Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Those SD835 Cinebench results look good to me: this certainly goes through
emulation, and given that it's very CPU intensive, it's a worst case. In
fact I'm surprised by how good that score is! I'd expect the native score to
be at least twice better, but we'll never know.
What this shows on the other hand is that if x86 compatibility of CPU
intensive programs is needed, Windows on ARM is surely not the way to go. For
other apps time will tell.
Recognize the green and white book under the left monitor?
Anyone know what the Programmable solutions group is? Is that Altera?
Altera will operate as a new Intel business unit called the Programmable Solutions Group (PSG)
Heise.de has run SPEC CPU2006 using various compilers on both Epyc and Xeon (in German).
It seems people forget Intel had a very similar issue on Skylake.
Intel Skylake bug causes PCs to freeze during complex workloads
It was patched by a microcode update.
The news is up and it has nothing to do with Intel:
MediaTek, Inc. (TSE: 2454; hereinafter “MediaTek”) and NavInfo Co., Ltd. (SZ: 002405; hereinafter “NavInfo”) today independently approved the signing of a Strategic Cooperation Agreement and a Framework Agreement (hereinafter, the “Agreements”) by the Board of Directors of each company. Per the Agreements, NavInfo plans to wholly acquire AutoChips, Inc. (hereinafter “AutoChips”), a Chinese subsidiary of MediaTek, for a purchase price of US$600 million. Completion of the above is scheduled for 4Q16, subject to various relevant regulatory approvals. Concurrently, MediaTek plans to invest in or co-invest with NavInfo, an amount of no more than US$100 million, to cooperate strategically in the fields of Automotive ICs and the Internet of Vehicles (hereinafter “IoV”).
Broxton was supposed to show up in 2015, but this thing is MIA. Didn't even make an appearance at MWC '16, probably will power the ZenFone 3 and that's about it.
Also see how 176.gcc is 2000 for both A53 and A57, while it's 3000 and 4000 for 253.perlmbk. Statistically that's almost impossible to have 4 multiples of 1000
Definitely dubious results.
They ran both 32-bit and 64-bit versions of SPECInt2000 on the Exynos 7420 (Cortex A57), so I would think AT used the 64-bit build for Kryo.
What's your take on these results? Kryo looks good in pretty much everything but SPECint, but IMO SPECint is the best, most representative CPU test in this suite.
There is no excuse for SPEC not releasing the retired SPEC_CPU_2000 for free now
https://twitter.com/FPiednoel/status/665246003390345216
https://twitter.com/FPiednoel/status/665249308346724352
Eh? How can he take a binary from iOS and run it on an Intel machine?
He was right that the 8 ARM core design is stupid lazy engineering. Apple A8/A9 and Qualcomms 820 prove that.
Atom is a good chip as well.
Definitely not the ~1% that GB3 author claims. Ouch.
I hope my answer to Paul clarifies what I think.
To add one point: I don't care at all about Geekbench MT score. As you
wrote those 8 cores (or even 4 cores for phones) SoCs are dumb.
You'll understand that there are some things I can't obviously say due to
NDA.
That being said, I'm fully aware of Geekbench shortcomings as they are
pretty obvious:
- too many compression/decompression benchmarks
- too many benchmarks using dedicated instructions
- too small datasets.
I still find it useful as long as you examine individual scores (Lua and
Dijkstra are interesting IMHO) and keep in mind that drawing a conclusion
on a chip performance based on a single benchmark (no matter what that
benchmark is) is utterly dumb. I think John Poole is really trying to
improve his benchmark and accusing him of having an agenda isn't fair.
My understanding is that he's taking inputs from various companies for
GB4 and I guess our preferred one is being listened to.
As far as SPEC 2006 goes memory requirements did not allow Apple phones
to run it (64-bit requires 2GB as you know). Also one has to be careful
with SPEC, given that many companies have dedicated huge efforts to tune
their compilers (icc on libquantum is the obvious example), so to use it
to compare different CPU's it would be fair to use gcc. Perhaps we'll
see someone qualified do that at last (not a random Anandtech guy
unable to properly compile SPEC as was the case in the Exynos 7
review...).
So you won't read me saying how great Geekbench is, but you won't read me
saying it's a PoS. It provides some useful information. Just one data
point, far from enough to draw any conclusion about an alleged catch up
of Intel by Apple.
I don't need input from Linus or any other about Geekbench. As a
benchmarking engineer working in a company that bought a license, I have
access to the source code
I think you're being a little too hard at Geekbench author. Here is what
he wrote on Realworldtech:
In our testing the performance difference between the mobile and the desktop workload sizes is less than 1% on the same hardware.
The two datasets was the only way we were able to produce a benchmark that ran on both desktop and mobile and had a reasonable runtime on both given the huge delta between mobile SoCs and desktop CPUs. This delta has decreased dramatically and so Geekbench 4 will remove the distinction between mobile and desktop workload sizes.
Let's give him the benefit of the doubt.
I heard that Mac sales were up but don't have numbers.
You still can't deny it was partly broken by several compiler vendors. That's
the price to pay for the source being available
Also only relying on SPECCPU alone is very misleading. As an example, the
Icache footprint is ridiculous, much lower than a simple browser benchmark.
They must have referred to a synthesized netlist for power simulation. Everything else would be really inaccurate, to say the least.
Looks like AVX-512 is only for the Knights * devices for now - I don't see an announcement for future CPUs for AVX-512 yet.
Also there now seems to be some sort of ban on the Chinese labs that built the current reigning monster from Intel Phi.
http://www.hpcwire.com/2015/04/08/chinese-supercomputing-orgs-placed-on-us-entity-list/
I wonder if someone finally figured out that the Intel stuff is about to explode in performance and decided to keep it at home and out of the "wrong hands."
That said, I had a look at the specs of the Intel stick and they're too low. The 32 GB of flash is okay but 2 GB of RAM is probably too low. It would be nice if there were additional models of the Intel stick with more RAM and flash - though you can add an SD card for more storage. The price of the Linux version is $89, not $110.
In this case Intel isn't promising anything like that. They say they'll use Core M (or did they mean the little m?) at the top end which is a much more reasonable aim even if I rather worry that ARM now has to be countered with Haswell chips. Perhaps they meant Broadwell instead of Broxton? I get confused enough with all the names.
Do you really think SoFIA will be able to achieve that in 28nm and without FinFET in its thermal envelope? It is much more likely to be limited to 1.5GHz at the very best.
There are much better Snapdragon 810 Geekbench MT score close to 5k: http://browser.primatelabs.com/geekbench3/1882896
A57 was announced by ARM October 2012. How long did it take until the first designs arrived in the market? 2 years maybe?
No matter how bad AnTuTu is (and it really is bad, even Geekbench is
better), the scores posted by HotHardware make little sense: see how
the Z3770 tablet is significantly worse for CPU Int than then Dell Venue
while it is much better for CPU Float, even though the cores are the same.
Also other sites have reported >55K AnTuTu scores for NVIDIA SHIELD Tablet.
IMHO AnTuTu definitely is a useless benchmark.
Prior to the deal, there were reports that Spreadrum had let go its entire development group.
Spreadtrum tells me they will continue to make ARM-based mobile ICs following the investment by Intel in Spreadtrum’s owner, Tsinghua Unigroup.
Intel said it would pay $1.5 billion for 20% of Tsinghua Unigroup earlier this week.
Intel’s CEO said at the time that: “This partnership will also enhance our ability to support a wider range of mobile customers in China and the rest of the world by more quickly delivering a broader portfolio of Intel architecture and communications technology solutions.”
It was said that x86-based mobile ICs would be designed by Spreadtrum and Intel and would be marketed by both companies sometime next year.
However it seems that even $1.5 billion doesn’t buy exclusivity and Spreadtrum will continue to design and market ARM-based mobile ICs.
These are not just partners, Intel has taken a 20% ownership stake in them.
That's a better balanced design that delivers what was promised from Core-M, a fanless genuine Core experience.
As a side note though, performance in Chrome is poor. In fact, Chrome takes a much higher toll on this hardware, pushing the CPU to max-load when streaming video from Youtube or even when casually browsing, and that leads to choppiness. From what I’ve read, Google decided to always stick to the VP9 codec when it comes to handling videos in Chrome, while IE/Firefox rely on the older H.264/265 codecs and that seems to help a lot (see this link for more details on these codecs). But that doesn’t explain what happens when browsing through ordinary pages with no videos. So for now, I could summarize that Chrome is just not optimized for this Intel platform and that takes it toll when it comes to both performance and battery life. Until that’s getting fixed, stay away and go with IE or Firefox.
China probably cancelled contracts for chips and the banks demand pay-back of loans secured by the contract.
This may explain the dramatic fall in Samsung’s revenue.
PC Pro Yoga 3 Pro review
This appears to be a production model. Much better battery life here with 8 hours though the performance is hard to gauge because of the way they benchmark. Their baseline is a 3.4GHZ Core i7 2600k and the core M scores .45 of that which seems to be pretty decent but their benchmarks are non-standard and a bit opaque. They do rate its responsiveness very well though. I think this proves that the other pre-production model had some issues. It is obviously not as fast as a Core i5 4200U which the review seems to want to compare it too but that would never fit into the Yoga Pro 3 frame.
Denver is a 64-bit ARM chip and the SHA subtests indicate it is being tested in that way.
The second
step is leverage the very high degree of uniformity of architecture
and features in x86 compared to the ARM tower of babel to allow
extra performance and/or capabilities to be delivered by apps so
that eventually consumers prefer to buy x86 based phones.