = randn(400); a = similar(b);
julia> @btime memcopy128!($a, $b)
53.923 ns (0 allocations: 0 bytes)
julia> @btime memcopy512!($a, $b)
28.376 ns (0 allocations: 0 bytes)
julia> b = randn(800); a = similar(b);
julia> @btime memcopy128!($a, $b)
98.277 ns (0 allocations: 0 bytes)
julia> @btime memcopy512!($a, $b)
57.960 ns (0 allocations: 0 bytes)
julia> b = randn(2_000); a = similar(b);
julia> @btime memcopy128!($a, $b)
232.138 ns (0 allocations: 0 bytes)
julia> @btime memcopy512!($a, $b)
133.229 ns (0 allocations: 0 bytes)
julia> b = randn(20_000); a = similar(b);
julia> @btime memcopy128!($a, $b)
3.688 μs (0 allocations: 0 bytes)
julia> @btime memcopy512!($a, $b)
2.759 μs (0 allocations: 0 bytes)
julia> b = randn(200_000); a = similar(b);
julia> @btime memcopy128!($a, $b)
112.598 μs (0 allocations: 0 bytes)
julia> @btime memcopy512!($a, $b)
117.437 μs (0 allocations: 0 bytes)
Of course there's a performance advantage from using 512 bit registers for a memcpy - but a memcpy is rarely a major performance bottleneck by itself and is usually surrounded by other code. Unless that code is also AVX-512, you've just made it slower by optimizing the memcpy. My point was that a compiler can't usually decide whether it's worth making the optimization in light of the broader context.
The other point was whether using AVX-512 while sticking to xmm registers is faster than just using xmm SSE/AVX code. I don't have an AVX-512 capable machine at the moment, perhaps you'd like to check if your 128 bit version is any faster than just doing "gcc -march=skylake -O2 -mprefer-vector-width=128" (thereby retaining the microarchitecture optimizations, but sticking to AVX2)?
Not necessarily, for large operations and depending on processor generation a simple rep stos/movsb will be simpler (no alignment requirements) and saturate your memory bandwidth just as well as any AVX sequence will with less icache pressure.
Seems to be a feature in Ivy Bridge and later, which happens to be around the time AVX2 started.
You forgot the most important: i9-9900k has TWICE the number of cores as i7-2600k.
And besides what we're seeing now and why CPU's are viable for much longer is they got extremely powerful and user demands for more power win general computing kind of tapered off. I mean nowadays you can run multiple programs, a browser with dozens of tabs, multiple other utilities and background task pretty easily. You can run way more stuff then a person can reasonably multi-task. You have to get into some pretty specific workloads to really utilize all the power a modern CPU has and be wanting more.
And yeah the performance difference between generations has been pretty meager, but over 5, 6, 7 generations those modest improvements do add up. Going form a 2600k to a 8700k or 9700k is a big leap. But if you're not desperate for a lot more power you just might not care about 50% more performance.
On CPU Benchmark the i7-2600K scores an 8449 [1], while the i7-8700K scores a 15970 [2]. The i7-8700K is the predecessor to the i7-9700K reviewed in the article.
[1] https://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-2600K...
[2] https://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-8700K...
The score you gave is the multi threaded benchmark that is just higher for having 6 cores rather than 4, that regular applications don't use.
If you're playing games, you'll get better performance with this i9.
But if you're playing games, you're probably not CPU-bound anyway.
I'd be interested in what niches people here work in and what CPUs they look for because of their unique needs.
We used to say this, but I don't believe it anymore. A lot of games these days make heavy use of simulation and physics engines. Games like KSP that perform continuous physics calculations are absolutely CPU-bound.
Newer intel CPUs turbo higher at stock (you could say that OC is sort of built in) so single threaded performance looks better than it really is.
Now 32nm at 4.8Ghz would be quite power hungry but that is not the issue here.
Pretty much all the progress Intel has achieved in the last 7 years since Sandy Bridge has been in achieving the same performance for less wattage focusing on mobile use cases.
(on server side Xeons have increased the core count substantially but again single threaded performance IPC has stagnated)
Thus mobile i5-8250U gets to claim that it does about the same in 15W that 2600k did at 95W.
Remember we are still on the same Skylake Arch from 3-4 years ago, just a few minor process tweaks on 14++++nm plus obviously 2x the number of cores.
https://www.tomshardware.com/reviews/nvidia-geforce-gtx-1070...
Though the point of this discussion was the CPU not being critical anymore. If you need more FPS, buy a more expensive video card, not a CPU.
Desktop productivity improves dramatically with a bunch more cores. Multiple browser tabs, background services, code editors open, VM's open... you can throw a lot of things at the Ryzen machine before you start noticing it starting to slow down.
On the 6600 machine it doesn't take much before you notice it chugging.
One example I can think of off the top of my head is gaming and transcoding (Plex server) at the same time. Even with QuickSync enabled and running, transcoding HEVC->x264 would regularly stutter and drop frames while a game is running, while also affecting the games framerate. The 1800X (without hardware accelerated video transcoding) just plows through it like it's nothing. I've even had it transcode 10-bit HDR video while playing Forza Horizon 4, not a single loss of framerate in either the game or the video.
The extra cores/threads are nothing to scoff at when software can take advantage of them.
My upgrade consisted of only the board and CPU; same RAM, GPU, PSU, SSD, etc.
This could represent a real additional long term opportunity to AMD as well, because historically Intel has been pretty fond of creating profitable product line segmentation via only allowing certain features on certain chips. And they've been mediocre about boosting interconnect capacity as well. Perhaps AMD doesn't necessarily need to match them as well in single thread performance if they can get the absolute difference down to a few percent, but be much friendlier then Intel in terms of not artificially restricting useful extra CPU features? I wonder if that becoming an area of competition again could bring some nice energy back to the market?
It's been something that's personally making me reconsider at least. I used to stick to Intel more despite price differences because of their performance advantages, but now I look at the rest of the ecosystem and start to calculate adding a few GPUs, NVMe storage, 10G networking, possibly other accelerator/utility cards and future interconnects, and the PCIe lanes seem to get sucked up a lot faster then they used to. With Ryzen/Epyc AMD is offering a full 64 lanes it looks like vs Intel's 24 here, and even with the latest Xeon I think it's now up to 48 lanes from 32? The difference there isn't nothing if overall system performance starts to depend less on just the CPU then on the overall ability to move data around and keep various more special purpose accelerators fed.
It tells you more than the spec sheet comparison.
Edit: Oh, wow, as taspeotis points out it's not even the same two CPUs in the original Intel Ark comparison (which I was working off of from memory when noting cores and threads). The i9-9700k is actually an 8-core and 16-thread (hyperthreads) CPU, so it is actually literally twice as many cores and threads. It's still better performance, but not nearly as lopsided as it even seemed before, given those resources.
Transistor count has increased from 1B to about 7B, which is roughly in line with Moore's law though...
https://www.reddit.com/r/intel/comments/9pidjh/any_9900k_ama...
The October 19th date that everyone was given thus far seems to be a fantasy. The new processors aren't in stock anywhere, and as of yesterday every single retailer said they were shipping today. After waiting weeks for the 19th to arrive, it's rather infuriating.
> or AMD’s R7 2700X is very competitive in almost every test, while they might not be the best, they’re more cost-effective.
I'm truly loving the 2700X I built last week, it's a nice bump over the 1700 at work.
The compatibility problem was probably most annoying. The 1700 didn't work with a lot of RAM when it launched, requiring BIOS settings to change. You'll still see AMD builders praising the "Samsung B-Die" (the most compatible RAM with AMD Ryzen).
The 2700x just works overall better. I'd personally consider it due to all of the little changes and little improvements.
The 9600K, 9700K or 2600X are really what gamers should be looking at. The only thing that's going to make use of a 2700X or a 9900K right now is productivity tasks (video encoding, 3D modelling, etc).
Intel still has a lead in high-refresh gaming (which is a coded way of saying they have better gaming performance that is currently bottlenecked by GPU performance), so if you're primarily a gamer it's still worth leaning to the 9600K/9700K. And the 9900K, while expensive, combines that gaming performance with a ~20-25% lead in productivity tasks. Expensive, but hands-down the fastest thing on the market for at least a year.
Threadripper is great for productivity stuff, but since it's NUMA it really doesn't perform any different than a 2700X for gaming. Also, the 2990WX has an even weirder configuration where half the dies don't have access to RAM, so it's hard to recommend except as a specialty product. 2950X is the generalist recommendation there.
Next year the Zen2 upgrade will hopefully bring AMD up to parity with Intel in gaming, and probably beat them in productivity. But it's hard to recommend planning on buying two different processors to "save money", so if that's your cup of tea, just wait until next year.
So, the recommendations look like:
60 Hz gaming => buy a 2600X
high-refresh gaming => buy a 9700K
60 Hz gaming+productivity or productivity only => buy a 2700X
high refresh gaming and productivity => 9900K
really heavy productivity stuff => 1950X/2950X/wait for 3800X/3950X next year
These days even Chrome is very aggressive about throttling background tabs, much to the consternation of some webdevs here.
https://arstechnica.com/information-technology/2017/03/chrom...
But then so is the 9900K, or even the 9700K. If you legitimately don't have anything that can use more than six threads, why are you paying for eight cores? Both vendors offer processors with fewer cores for less money.
I think basically as you're building up this tree you'd add annotations at each branch/ply which call out things the human user (opponent) could do that strengthen or weaken the position. Then as the human performs their actions, the tree is pruned/rebuilt accordingly.
The other complication is it isn't just 1 on 1, its 1 player vs ~10 players. So your move changes player 2's move changes player 3's move and so on.
0: https://www.pcgamesn.com/intel-i9-9900k-review-benchmarks#nn...
1: https://www.tomshardware.com/reviews/intel-core-i9-9900k-9th...
If you want to run the latest games on max settings at 4K/120hz, it's a different story. Personally? I remember when reviews gushed over 30fps at 800x600.
[0] https://techreport.com/review/34192/intel-core-i9-9900k-cpu-...
No one much cares about the framerate in civ6. It's about AI turn calculations in the end of the game, which is what the benchmarks people were linking are.
Overwatch is one game where in order to get 240hz, your RAM speed and timing can be a bigger bottleneck than your graphics card.
Games like X-COM have the luxury of only having to advance the simulation every so often. The rest is just rendering the game's view.
as well competitive gamers want ~140+ FPS so even when gameplay code is not super complex per frame, it is a lot of frames.
For the time being, even with latest versions, single-threaded cpu performance is the largest concern in these sims.
Alt: https://www.anandtech.com/show/13400/intel-9th-gen-core-i9-9...
This line of reasoning feels unconvincing. If there were really worried about this they should disable it everywhere.
So: cloud computing, and consumers that might be running dumb code from the internet (= i{3,5,7} and Xeon = volume production) should get chips with no hyperthreading; while HPC, locked-down workstations, and gaming PCs that probably don't have any important data on them (= i9 = low-batch-size production) should get chips with hyperthreading.
For HPC it's usually turned off, and I think for gaming and workstations it also has little if any use.
Also, HPC doesn't necessarily imply a uniform workload; it just implies that all workloads are trusted. Really, modern HPC clusters are still "multi-tenant"; they're just tenants who are all e.g. employees of the same company, or researchers at the same university. So you can describe them, from a security perspective, as one "tenant." None of them would have anything to gain by doing a side-channel attack against one-another.
Does this give AMD an advantage with Ryzen and Threadripper, since AMD supports hyperthreads?
The Intel i9-9900k sells for $550+ to $580+, requires an added cooler, and 8c/16 threads.
The difference is that the i9-9900k can single-thread turbo to 5GHz, which makes a big difference in single-threaded situations. The i9-9900k also has 256-bit AVX pipelines, while AMD Ryzen is 128-bit based AVX pipelines.
So any AVX-heavy code (H264 rendering, Blender benchmarks, etc. etc.) will be hugely beneficial on the Intel platform.
-----------
AMD still wins price/performance. I'd personally still take a Threadripper 2950x for example. But the i9-9900k is impressive because of that 5GHz clock, even if its a bit restrictive to use.
Anyone who cares about maximum single-threaded speeds is looking at the 5GHz clock. That's huge.
minor nitpick: the 9900k can actually turbo to 5GHz on two cores simultaneously.
But I wouldn't doubt that things like Video encoding / compression uses this. Again, I was not saying it's not valuable just saying that for my use case where I don't encode videos for a living or run a large Handbrake/Plex video farm / library the i9 cpu is not worth as much to me as the cheaper, cooler running Ryzen 2700.
The i9-9900x does. (x, not k). Confusing? Yeah. Blame Intel's marketing.
Still, the i9-9900k is faster at AVX2 (256-bit), which is useful in H264, Photoshop, and a number of graphical productivity applications. And you can see the benefits of it: the i9-9900k gets close to Threadripper performance on Handbrake due to the AVX2 support, even with half the cores.
Media applications makes sense to use such cpu features.
The 9900k is coffee lake v2. The 9900x is skylake (server) v2.
Two totally different chips.
2-memory channels is what is expected on a normal, consumer platform. Especially since DDR4 effectively doubled the bandwidth from DDR3 anyway.
In effect: DDR4 2-channels is roughly the same speed as DDR3 4-channels. So that's how RAM scales: it gets faster without necessarily needing more sticks.
In two or so years, DDR5 is going to double-bandwidth again. So 2-channel DDR5 will be the same bandwidth as 4-channel DDR4.
Latencies are all going to be the same though, but that's the nature of the technology.
----------
If you need 4-channels of RAM, x299 exists (and x399 for AMD Threadripper). If you need 6-channels, you got Xeon Scalable, 8-channels got EPYC, 12-channels (or more) for higher-end Xeons. So these high-end systems exist for people who need them.
But most consumers are fine with the ~40GB/s speeds that 2-sticks of modern DDR4 get you today.
Threadripper or cheap EPYC are looking quite attractive to me because of the high bandwidth per core and price-point. I just wish Intel would also compute on this point.
Intel's main price/performance competition are Dual Xeon Silvers. Its not talked about very much, but 2x Silver 4114 (https://www.amazon.com/Intel-Xeon-4114-Deca-core-Processor/d...) isn't really that bad. 10-cores x2 == 20 cores at $1500 total.
That's 12-memory channels too (6-channels per socket). You'd have to buy an expensive dual-socket motherboard, but its relatively high-end / server-class for a reason.
Xeon Gold and Xeon Platinum are just WTF with the pricing though. I guess those are for people who stopped caring about price/performance.
AMD EPYC still seems better on a price/performance front than Dual Xeon Silvers, if only because you can get a single-socket EPYC with 8-channel RAM. But Xeon Silvers aren't really that far away.
Isn't the Epyc 7281 basically the same price with 32 cores and 16 memory channels (16 cores and 8 channels per socket)?
10-cores x2 == 20-cores on 2x NUMA nodes.
2x EPYC 7281 basically gives you 8x NUMA nodes. Your threads are really spread out, so its bad for major workloads like Databases (which may communicate heavily, and also benefit from the "combined" L3 caches on the Intel Xeon Silver setup)
In many ways, I'd argue that the "real" competitor to the Dual-Xeon Silver 4114 setup is the single-socket EPYC 7401P for $1200ish.
The EPYC 7401P gives you 4x NUMA nodes with 24-cores / 48-threads total, fits into a single-socket motherboard (way cheaper). This minimizes the NUMA crosstalk issue, still gets you a respectable 8-channel memory system and a large core-count.
The EPYC 7401P is the price/performance champ. Still, AVX512 or Database purposes may keep dual-Xeon Silvers in the running.
As soon as you add the second node for a workload with poor locality like that, half your accesses are on the wrong node. No matter how many more nodes you add, the worst it can cost is that much again.
The unified L3 is also probably immaterial when you have a large database like that which exceeds the L3 by hundreds of gigabytes and it's ~100% cache misses either way.
On the other hand, you have 60% more cores to make up for it.
I'm not saying there are no circumstances or workloads where the Intel system makes sense, but it's kind of telling that it's a matter of finding them as exceptions to the rule.
For the general purpose things like running a bunch of unrelated VMs that aren't affected by number of nodes much if at all (or benefit from them because one misbehaving guest or process thrashing all the caches and flooding memory bandwidth only impacts a single CCX/node), it seems like an obvious choice.
It's also going to be interesting to see how databases optimize for multiple NUMA nodes now that they're common. There should be ways to determine which parts of the database are on which node and then prefer to dispatch queries to cores on the same node, or keep copies of the highest hit rate data on multiple nodes etc.
Which is just as legitimate as AMD's 2x4 memory channels on their quad-NUMA single-socket solution IMO.
I did evaluate about 2 months ago if to go for a 2700X or the usual Intel, and based on the benchmarks I saw, the 2700X is still a bit too weak when running single-threaded SW => I then bought an i7-8700 (no "k") just to be "safe" when running semi-single threaded SW (e.g. my current prototype uses 10 threads but has a single coordination bottleneck).
Let's see what happens in ~January (I think that's when AMD is supposed to present the new CPUs that are supposed to be the "real" next-gen with higher IPC?)
When the GTX 1080 came out, I bumped up to that with a 1TB NVME drive (Samsung) which has been fantastic in limited use. Mostly notice the difference in my npm scripts (ci and build). I'm mostly just itching to upgrade more than feeling pain at this point. I'm pretty sure the Ryzen 2 CPUs (not 2xxx) will be sufficient, as will the mid-upper Intel. However, unless Intel brings their pricing into better competitive range, I may go AMD on my next build.
Linux runs great (I'm using Ubuntu 16.04 desktop currently)