Tag: fan

Heat sinks and fans could go the way of the Dodo

dodo The current method of cooling computers is about to get a rethink.

The Defence Advanced Research Projects Agency (DARPA) is supporting new research to provide on-chip liquid cooling in field-programmable gate array (FPGA) devices and it looks like the technology could be easily adapted for CPUs and GPUs.

This has the potential to reduce the size of devices, allow for chip stacking, dispense with heat sinks and fans and significantly extend the life-span of chips.

Speaking at the IEEE Custom Integrated Circuits Conference, Thomas Sarvey, from Georgia Institute of Technology (GIT) presented the paper with the catchy title “Embedded Cooling Technologies for Densely Integrated Electronic Systems.”

What they managed to do was get rid of the heat sink atop the silicon die by moving liquid cooling just a few hundred microns away from the transistors.

The technique involves cutting microfluidic channels into the die of FPGA devices, which were chosen for the research and trials because of their flexible configuration and extensive use in the military.

This locates the cooling just microns from the problem, and even allows for the possibility of chip-stacking, which very few devices currently have the room or efficiency to achieve, given the necessity to dissipate heat from a central locus of adjacent chips.

The group successfully developed a standard demonstration test, including one for DARPA officials, in which a converted FPGA with bespoke Altera-supplied architecture operated, with no other cooling, at less than 24 degrees Celsius, and was compared to an analogous air-cooled device operating at 60 degrees Celsius.

On-chip liquid cooling also opens up the possibility for a new level of compactness in device design, which frequently has to use available surface space for dissipation purposes.

 

Intel’s fanless chip gets a fan

While Chipzilla is marketing its anorexic fanless Core M reference designs at Computex this week, it seems that that the company does think that the designs will actually run better with a fan.

Cnet said that Intel showed off a design for a docking station that includes a set of fans to add some airflow into the Core M.

The reason is that to allow for greater performance while using such a system at your desk. It works by targeting airflow over the front and rear surfaces of the tablet. CNet tells us that the dock was switched on at the demonstration, making very little noise and the airflow was hard to notice. However it does defeat the purpose of having a fanless computer an awful lot.

It is a bit like offering a punter a hybrid car but telling them that it will go faster if you put petrol in. The performance gain is not small either. Intel’s testing indicates that running a Core M system with the dock would boost performance by around 30 percent.

Still if Intel does hold to its promise to have all wireless systems including docking, charging and displays, in 2016 the idea will be toast, unless it tells us that the way to improve wireless performance by 30 per cent is by connecting it with a wire. 

AMD says Kaveri up for pre-order

AMD has announced that two of its A-series “Kaveri” chips are now available for preorder, and will be in the shops on January 14.

Inside the new A10-7700K and A10-7850K are 4 CPU cores and 8 Radeon R7 GPU cores. Now, however, AMD is simply referring to both as “compute cores,” and failing to distinguish between the two. Both chips contain 12 “compute cores”.

At the CES show at Las Vegas AMD also announced three new mobile GPUs for notebooks, including the R9 290 GTX.  and the R5 M230.

The chipmaker also talked about its the Discovery Project, a sleeve that designed to plug into and enhance AMD-based tablets, and an envelope-sized PC prototype that uses one of AMD’s next-generation ultrabook chip, code-named “Mullins.”

What appears to be happening is that the lines between CPUs and GPUs are blurring, as specialised tasks once performed by high-end microprocessors are being passed over to more specialised GPUs serving as coprocessors.

Lisa Su, senior vice president and general manager, global business units, told an audience at the AMD press conference that the number of APUs is expected to grow to 160 million units by 2015, doubling in the last two years

She said that Kaveri is the best and most powerful APU that AMD has ever put out there,” Su said. She ran a brief clip of what of Tomb Raider, running on Kaveri to prove it.

Kaveri is the first AMD processor to support HAS and Su also announced some aggressive benchmarks in a bid to stick two fingers up at Chipzilla which was also having a bit announcement.

The new chips should offer a total compute power of 856 gigaflops which is fairly floppy. Both chips will have support for AMD’s TrueAudio technology, Graphics Core Next architecture with AMD’s Mantle technology, and will be powered by AMD’s Steamroller processor architecture. 

Retail Radeon chips still go slower than review copies

Chimpzilla’s Radeon R9 290 and 290X review copies appear to be performing better than retail versions.

Techreport penned a review which said that AMD upended the high-end graphics market by offering performance competitive with Nvidia’s existing products at substantially lower prices. The R9 290X captured the overall GPU performance crown, wresting it away from the GeForce GTX 780 and Titan.

However, some of its readers said that Radeon R9 290X cards bought at retail do not seem to perform as well as the review units AMD supplied to the press.

The retail R9 290-series cards had the same basic hardware and specifications as the review samples, but at least some retail R9 290-series cards seemed to operate at lower clock speeds.

Techreport got on the blower to AMD which investigated and found that the fans on some retail cards were not spinning as fast as expected, and the reduced cooling capacity resulted in lower clock speeds.

It issued a fix in the form of a software update. The Catalyst 13.11 beta 9v2 driver sought to equalize blower speeds from card to card by monitoring RPM directly, thus hopefully improving performance on retail cards that seemed to lag behind.

Techreport wanted to test a retail card but could not afford it, until Nvidia offered to buy a couple of retail R9 290X cards for them to test. Nvidia did not touch the cards which were shipped from New Egg. It looked like Nvidia were happy to foot the bill because it was fairly sure that AMD would look very bad.

It tested two Radeon R9 290X cards: one from HIS and the other from Sapphire. Apart from the stickers on the cooling shrouds, the two look to be identical and to the two R9 290X review samples.

The testing showed that AMD’s software fix to equalize blower speeds in the Catalyst 13.11 beta 9v2 driver release definitely improves the worst of the low-clock-speed problems. The fix appears to raise fan speeds overall for 290X cards, slightly for our initial review unit and more dramatically for our HIS retail card.

However the initial R9 290X review unit was superior to the two retail cards tested. Even with the blower speed fix working the first review unit runs at 5-10 per cent higher clock speeds than the retail cards, depending on the workload. That deficit translates into a 5-10 per cent advantage in frame rates, though usually toward the lower end of that range at 4K resolutions and lower voltage.

The initial review was skewed so that the Radeon R9 290X so that the card’s performance was better than it should be and it was quieter.

It thinks that AMD chose to include some of its very best Hawaii GPUs aboard the review samples it supplied to the press. 

AMD rushes out a new Radeon R9 290 fix

AMD has quickly rushed out another Catalyst driver update which fixes an embarrassing bug in its Radeon R9 290 series.

Last week tech magazines noticed that the Radeon R9 290 series was making low level operational changes that will have a direct impact on the power, noise, and performance of the cards.

It would appear that fans were running too slow, which could have resulted in lower performing cards.

The new catalyst update fixes a problem with the variability in the fan speeds on the cards. Specifically AMD has changed the algorithms for how their drivers handle the fan speeds and overridden the BIOS defaults.

Currently fan speeds are based on percentages, but now it seems that AMD has switched to controlling fan speeds on an absolute basis, using the measured RPM of the fan.

For the release of the new driver, AMD said that it had identified that there’s variability in fan speeds across AMD R9 290 series boards. This variability in fan speed translates into variability of the cooling capacity of the fan-sink.

Using PowerTune technology this could be corrected this variability in a driver update and will normalise the fan RPMs to the correct values.